Mode of the Reposturgeon!

It was inevitable, I suppose; reposurgeon now has its own Emacs mode.

The most laborious task in the reposurgeon conversion of a large CVS or Subversion repository is editing the comment history. You want to do this for two reasons: (1) to massage multiline comments into the summary-line + continuation form that plays well with git log and gitk, and (2) lifting Subversion and CVS commit references from, e.g., ‘2345’ to [[SVN:2345]] so reposurgeon can recognize them unambiguously and turn them into action stamps.

In the new release 2.22, there’s a small Emacs mode with several functions that help semi-automate this process.

Fear the reposturgeon!

214 comments

  1. Joking aside, this is exactly why the mentality that “comments don’t have to adhere to strict rules, because they’re for humans to read, and we can figure out what you really meant” (expressed on the asterisk thread) is so short-sighted.

  2. Have to agree with the monster. I’ll admit to having had a good many comments, back when I was blogging, that made no sense at all. AND, I’ve read a good bit of code with comments that made just as much sense as the spammers on my (late and unlamented) blog.

    Please, if you are coding and leaving comments, leave something your 80-year-old grandmother might have a chance of understanding? I’m not expecting English. Leave it in your local language, Tagalog, Hindi, Chinese, Japanese, even Australian… just leave it in something that can be translated by a reasonably intelligent programmer? Hackers aren’t everywhere – let the programmers in sometimes?

    Caveat: I’m about 1/3 hacker – I can do a lot that others can’t (well, much of present company excepted) but I’m not a hacker. I’m, at best, an adjuster. I fix things. And I just LURVE me some spaghetti code. I suspect the Flying Spaghetti Monster approves of it too. And, of course, my goddess Eris. Anyone want an apple? Or a hot dog? :-P

    1. >I thought the classic line was ‘every program expands to read email’, not send it.

      Damn. You’re right – and here I had just thought of a good rationalization for making reposurgeon send mail. Sometimes repositor….OH SHIT.

      reposurgeon already reads mail.

      Or, well, RFC822-format mailboxes, anyway. Read up on the ‘mailbox_in’ command.

      *shudder* JWZ was right.

  3. Has any progress been made in getting tools other than reposurgeon to understand action-stamp syntax?

    On another note, I’ve been rereading your posts on the forge problem. Is it your intention that reposurgeon eventually learn to translate issue references to other formats—perhaps to Roundup/VaporForge-style? And what of the complementary problem: will forgeplucker learn to rewrite repository references?

    1. >Has any progress been made in getting tools other than reposurgeon to understand action-stamp syntax?

      No, but I have a to-do item to ship a gitweb patch that will do this. Shouldn’t be too difficult, but I’ll have to write Perl to do it. Is this something you could take on?

      >On another note, I’ve been rereading your posts on the forge problem. Is it your intention that reposurgeon eventually learn to translate issue references to other formats—perhaps to Roundup/VaporForge-style?

      No such plans, but I could be talked into it with a well-presented technical case. That kind of thing is easy to do within reposurgeon’s framework.

      >And what of the complementary problem: will forgeplucker learn to rewrite repository references?

      Unclear. I’m not sure forgeplucker is the right level for that.

  4. And I’ve been thinking of how best to formulate Raymond’s Corollary to Zawinski’s Law: “Every programmer’s tool grows until it includes an EMACS mode”.

  5. >>Has any progress been made in getting tools other than reposurgeon to understand action-stamp syntax?

    >No, but I have a to-do item to ship a gitweb patch that will do this. Shouldn’t be too difficult, but I’ll have to write Perl to do it. Is this something you could take on?

    Eric, what would you want from gitweb in understanding action-stamp syntax in commit messages?

    One of those days I’ll go back to working on gitweb (I promise). I just have to set up down and up email to Gmail in MUA-agnostic way (msmtp or sendmail, or sth. for up, but I don’t know what for down).

    1. >Eric, what would you want from gitweb in understanding action-stamp syntax in commit messages?

      Nothing complicated. Each string that’s a valid action stamp needs to tun into a hyperlink to the commit that matches it.

  6. >>Eric, what would you want from gitweb in understanding action-stamp syntax in commit messages?

    >Nothing complicated. Each string that’s a valid action stamp needs to tun into a hyperlink to the commit that matches it.

    Ah, so it would be just extension of an existing gitweb mechanism to turn SHA1-looking fragments into hyperlinks (currently without checking upfront for performance reasons).

    1. >Ah, so it would be just extension of an existing gitweb mechanism to turn SHA1-looking fragments into hyperlinks

      Yes, exactly so. Probably almost trivial for anybody familiar with the codebase.

  7. Hijack, yes, but

    http://www.comscore.com/Insights/Press_Releases/2013/3/comScore_Reports_January_2013_U.S._Smartphone_Subscriber_Market_Share

    US market share rising (3.5%) for Apple declining for everyone else, including Google (Android: – 1.3%)

    As a reminder, it was 11 months ago when Eric rejoiced about Android obtaining 50.1% market share, and predicted a complete market tip due to super-majority effects. http://esr.ibiblio.org/?p=4265

    I know that Eric now rejects the comscore data, conveniently this occurred just about the time it no longer supported his assertions.

  8. The new HTC One unveiled last month is increasingly feeling like last-chance saloon for the troubled Taiwan handset maker. Today HTC noted that its sales for the month of February fell by nearly 44% to 11.37 billion Taiwan dollars ($384 million), from NT$20.3 billion for the same month one year ago. Looking at the bigger picture, that NT$11.37 billion is barely higher than what HTC made in January 2010, when it reported NT$11.12 billion in sales.

    Source: http://techcrunch.com/2013/03/06/htc-sees-lowest-sales-since-jan-2010-worst-revenue-decline-in-4-months-as-february-sales-fell-44/

    This follows a quarter in which profit fell 91 percent. And that followed a quarter in which profit dropped 79 percent.

    This is a two-horse race. Apple is winning the U.S. and Samsung is winning the rest of the world (with Huawei being an up-and-coming wildcard). It’s that simple.

    Yes, there are other companies which help extend Android’s market share but none of them can seem to make any money. That’s not a viable long-term strategy.

    But Android is winning.

    Oh, and this will (not) be the year of linux on the desktop. Miguel de Icaza loves his Mac because “it just works”. http://tirania.org/blog/archive/2013/Mar-05.html

  9. Google last gave data for Android activations on 12 September 2012, when it said that there had been 500m activations and the run-rate was ‘over’ 1.3m a day. A week earlier Eric Schmidt gave the same 1.3m number and cumulative activations of 480m. Those numbers are not, obviously, consistent (the daily run rate between the two announcements would actually be 2.86m), but then they’re round numbers at scheduled events, so one must make allowances. ?

    Since then, though, silence. ?Google has repeated the same 1.3m number from time to time, but there is no new data. (As an aside, Google’s monthly developer stats for Android device screen sizes haven’t been updated since the end of September 2012.)

  10. The ‘Android has problems’ narrative is very clearly established (as established, of course, as the ‘Android is selling in vast numbers’ narrative). There’s the fragmentation by OS version, the fragmentation by hardware, the consistently lower engagement and monetisation, and of course the financial weakness of all the branded OEMs except Samsung, which has an equally unhealthy (for Google) 45-50% or so of Android device sales.

    However, it is far from clear to me that problems for Android developers and OEMs are the same as problems for Google.

    As I see it, Google really has three strategies for Android, as it has developed over time. 

    First, there was the publicly stated objective; to make sure that Google was not somehow shut out from the mobile internet by a dominant OS provider that chose to exclude it. Originally (i.e. pre-iPhone) this fear was directed at Microsoft; subsequently it may also have been consciously directed at Apple. 

    Android has been a complete and unambiguous success on this front: it is extremely hard to see how any one company could now control the mobile OS market at all, let alone try to exclude Google. (The exception, of course, is China, to which I will return.)

    Second, whether deliberate or not, Android has had the effect of hugely increasing the number of people with access to the mobile internet. Just as Wintel 25 years ago powered an army of cheap PC ‘clone’ makers churning out tens of millions of cheap commodity PCs (in-turn powering the linux revolution), Android and a small group of mobile chip companies (mainly Qualcomm, EMP, Mediatek, Spreadtrum) have enabled a flood of cheap commodity smartphones and tablets. Do a search on Alibaba for ‘Android 2.3’ to see what I mean – the entry price for an Android smartphone is now $45 wholesale, with wifi tablets not much more and a vast range of other devices (ersatz netbooks, in-car PCs and DVD players, set-top-boxes and lots else besides) following on behind. 

    These devices in turn are percolating out to emerging markets around the world and to prepay customers everywhere, quite apart from the relatively small numbers of people in developed markets buying relatively high-end devices like Samsung’s Galaxy S3 (which made up under 10% of Android device sales in 2012). All of those devices represent more people and more time online, and that means search, which means Google web search and revenue for Google. 

    On this front too Android has been a huge success, and many of the criticisms of Android as an ecosystem are of limited relevance to Google. A forked, fragmented, Android 2.2 phone in Jakarta with no Google services on it still accesses the web (if it has a data plan, of course) and drives search volumes, and it is still a great product for a laborer with $50 to spend. The decline of HTC matters little compared to that. Even the dominance of Android sales by Samsung, though arguably unhealthy, doesn’t fundamentally threaten these strategic gains so long as the ‘cheap Chinese’ are providing a flood of alternatives. 

    Equally, iPhones, which have around 20% of the smartphone market, do almost all of their web search (outside China) on Google. Given the fact that $650 phones tend to be bought by higher-end consumers than $100 or $300 phones, it is quite possible that iPhones generate more advertising revenue for Google than all Android phones combined. 

    Then we come to the third strategy.

    An Android device, properly signed into a Google account and running all the Google Apps, generates an endless stream of little bits of ‘signaling’ information, way beyond what Google gets from a desktop search user even if they’re using Chrome. It knows where you live and work, how you commute – and which phone numbers on web ads you dial. Unlike a web browser, you are probably always signed in to Google, so all of your interactions with Search, Maps and anything else can be linked together. (This, of course, is also the main purpose of Google Plus.) 

    Google Now is just one manifestation of this: doing useful things like telling you that your meeting will take 45 minutes to get to because there’s heavy traffic, so you should leave now.  But this value flows both ways: Google is giving you useful tidbits, but it is also mining far more data than it would get from a PC user – data to improve search relevance and advertising relevance. 

    In other words, Android, like Plus, allows Google to tie searches and advertising to individual people and places. In the long term, the data that Google gets from Android users is probably just as important as Pagerank in understanding intent and relevance in search. 

    Hence, the real structural benefit to Google from Android now comes from the understanding it gives of actual users, and the threat comes from devices that do not provide this data – even if, like the iPhone, they do provide plenty of search traffic. 

    Obviously, Google’s access to this data on non-Android platforms is only partial, but the problem also applies to significant parts of the Android base. In China the problem is near-absolute. Google services are mostly blocked anyway, almost all Android phones ship with few or no Google services installed (i.e. they are based on the open source AOSP version of Android) and Google Play, where it is installed, appears to have a very low share of actual app installs. Hence, Google gets close to zero practical benefit from the explosive growth of Android there. There is a similar issue in other emerging markets – a significant portion of those $45 handsets skimp on Google apps just as they skimp on IMEI numbers. 

    The clearest expression of this is in tablets. Google gets no data from a Kindle Fire, only web search traffic. More importantly, at scale, it gets no data from many of the generic Chinese Android tablets that are starting to well out of China at $100 or less. These devices are like dark matter: everyone suspects there are a lot of them, but no-one quite knows (publicly) just how many. I’ve seen credible people claim it could be well over a hundred million units in 2013. Possibly much more. How many of these devices will have Google Play? How many users will install Google Maps? How many will come with a third-party web browser (from Tencent, say), one or two of the dozens of major Android app stores operating in China, or Amazon’s app store?

    In view of the above, Apple’s “Maps” debacle seems like an attempt at doing the right thing for consumer privacy.

    Someone described Android as an unguided missile: very powerful but spiraling semi-randomly with no clarity on where it would land. There is the fragmentation issue, and the weakness of most of the OEMs. There is the threat of Amazon or Samsung forking the platform. But there is also the threat that an increasing number of Android devices might have no more connection to Google than does an iPhone.

    To put that another way, Google’s penetration of Android is as important as Android’s penetration of the handset market. 

  11. As an aside, Google’s monthly developer stats for Android device screen sizes haven’t been updated since the end of September 2012.

    I dunno why that matters…platform versions matter more to me and that was last updated March 4.

    You have to handle the range of screen sizes/densities anyway and normal size + hdpi and xhdpi covers 75% of the market. If you target 3.x+ then you essentially need to handle 320dp (normal handsets), 600dp for 7″ tablets and 720dp for 10″ tablets.

    The biggest annoyance is that it’s hard to get pixel perfect layouts like you can on iOS but 4.x is a lot better.

  12. >To put that another way, Google’s penetration of Android is as important as Android’s penetration of the handset market.

    I’d generally agree with this. I’m not sure, however, that I want to actually see Google having a lot of penetration of Android. From the consumer perspective, I’d say the ideal would be for Android to have 90%+ penetration of the mobile market, but for no player in the Android market to have more than 50% penetration and no two to have more than 75%-ish.

    A situation like IBM with the PC, where the parent company eventually totally loses control of the platform, but the platform is dominant for decades after, wouldn’t be at all unwelcome. That said, I wouldn’t be disappointed if it didn’t happen either, as long as Android’s penetration of the overall market is high enough that people don’t get vendor locked into the Apple platform or anything like it, and as long as Google’s penetration of Android is low enough to keep them honest with regards to privacy, or at least to give users a safe-haven if they don’t stay honest.

  13. @LeRoy
    “Since then, though, silence.”
    “The ‘Android has problems’ narrative is very clearly established (as established, of course, as the ‘Android is selling in vast numbers’ narrative). ”

    Android sales in 4Q2012 were 147.3 million. For 92 days, that is 1.6 million per day. Up from 1.3M/day in September, mid 3Q2012. So, an increase of 100K per month. At this rate, I expect them to reach 2M/day in April. And Android will reach 1B sold this summer.

    Why Google stopped giving out point release activation updates seems to be unknown outside Google. Maybe they are waiting to make a smash impact, 1 to 2 million per day?

    @LeRoy
    “Someone described Android as an unguided missile: very powerful but spiraling semi-randomly with no clarity on where it would land.”

    I see it aimed at every pocket on earth. This summer there will be more mobile subscriptions than humans. There are already over 5 billion phones in use by over 4 billion unique users. They will all switch to smartphones in the next 5 year or so. As it looks now, +95% of those smartphones will run Android.

    The Annual Mobile Industry Numbers and Stats Blog – Yep, this year we will hit the Mobile Moment..
    http://communities-dominate.blogs.com/brands/2013/03/the-annual-mobile-industry-numbers-and-stats-blog-yep-this-year-we-will-hit-the-mobile-moment.html

  14. @LeRoy
    To see how deep Android is in trouble, it is instructive to look at the numbers below (taken from Tomi Ahonen’s quarterly numbers)

    The cumulative numbers for Android as of 4Q2012 are 740M sold. Two quarters like 4Q2012 would bring the total well over 1B. iOS has a slow increase with large seasonal variations. WP is noise.

    @LeRoy
    “I know that Eric now rejects the comscore data, conveniently this occurred just about the time it no longer supported his assertions.”

    I suspect the USA networks manipulate sales to get a 50/50 market share for iOS and Android. Price manipulations cross subsidies are very obvious in the USA market. That makes the ComScore numbers meaningless for global consumer preferences.

    Estimated mobile phone sales numbers in millions, recalculated to absolute numbers sold per day for Android:
    http://communities-dominate.blogs.com/brands/
    (I have a list of links, if someone wants them, _ for layout)

    Quarter___Rate/Day__Andr(M)__iOS___WP____All_
    4Q2012__1,636,666__147.3____47.8___5.8___217.2
    3Q2012__1,346,666__121.2____26.9___3.3___171.4
    2Q2012__1,137,777__102.4____26.0___4.6___153.0
    1Q2012____897,777___80.8____35.1___2.3___145.2
    4Q2011____844,444___76.0____37.0___2.0___155.0
    3Q2011____626,666___56.4____17.1___1.3___117.7
    2Q2011____544,444___49.0____20.5___0.5___108.0
    1Q2011____393,555___35.4____18.2___2.0___101.2
    4Q2010____334,444___30.1____16.2___1.7____99.4
    3Q2010____222,222___20.0____14.1___2.4____80.0
    2Q2010____126,666___11.4_____8.4___1.6____61.7
    1Q2010_____60,555____5.4_____8.8___4.0____54.5
    -Q2009_____19,178____1.7_____8.3___2.6____43.8
    (last row is the average per quarter over 2009)

  15. I suspect the USA networks manipulate sales to get a 50/50 market share for iOS and Android. Price manipulations cross subsidies are very obvious in the USA market. That makes the ComScore numbers meaningless for global consumer preferences.

    LOL…you believe this? Really? For what reason would Verizon participate in price manipulations so Apple can maintain or grow share? Verizon only accepted the iPhone because AT&T was using it successfully to capture the most lucrative customers.

    The fact is that Apple makes great products that users like. And he’s right: ComScore numbers were not “meaningless” to you and esr when it positive for you util it stopped supporting the assertion that a disruptive market share collapse of iOS in the US was imminent after Android hit 50% share because of network effects and that Apple revenues were going to crater.

    Nope. Not so much.

  16. @Nigel:

    For what reason would Verizon participate in price manipulations so Apple can maintain or grow share? Verizon only accepted the iPhone because AT&T was using it successfully to capture the most lucrative customers.

    You got the first part of the answer right. The only part you missed was where the evil geniuses at Apple noticed that the carriers really, really wanted their offerings to ward off the competition, and managed to say “OK, but only if you sell lots of them.” Sprint managed to sell lots of them and make a huge loss, and Leap wireless didn’t actually manage to sell enough..

    Sprint has finally realized they need to sell more high-end Android. In doing so, they single-handedly put Android back on the top of the heap in January.

    So yeah, customers are looking for good stuff. For the right price. And it could be Android. But usually isn’t, in the states. T-Mobile is looking to really shake things up — it will be interesting to see what happens. I now have a Nexus 4 on GoSmart Mobile (a T-mobile-owned MVNO), with unlimited everything (but slow speed data) for $35/month.

  17. @Nigel
    If you are right Apple only makes great products for the USA. Worldwide, Android outsells iPhone 3:1. But outside the US, people have to pay more for iPhone than for competing Android phones.

    Why US networks would manipulate prices and market shares is something you USAians should be better able to understand. But I imagine it is part of the oligopoly politics that makes USAians pay what looks like double our EU rates.

  18. @patrick

    The company sold 6.6 million iPhones during 2012, with 2.2 million units being sold in the fourth quarter, of which 38 percent of subscribers were new to the carrier.

    That 38% are new to the carrier is why Sprint desperately wanted the iPhone: to stop hemorrhaging customers that left because they didn’t have the iPhone and to attract new customers who would not move to a carrier without an iPhone. The problem is their network sucks.

    As far as losses, revenue was higher and losses lower than analysts expected.

    Sprint reported a loss of $1.32 billion, or 44 cents per share…Revenue, though, increased to $9 billion. Analysts expected a per-share loss of 46 cents a share and revenue of $8.92 billion.

    The problem with ComScore is the same problem as Kantar. Both rely on surveys and not hard numbers. The only think you can safely say is that they probably are in the right ballpark. iOS and Android around the same and every one else hanging in the margins.

    The “around the same” is something that you guys insisted wasn’t going to happen once Android hit the magic 50% mark. Which happened what? 6 months ago?

    I’d probably go with one of the AT&T MNVOs simply because their network is better than T-Mobiles. Straight Talk or Black. T-Mobile should get you HSPA+ speeds…which isn’t that slow. At least they weren’t on AT&T. I was getting 4-6Mbps on my tower over HSPDA and we’re a moderately congested area.

    But with a family plan it works out cheaper to stay with AT&T for now rather than move to Verizon.

  19. @Nigel
    “USAians? LOL. You don’t cease to amuse.”

    I hear that often from people who run out of arguments. Thank you.

    As for the term, I would welcome a better one that does not presupposes that the 300M from the USA speak for all 1B living in the Americas.

  20. Winter, they don’t run out of arguments. They run out of patience.

    This is the problem many FSF zealots have: the inability to have a discussion without insisting folks adopt their wierdo language for “clarity” or because “words have meaning”. No shit. Nobody in the world misunderstands when someone writes American that they mean citizens of the United States and not someone from Columbia.

    When you write dumb words like USAians folks roll their eyes and laugh at you because you aren’t worth the effort of refuting since you just neutered your own credibility.

  21. @Nigel

    So when I write South American it is obvious I mean people from the south of the USA? Like what happened to me here. And when Americans live in the USA, what should I call people living on the continent? Non-American Americans?

    And you are sprouting nonsense about Apple market share and start arguments about my choice of words to avoid going into the subject of why the US market differs so much from the rest of the world.

  22. Entirely off-topic question:

    ESR, are you still maintaining the jargon file? I ask because I was searching for a version I could read on the kindle, and found this page:

    http://cosman246.com/jargon.html

    Which contains an apparently-newer version than the one on your site (which I think dates from 2003).

    Is this new version an officially-sanctioned (by you) continuation of the old file, or a bogus imposter?

    1. >Is this new version an officially-sanctioned (by you) continuation of the old file, or a bogus imposter?

      It’s not sanctioned by me. But Bryant’s answer is also pretty good :-)

  23. South Americans means folks that live in South America. Just like North Americans means US, Canada and Mexico.

    Nonsense about Apple market share? Where? Apple has around 50% share in the US contrary to what you and other folks here predicted. It is because they make high quality products that people like, not because of some weird conspiracy theory about market manipulation.

    US cell phone companies are in business to make money so if Apple didn’t provide a significant competitive advantage then they wouldn’t be paying more subsidy than for Android phones.

    The reason we went down this stupid tangent because you turned what could have been a useful comment into some oddball snipe against Americans and the implication that Apple is only successful because of “oligopoly politics”.

  24. Back to topic.

    The problem with supporting 2011-10-25T15:11:09Z!fred@foonly.com action stamp in gitweb (or other web interface like e.g. cgit or Gitalist), I guess to refer to commits, is that turning it into hyperlink is not cheap.

    Current solution for SHA1-like in gitweb is that gitweb turns it into actionless hyperlink, and checking if such object exists and what is its type (to select appropriate view / action) is done after clicking it.

    Hmmm… well, it could be done this way also for reposurgeon action stamp. Still, it is more expensive (search DAG from tips, unless there is some cache and we can assume monotonic time), than checking the type of object (which also checks if it exists).

    See e.g. http://etherpad.osuosl.org/ep/pad/export/git-clock-skew/latest?format=html

  25. Well, I thought we’d hit a tipping point around 50%, and then Apple would wind up on the wrong side of the network effects. And I was wrong, at least for now, in the US.

    So, then, the question is, why was I wrong?

    First, the US subsidy model eliminates the Android price advantage to the consumer. Certainly, the phone companies would like customers to get the cheaper phones, but the steering effort can only go so far before alienating a customer. If one of (AT&T, Sprint, Verizon) tried charging people more for the iPhone, the result would be bleeding customers to the other two instead. As US health care shows, price control efforts in a non-monopolistic thrid-party-payer system are not particularly effective. There is no effective price competition between Apple and Android among a large segment of people selecting phones, so the open platform’s price advantage is seriously reduced.

    Second, Apple managed to keep gaining US marketshare (Sprint) while Android was making its gains, and the Apple “ecosystem” had more cohesion. Android had version and hardware “fragmentation”, while the specific subsidized-handset model of the US resulted in a regular upgrade cycle for Apple. The result is that to an app vendor, the 35%-share Apple network is as valuable as the 50%-share Android network. Thrown in the usual frictions involved in internationalization, and there is a solid US Apple market segment that is not any less attractive to devs than the US Android one, and thus no development effort switchover.

    As a result, while the overall world market is going to keep moving to an Android monoculture, Apple persistence in the US market is going to be unusually strong. Much like some market segments (like graphics) remained dominated by Macs for years after Windows gained 80+% market share in the overall market, the US is going to have a local outsized iPhone share going forward.

  26. @Steven Ehrbar

    iPhone, by “US Americans” for “US Americans”

    @Nigel
    So we have North America, Central America, South America, and America?

  27. >So we have North America, Central America, South America, and America
    The common name for ” Republic of Ireland” is “Ireland”
    The common name for “Bundesrepulblik Deutschland” (“Federal Republic of Germany”) is “Germany”
    The common name for “Estados Unidos Mexicanos” is “Mexico”.
    The common name for “Estados Unidos de Colombia” was “Columbia”.
    The common name for “Estados Unidos de Venezuela” was “Venezuela”.
    The common name for “The United States of America” is “America”

    A pattern seems to emerge here. A country with a formal name in the format “${GovernmentType} of ${GeographicArea}” is often commonly known as “${GeographicArea}”.

    When one refers to the continent of North America or South America, there is no ambiguity. One may speak of inhabitants of South America as “South Americans”, and inhabitants of the southern US as “southern Americans”.
    I suppose “Panamerican” is as good a word as any to refer to the entire population of all of the Americas (including the islands not considered integral to either continent).

    What do you call residents of the “European Union”? The logical name would be “European”. But what is common usage?

  28. > First, the US subsidy model eliminates the Android price advantage to the consumer

    So what you’re saying is that, at price-parity, a majority of consumers will pick the iPhone over any Android device.

    That’s an interesting result, because as soon as Apple wants to price into the curve, there is no price advantage for Android. Should Apple wish to pursue a low-margin business, it’s well-able to do so.

  29. @LeRoy:

    So what you’re saying is that, at price-parity, a majority of consumers will pick the iPhone over any Android device.

    Hard to say. When I looked around a year ago, a high end Android phone that cost the carrier less would actually cost the consumer more on most carriers. This quarter, Sprint started pricing higher end Androids to sell, and they are selling.

    Historically, carriers definitely subsidized iPhones more heavily, partly because they got fewer returns. Of course, this trend started before there were any comparable Android phones, and there is no doubt that a lot of customers do prefer the iPhone. There is also the Apple store penetration in the US — when a carrier sells an iPhone, in general it doesn’t need to worry about it any more. There is no question that people who definitely prefer iPhones aren’t going to be happy with anything else. The same is true of people who definitely prefer Android, but they aren’t going to settle for an iPhone and then take it back. The ones in the middle are getting steered towards iPhones. It will be interesting to see if Sprint’s strategy spreads to the other carriers, or if Apple

    That’s an interesting result, because as soon as Apple wants to price into the curve, there is no price advantage for Android. Should Apple wish to pursue a low-margin business, it’s well-able to do so.

    Their margins are already eroding, but I doubt they have the stomach to drop them very quickly. And it turns out that iPhones cost too much for a lot of potential buyers at present.

  30. The problem with supporting 2011-10-25T15:11:09Z!fred@foonly.com action stamp in gitweb (or other web interface like e.g. cgit or Gitalist), I guess to refer to commits, is that turning it into hyperlink is not cheap.

    Does gitweb have a sort of cache that it can look up? Perhaps it’s too expensive to go through all the commits in every repository immediately, but perhaps on viewing a commit with an action stamp, gitweb (or similar) would do a quick look up of whether the action stamp is already in the cache (possibly even marked as not being a valid stamp for the current repository) and then scan through the repository for the date and committer, before generating the final page and deciding whether it’ll be a hyperlink.

    Of course I may be misunderstanding the problem.

    (also.. how did this thread get derailed so badly with the smartphone stuff?)

  31. @The Monster
    “What do you call residents of the “European Union”? The logical name would be “European”. But what is common usage?”

    Europeans. But we call Russians and Swiss Europeans too, although they do not live in the EU. Just as East-Germans living in the former DDR were still Germans, even though they were not living in the Bundesrepublik Deutschland. But Americans are exclusively those living in the US, I hear.

    So this is settled, Asians are from Asia, Africans from Africa, and Americans from the USA. And there is no continent that unifies Cubans, Mexicans, Canadians, and Argentinians.

  32. @Nigel
    “Nonsense about Apple market share? Where? Apple has around 50% share in the US contrary to what you and other folks here predicted. It is because they make high quality products that people like, not because of some weird conspiracy theory about market manipulation. ”

    Indeed, and world-wide we saw the effect that Android is taking over the Smartphone market, currently outselling iPhone 3:1.

    But it is basic economics 101 how subsidies distort markets. It is clear that if Volkswagen cars are more expensive than cars from Mercedes Benz, Mercedes will sell more cars than Volkswagen. If a Galaxy Note is more expensive than an iPhone 5, like I hear it is in the US, it is no surprise that they sell less of them. But elsewhere in the world, a Galaxy Note is cheaper, eg, 509 vs 627 Euro over here.
    http://www.beslist.nl/products/elektronica/elek_mobiele_telefoons/?m_as_nm=1

  33. Good grief, Winter.

    “And there is no continent that unifies Cubans, Mexicans, Canadians, and Argentinians.”

    Exactly how often do you need to refer to that group? Compared to residents of the USA?

    Your politics are making you as stubborn as a Stallmanite.

  34. And there is no continent that unifies Cubans, Mexicans, Canadians, and Argentinians

    Maybe because they all don’t live on the same continent…

    And I thought only Americans were geography challenged.

  35. @The Monster
    “Exactly how often do you need to refer to that group? Compared to residents of the USA?”

    There are more people from Surinam and the Caribbean in the Netherlands than from the USA by orders of magnitude. In April, we have an Argentinean queen sitting next to our head of state. A Mexican is buying our biggest network provider.

    But maybe Nigel is right, and there are two continents combined called the Americas?

    But that seems to be a language thing, according to Wikipedia.

    Eg, in Dutch it is common to call them simply “Amerika” (singular) but this status is debated and the Germans even have a special word for it “Amerika ist ein Doppelkontinent”.
    https://de.wikipedia.org/wiki/Amerika
    https://nl.wikipedia.org/wiki/Amerika_%28continent%29

    So, in the end it might be just my age. I must update my primary school geography.

  36. Sorry, that should have been

    @Jay Maynard
    “Exactly how often do you need to refer to that group? Compared to residents of the USA?”

  37. Winter, North America and South America sit on two different continental plates which is an exceedingly strong objective reason to state they are two different continents. The only two continents that live on the same plate are Europe and Asia.

    An ironically Eurocentric view of continents given the heartburn you have with the term “Americans” as arguably the proper term for folks that live in Europe at the continental level is Eurasian and not European.

    If you’re going to be pedantic about contiguous landmass as the “proper” definition (in order to group the Americas as one continent) then you folks are called Afro-Eurasians. When it becomes common to refer to Europeans as Afro-Euroasians then perhaps Americans might be considered confusing as to whom you are referring to when speaking English.

    Until that day, not so much and by then we’re going to be speaking Mandarin anyway and there’s no such ambiguity in Chinese…

  38. @Winter
    Just as East-Germans living in the former DDR were still Germans, even though they were not living in the Bundesrepublik Deutschland.

    When the DDR existed, there were “East Germans” and “West Germans”, as well as “Germans”. When the Republic of Vietnam existed, it was known as “South Vietnam” and the “Democratic Republic of Vietnam” was known as “North Vietnam”. Now there is only one Vietnam and one Germany.

    There is only one country with “America” in its formal name. It is also the first country in the Americas to gain independence from a European colonial power. It therefore got the “first mover” advantage in mind share, which might explain why there are no other countries named “*America*”. Had the CSA won the “Civil War”, there would have been two “___ States of America”, and the name “American” wouldn’t unambiguously refer to either one. Perhaps “American” would still have referred to the USA, and the newer nation’s inhabitants would have been “Dixians”.

    There isn’t a convenient nomenclature to distinguish between residents the island of Ireland as a whole, vs. the Republic. There are sports teams called “All Ireland”, which includes UK subjects as well as Republican citizens, but the usage doesn’t seem to carry over to other contexts. The difference is that we don’t call the Republic “South Ireland”. We don’t have a name for residents of {Virginia, West Virginia} either, while “Dakotan” can easily be understood to mean “Resident of {North Dakota, South Dakota}”.

    Does anyone ever need a name for residents of {Africa, Asia, Europe}? What would that name be? Now throw Greenland into the pot. It is geographically part of the Americas, but socially closer to Europe. Is a resident of Greenland an Afeurasian or a Panamerican?

  39. There are more people from Surinam and the Caribbean in the Netherlands than from the USA by orders of magnitude. In April, we have an Argentinean queen sitting next to our head of state. A Mexican is buying our biggest network provider.

    Gee, you think that might be because you guys owned Suriname for 300 years? Until 1975?

    It also strikes me that the majority of the time folks wish to group Surinamese, Argentinians and Mexicans into the same group as Americans is when they wish to rail against annoying foreigners…

    Outside of Europe there’s no great desire to subsume national identity into a continental one. So when someone says “Those Americans are so dumb because they believe…” Argentinians, Mexicans and Surinamese laugh.

  40. “Is this new version an officially-sanctioned (by you) continuation of the old file, or a bogus imposter?”

    If you meet the Buddha on the road, kill him.

  41. @Nigel
    “Outside of Europe there’s no great desire to subsume national identity into a continental one.”

    As do Asians, Australians, and Africans on a regular basis.

    New Asian Hemisphere: The Irresistible Shift of Global Power to the East
    Kishore Mahbubani
    http://www.mahbubani.net/book3.html

    For centuries, the Asians (Chinese, Indians, Muslims, and others) have been bystanders in world history. Now they are ready to become co-drivers.

    Asians have finally understood, absorbed, and implemented Western best practices in many areas: from free-market economics to modern science and technology, from meritocracy to rule of law. They have also become innovative in their own way, creating new patterns of cooperation not seen in the West.

  42. The Monster
    “Does anyone ever need a name for residents of {Africa, Asia, Europe}?”

    Actually, people do use Eurasian quite regularly, especially in linguistics, history, and archeology. There is little use in adding Africa, except when talking about the Mediterranean.

  43. For centuries, the Asians (Chinese, Indians, Muslims, and others) have been bystanders in world history. Now they are ready to become co-drivers.

    The only folks that really believe this are Europeans (and I count Americans in this context probably more so). As well as the desire to lump Chinese, Indians and Muslims into one group. That some south asian authors cater to this worldview isn’t unexpected given that it’s probably a lot more lucrative than teaching actual history.

    Anyone who believes that Muslims were bystanders in world history doesn’t know world history or even European history. Ever hear of the Hasburgs? Holy Roman Emperor and Defender of Christianity against Islam? That wasn’t just an honorary title. Suleiman was a significant threat to Europe although his descendants not so much. Thus the Ottoman Empire was a significant influence on European politics and power balance even in times of peace and decline all the way through World War I. Whatever intrigues the Hasburgs indulged in Europe they had to keep a keen eye on the Ottomans…

    And this ignores the earlier Umayyad conquest of most of Spain and the following Reconquista as well as the interminable Crusades and the fall of Byzantine Empire.

    Seriously WTF?

  44. “But Americans are exclusively those living in the US, I hear.”

    That usage is common in US English, but residents of the rest of the western hemisphere don’t go along. When I worked for that Spanish language radio station, a common complaint I heard from the Spanish speakers was, “Why do people in the US think that they are the only Americans? I’m from Chile and I’m an American, too. We all are.”

  45. @Nigel
    “Seriously WTF?”

    The believe in distortions of history and geopolitics is not limited to (South-) Asians.

    Further, you should not complain. You asserted that “Outside of Europe there’s no great desire to subsume national identity into a continental one.”. I just supplied a prominent example from Asia that does the same. I never said I believed his account. But he is read not only outside Asia.

  46. @Winter:
    >So when I write South American it is obvious I mean people from the south of the USA?

    Nope. Those are “Southerners” (in American English at least, I can’t speak for the rest of the Anglosphere). Even here things aren’t entirely obvious as “the (American) South” refers only to the Southeast of the country (Texas and points east), so someone from Arizona or New Mexico isn’t a “Southerner”.

    “South Americans” are people who live in South America.

    >Like what happened to me here. And when Americans live in the USA, what should I call people living on the continent? Non-American Americans?

    When referring to people by where they reside in terms of physical geography, one speaks of “North Americans” and “South Americans”, according to which continent they live on. When referring to people by where they reside in terms of political/cultural geography, one speaks of “North Americans” (Americans and Canadians), and “Latin Americans” (those countries where the national language is Spanish or Portuguese, or, alternatively, any point in the Americas south of the US). On the rare occasion that one needs to refer to everybody living on both continents, a phrase along the lines of “residents of the Americas” would probably be used.

    @Jay Maynard:
    >Your politics are making you as stubborn as a Stallmanite.

    I’m not sure it’s politics (or, at least, primarily politics). European languages generally have terms to distinguish people from the US from residents of the Americas in general (when my nationality was mentioned by others during the year I spent in Germany, I was generally called a “US-Amerikaner”), so the chances are good that Winter is mostly just trying to find a concise exact match for a pair of terms he’s used to in Dutch.

    Obligatory link in any discussion of the word “American”: http://www.youtube.com/watch?v=29g57XTYgLE

    @Winter and Nigel, regarding the argument over “bystanders in world history”:

    Up until a few centuries ago, *everybody* was a bystander in world history. Europeans had no awareness of or influence over events in China and vice versa. Europeans were lucky enough to be dominant at the start of the period in which it has been possible not to be a bystander (which is to say, the period in which at least one part of the world has been in fairly close contact with every other part), so the sentence at the beginning of Winter’s article is trivially true.

    The bulk of the article seems to have a fairly shortsighted view of history though: a lot of the stuff mentioned as “western” has only been common in the west relatively recently.

  47. >> The problem with supporting 2011-10-25T15:11:09Z!fred@foonly.com action stamp in gitweb (or other web interface like e.g. cgit or Gitalist), I guess to refer to commits, is that turning it into hyperlink is not cheap.

    > Does gitweb have a sort of cache that it can look up? Perhaps it’s too expensive to go through all the commits in every repository immediately, but perhaps on viewing a commit with an action stamp, gitweb (or similar) would do a quick look up of whether the action stamp is already in the cache (possibly even marked as not being a valid stamp for the current repository) and then scan through the repository for the date and committer, before generating the final page and deciding whether it’ll be a hyperlink.

    Of course I may be misunderstanding the problem.

    First, there is question if to validate and detect type of object (for SHA-1 like), or translate into SHA-1 identifier (for action stamp) when displaying i.e. generating link, or only when trying to access the page i.e. after clicking the link. I think for performance reasons it should be the latter, especially for pages that contain commit messages of many commits like e.g. ‘log’ view.

    In this case there is also a second question of detecting either SHA-1 identifier, or action stamp. I think that here it is probably easier to find something that looks like shortened SHA-1 but isn’t that something that looks like action stamp but isn’t.

    Third, there is a case of finally validating of (shortened) SHA-1 and if valid finding type of object pointed to by SHA-1, or finding a commit that corresponds to action stamp. In this case validating SHA-1 is easy: try to find it’s type and catch errors. Turning action stamp to SHA-1 is hard, and requires walking the history of project. Cache might help here, but [canonical] gitweb currently doesn;t support any such cache.

    Note that if we can trust timestamps to be monotonical to some extent at least (no time skew in repository, or time skew limited in time and number of commits), we will be able to use this cache to speed up access also for action stamps that are outside cache: find closest time and walk from it.

    I hope this makes it more clear.

  48. In the English language “American” tends to mean someone from what is now the United States. I notice that Englishmen are far more likely to use “America” to mean the United States specifically than even (US) Americans are. This usage probably dates back to before the USA was a country in its own right; and a Londoner saying “I’m going to America” would almost unambiguously mean “I’m going to the British colonies in America” as opposed to the Spanish lands in central or South America.

    That’s just my guess and it has flaws; for example it doesn’t really explain why Canada doesn’t count as America to modern Brits, but it’s a start.

  49. @Jon I understand what you’re saying but it’s not even trivially true unless you wish to also ignore the entire Pacific War in WWII as well as the already cited WWI involvement of the Ottoman Empire.

    Whatever you wish to say of the Japanese, “bystander” isn’t one of them during the period of 1904 through 1945 (or after for that matter).

    While Europeans had little influence on China, the Mongols had influence on both. Had it not been for the death of Ogedei the Mongols could have conquered Europe as well. They beat every feudal army sent against it. Last I heard Mongolia was in Central Asia.

  50. Further, you should not complain. You asserted that “Outside of Europe there’s no great desire to subsume national identity into a continental one.”. I just supplied a prominent example from Asia that does the same. I never said I believed his account. But he is read not only outside Asia.

    I was trying to remember who this was and finally googled him and remembered. He’s the guy who wrote “Can Asians Think” back in the 90s which in my vague recollection was yet another set of essays on how the US was doomed and the Asian tigers and the Japanese were going to kick our butts.

    Yah, not so much. And yeah, I stand by my assertion. Certainly the Chinese don’t think that way.

  51. @esr

    >It’s not sanctioned by me. But Bryant’s answer is also pretty good :-)

    Indeed.

    In case anybody’s interested I found a decent set of ebookified versions here:
    http://xy22y.tumblr.com/post/16225563626/jargon-file-4-4-7-typeset-pdf-epub-mobi

    @LS

    >That usage is common in US English, but residents of the rest of the western hemisphere don’t go along.

    It’s extremely common in British English as well. References to “Americans” or “an American” here are unambigiously understood as denoting people from the US.

  52. Action stamp question:

    I make some change simultaneously on multiple branches of a repository. How are these disambiguated? Recall that Git allows you to set the authoring and committing times for a commit arbitrarily—and if it didn’t, the fast-export format would.

    So which commit is timestamp:1 and which is :2?

    1. >So which commit is timestamp:1 and which is :2?

      They are numbered by arrival order at the repository. We don’t know how skew clocks elsewhere might be, but we know the local clock is monotonic.

  53. @Nigel
    “Had it not been for the death of Ogedei the Mongols could have conquered Europe as well.”

    Nope. They already were in Poland but never bothered to come back.

    There was nothing worth fighting for in Europe. Also too many forests and mountains and too little room to maneuver with a few hundred thousand soldiers on horse back. The Mongols did stay in Russia though.

  54. @Nigel
    “The reason we went down this stupid tangent because you turned what could have been a useful comment into some oddball snipe against Americans and the implication that Apple is only successful because of “oligopoly politics”.”

    Indeed, the whole point of this wild goose chase was that I was specific about where Apple is so successful, in the USA (and I really do not care whether you call me Eurasian or Eurafrican). We see that Apple gets their 50% market share only where there is an oligopoly that manipulates prices with cross-subsidies that lowers iPhone prices wrt Android prices.

    I am not proposing some conspiracy, just standard oligopoly tactics, straight out of the Economist’s text books: In any supply chain, the profit will end up at the most consolidated level.

    In the USA, the two most consolidated levels are the network providers (Verizon and AT&T control 64% of the cellular market) and Mobile OS providers (Apple and Google control ~90% of Smartphones)/Handset producers (Apple and Samsung control ~60% of the market)

    If either Apple, Google, or Samsung get over 60% of the market, the networks will lose control over the “bottle neck” in the chain and therefore, lose profits. Simple rule of business to never depend on a single supplier or customer.

    So it is perfectly rational and predictable that Verizon and AT&T both have a big interest in preventing Apple, Google, and Samsung from getting market dominance. Which shows in the way they manipulate prices for these products.

    The simple fact that the networks subsidize iPhone more than Android phones already tells us that Apple would lose massive market share if iPhones were sold for their real prices.

  55. @Winter
    >There is little use in adding Africa, except when talking about the Mediterranean.

    Africa is more “connected” to Eurasia than North and South America are, but you see “little use” in concocting a word to describe that combined contiguous landmass.

    It is a pure accident of history that most Europeans learned of the New World from Amerigo Vespucci’s maps that introduced both continents at the same time. Were North America simply known as “America”, and South America known as “Vesputia”, the use of “American” for the USA would almost precisely correspond to “European” for the EU, and you’d climb down from your high horse.

  56. @The Monster
    “Africa is more “connected” to Eurasia than North and South America are, but you see “little use” in concocting a word to describe that combined contiguous landmass.”

    The Sahara used to separate the Mediterranean from the rest of Africa quite successfully.

    @The Monster
    “Were North America simply known as “America”, and South America known as “Vesputia”, the use of “American” for the USA would almost precisely correspond to “European” for the EU, and you’d climb down from your high horse.”

    I have no idea what you are after here. I really have no high horse to climb from.

    English speaking people seem to equate “America” with “the USA”. In other languages this can be confusing as “America” is used both to refer to the two continents combined and as an abbreviation of USA. This seems to include most non-USA inhabitants of the land masses between Alaska and Tierra del Fuego.

    The only thing I needed was a name for those living in the USA that did not confuse them with others living on the same continent(s). If you cannot stand the idea that there are other people living outside of the USA, I cannot help you.

    And the EU is indeed often abbreviated to “Europe” when there is no confusion, but that is non-exclusive. No one will deny that the Swiss, Icelanders, or Belarus are European. It is just a geographical marker meaning West of the Ural mountains and north of the Mediterranean. The fact that there are both European and Asian Russians and Turks bothers no one.

  57. So which commit is timestamp:1 and which is :2?

    They are numbered by arrival order at the repository. We don’t know how skew clocks elsewhere might be, but we know the local clock is monotonic.

    Take a look at https://github.com/jcsalomon/patient.git for a deliberate counter-example. There are commits on branches A, B, C, and D, all with the same commit time.

    1. >Take a look at https://github.com/jcsalomon/patient.git for a deliberate counter-example. There are commits on branches A, B, C, and D, all with the same commit time.

      I understand that this is possible. I don’t see why it matters. We must be talking past each other somehow.

      The commits all have the same commit time because git only records those to 1s granularity, but they didn’t arrive at the same time. When reposurgeon makes an action stamp, it looks for commits with a matching commit stamp, counts how many there are, and uses that count as a suffix if it’s nonzero.

  58. Actually, they did arrive at the same time: I created the branches in the git-fast-export stream with the same time-stamp and at the same time. Extend Git to nanosecond precision and they’ll have the same time-stamp.

    Now, reposurgeon working on this repository might well assign :1 to the commit in branch A, …, and :4 to the commit in branch D. But how is a future program (e.g., gitweb) to know which commit a time-stamp refers to?

    (Is the “mark :6” designation stable across repeated imports, exports, etc., etc.?)

    1. >But how is a future program (e.g., gitweb) to know which commit a time-stamp refers to?

      Oh, I think I see what you mean now. gitweb, looking at the deserialized repo, doesn’t actually have access to the implied sequence of the stream of commits that built it. I was led astray by the fact that reposurgeon always starts by parsing a stream file of commits, so there’s an implied monotonic arrival order even if there are timestamp collisions.

      It’s a kluge, but the obvious tiebreaker to use is sort order of the branchnames.

  59. “Some people, when faced with a software problem, think “Aha! I know, I’ll add an emacs mode!”. Now they have two problems.”

    I don’t think that’s literally true (in that I’m not sure it actually multiplied problems), but I found it so amusing I couldn’t resist posting.

  60. >>But how is a future program (e.g., gitweb) to know which commit a time-stamp refers to?

    >Oh, I think I see what you mean now. gitweb, looking at the deserialized repo, doesn’t actually have access to the implied sequence of the stream of commits that built it. I was led astray by the fact that reposurgeon always starts by parsing a stream file of commits, so there’s an implied monotonic arrival order even if there are timestamp collisions.

    It’s a kluge, but the obvious tiebreaker to use is sort order of the branchnames.

    The problem with that kluge is that branchnames can be different in different clones of the same repository (e.g. in Git, though nowadays with remote-tracking branches they usually are the same, just perhaps differently namespaced), or here can be no branchnames (older Mercurial with anonymous heads and no bookmarks extension) or branchnames can be the same (older Mercurial with so called “named branches” which in truth are just commit labels – nowadays the situation with bookmark branches is similar to the one in Git).

    If we want to avoid this, and not use ordering based on native commit identifiers (SHA-1 in case of Git), then perhaps sort order of commit messages, and if it is not enough, sort order of action stamps of parents of said commits.

    Though I think that it is fairly theoretical exercise; I don’t think that in real repositories there are commits with identical author and authordate.

    BTW. action stamp is about *author* date, or about *committer* date?

    1. >BTW. action stamp is about *author* date, or about *committer* date?

      Committer. But that would be an easy switch to flip if you think author has better properties.

  61. On slightly unrelated note: Kiln now supports transparently and interchangebly both Git and (original) Mercurial: http://www.joelonsoftware.com/items/2013/03/11.html

    I guess that they use the same mechanism that reposurgeon and git’s remote helpers (like git-hg), namely the fast-export stream… which was originally invented from what I remember for fast conversion of Mozilla CVS repository to Git, which didn’t happen.

  62. Author might be more valuable, committer is basically just whenever it got applied to the repository, but author name and date will be preserved across, eg, patches sent by email and applied by someone else.

  63. @winter: “English speaking people seem to equate ‘America’ with ‘the USA’. In other languages this can be confusing as ‘America’ is used both to refer to the two continents combined and as an abbreviation of USA.”

    So? The English word “day” means both the sunlit portion of a 24-hour period, and the full 24-hour period itself. There is plenty of ambiguity in English. There’s nothing unique about this case that’s gotten so many people here so hot and bothered.

    I can’t believe I’m replying to this incredibly off-topic topic…

  64. I pushed to https://github.com/jcsalomon/patient.git a new, definitely contrived, case. No branch names have been pushed to GitHub (because they’ve all been merged together) and three commits with distinct content have identical commit messages and time-stamps.

    Can you predict, without running reposurgeon, how the simultaneous commits “grepsling”, “kugelator”, and “whizding” will be ordered?

    Yes, it’s an artificial case. But I can imagine an automated tool generating such commits. (Isn’t this what happens when a Subversion cross-branch commit is split up? Two commits, on different branches, with the same commit-time and -message.)

    Even in “normal” cases of simultaneous commits, where you suggested ordering by the sort-order of the branch-names or the commit messages, what happens when you merge someone else’s tree? Is it inconceivable that fred@foonly.com submitted the same patch to several tines of a fork, which later merge together. When forks A and B merge, we can refer to :1 and :2 (and pre-merge references to the action stamp can be resolved to the patch in the ancestry of the referring commit). When forks C and D merge, they too will call their patches :1 and :2. When AB and CD merge, there will be glorious inconsistency.

    I’m beginning to think that the action-stamp idea is a bit of a kluge. It will work in most real-life cases, but every now and then it will fail.

    I don’t have a solution for the cross-DVCS problem, but so long as I’m using Git I’ll stick with the conventional truncated hashes. Action stamps are not powerful enough.

    1. >I’m beginning to think that the action-stamp idea is a bit of a kluge. It will work in most real-life cases, but every now and then it will fail.

      Undoubtedly so. Given the constraints on the problem, though, some variant of it is probably the best that can be done. We don’t have SHA1 hashes of content available when the output stream is being generated. Supposing we did, those hashes would be unresolvable by an importer trying to render that stream for a non-git system. Thus, we can’t “stick with the conventional truncated hashes”.

      On the other hand – as Mike Swanson points out – the failure case will be extremely rare. I think the thing to do is bite the bullet, accept that action stamps will sometimes have multiple matches, throw out the suffixing, and recommend that clients like github resolve the problem by throwing a popup that displays links to all matches when there is a multiple.

      Also, I intend to change reposurgeon to use author timestamps rather than committer when they are available; I think that suggestion is good.

  65. @Cathy
    That is why I am so surprised people get angry when I try to be specific. I seem to have insulted them.
    (we do have a word to specify a 24 hour period “etmaal”)

  66. @Joel

    Your case is extremely contrived. Can you show a case where a real repository will ever encounter this issue? I think anyone can imagine the scenario, but to call action stamps bust because of an extremely unlikely eventuality is going too far.

  67. @Mike: @Joel gave quite realistic example of splitting multi-branch commit from Subversion when converting to Git – you will have same author, same author timestamp, same commit message, same comitter, same committer timestamp.

    @esr: Using author timestamps mean that same action stamp might happen more often, as now you have also the case of cherry-picked commits (from one branch to other branch). I think…

    BTW. do other version control systems store committer time, or author time? AFAIK Git is unique in separating authorship (whi created a change, which might come from email) from committership (who entered change into repository, which might be from somebody else email, or from some other commit cherry-picked).

    @Joel, @esr: You can always use parents (e.g. lexicographical order of parents action stamps, recursively if necessary) to resolve action stamp conflicts. If commits have the same parents, they are the same commit.

    Though it could fail in some case where action stamps of commits are the same all the way down to root (e.g. parallel histories, e.g. with and without secret file), it should be much, much more rare.

    1. >You can always use parents (e.g. lexicographical order of parents action stamps, recursively if necessary) to resolve action stamp conflicts. If commits have the same parents, they are the same commit.

      Brilliant idea! I’ll think about this and write up something.

      >Using author timestamps mean that same action stamp might happen more often, as now you have also the case of cherry-picked commits (from one branch to other branch). I think…

      I think this is less common than the patch-shipping case that using author avoids – but it’s nearly a moot point if we’re going to order by parent marks.

  68. @Winter:

    I think it was the pseudo-word USAian that got attention. Frankly, that “word” (c’mon, you can’t combine an upper-case acronym with a lower-case suffix and not expect people to find it amusing at best or idiotic at worst) itself sounds like an attempt to distract from the issues, so when your response to Nigel’s LOL about the word was to claim that others were attempting to distract from the issues, that’s when the excrement collided with the rotating ventilation device. I could be wrong, but I think the discussion would not have gone anywhere near that direction if you had said “US residents.”

  69. (And yes, I know, in some quarters, that “word” is considered to be hip, but those quarters have little, if anything to do with most of the resident curmudgeons around here.)

  70. @Patrick
    “that “word”… itself sounds like an attempt to distract from the issues”

    Ok, then I chose the wrong word.

    Mixing upper and lower case in pronouncing acronyms is customary in my native language and I have encountered the equivalent of “US Americans”, so it did not look odd to me. However, the distraction was indeed complete.

  71. @Winter

    >Ok, then I chose the wrong word.

    There’s no need to come up with *any* new word (and, by the way Patrick, “US resident” does not fit the bill because residency is an orthogonal concept to nationality) because we already have a perfectly good one. “American” is clear and unambiguous in English, which is the language we’re using here. It is irrelevant that this word does not have the same meaning in other languages.

  72. @Tom
    What incites this religious fevor to claim to be the one and only true Americans? It is just a word.

  73. @Tom
    ” Patrick, “US resident” does not fit the bill because residency is an orthogonal concept to nationality”

    But Patrick was right. I was writing about those using USA cellular networks. Which equates to residency, not nationality.

  74. @Winter

    There’s no religious fervour, and I’m not an American. I’m just pointing out that in the English language the words ‘America’ and ‘American’ (when used unqualified) almost always relate to the United States. It’s just a fact about the language that you can either accept or obstinately ignore.

    By the way, every Canadian I have ever met has been very quick to point out that they are *not* American. ;)

  75. @Winter

    >But Patrick was right. I was writing about those using USA cellular networks. Which equates to residency, not nationality.

    You’re right.

  76. Indeed, and world-wide we saw the effect that Android is taking over the Smartphone market, currently outselling iPhone 3:1.

    But it is basic economics 101 how subsidies distort markets.

    Interesting how it is obviously now basic economics 101 today but given that the US market has been subsidized since before the iPhone was released it wasn’t obviously basic economics 101 the last couple years when the prediction that Apple market share was obviously going to crater because of “disruption” due to “network effects”.

    Perhaps that’s an indication that you guys should be less…ah…adamant…regarding the validity of your analysis…

    Yep, Android is outselling Apple 3:1 outside the US and EU. That would leave Apple with around a 20-25% share which is about what most folks have been predicting and not the sub-5% you and ESR were predicting.

    Interesting that even in China the market share is around 20-25% despite the lack of a inexpensive model although helped by subsidies but hurt by graymarket and very cheap android phones. They also gained in India to around 15% which was really surprising given that their presence was an afterthought six months ago and there are no subsidies.

    The only clear picture has been wherever Samsung and Apple makes even a minor effort Blackberry and Nokia share starts to crater. Incumbents disrupted by Apple from above and Samsung from below, middle and high.

    I do wonder how well Android would be doing in the US without Samsung given the lackluster performance of the other Android manufacturers but Apple is largely supply constrained for the first couple months anyway.

  77. Ok, then I chose the wrong word.

    Not if the intent was levity…

    However, the distraction was indeed complete.

    Hence the failure of many FSF issue discussions with normal people…

  78. @Nigel
    “Apple is largely supply constrained for the first couple months anyway.”

    I have heard that excuse since fall 2010?

    You simply state here that differential network subsidies do not affect market shares. Which is as ridiculous as it sounds. As Moore’s law trudges along, more of the market will be at the low and lower end. Apple has shown in over 30 years of history to never go there. Consequently, a continuously larger part of the market will be far below the iPhone price point. Subsidies will not stop that.

    @Nigel
    “That would leave Apple with around a 20-25% share which is about what most folks have been predicting and not the sub-5% you and ESR were predicting. ”

    First, the marked has not stabilized yet. Android is on it way to reach their first billion this summer, next billion 2014. Apple is never going to reach billions of people without a fundamental change in their pricing and product strategies. Things are still moving.

    Now for some serious predictions. In the current market (ie, if FireOS does not take off massively on the low end), Android will power ~5 billion phones in ~5 years. Optimistically iPhone will stick to ~0.5 billion. Hence, Apple will end up with at most ~10% of market share. Probably much less. The other OS’ will be noise.

  79. You simply state here that differential network subsidies do not affect market shares. Which is as ridiculous as it sounds.

    I’m simply stating that you guys were wrong before for reasons you now state are obvious Economics 101 and there’s no indications that you’re any more right today that Apple market share is going to suddenly crater.

    You can cite “differential subsidies” or “imminent disruption from below” or “network effects” all you like but predictions made two years ago didn’t happen beyond what most folks already accepted: android would do really well. That was never really disputed. It was the assertion that iOS would do really badly that was disputed.

    First, the marked has not stabilized yet. Android is on it way to reach their first billion this summer, next billion 2014. Apple is never going to reach billions of people without a fundamental change in their pricing and product strategies. Things are still moving.

    Indeed, things are still moving, things aren’t stabilized today so ESR’s predictions of Apple’s imminent doom were very premature a couple years ago. Still so today.

    So is your growth curve.

    Most folks (on Android sites) are still predicting fall (september). Maybe a little further out for 2B. Looking at the Android sites many seem to think that Android will be at 2B+ when iOS is at 1B in 2015.

    Now for some serious predictions. In the current market (ie, if FireOS does not take off massively on the low end), Android will power ~5 billion phones in ~5 years. Optimistically iPhone will stick to ~0.5 billion. Hence, Apple will end up with at most ~10% of market share. Probably much less. The other OS’ will be noise.

    My serious prediction is that any 5 year prediction is probably wrong. Certainly any 5 year market prediction made in 2007 were clearly wrong in 2012. In 5 years it probably won’t be like the “current market” caveat you applied to your “prediction”.

    Given that Apple has already sold 500M iOS devices your guess is…well…probably low. They sold 125M iPhones in FY12. There’s no indication that they wont hit that billion mark in 2015 as “optimistically” predicted on android sites (LOL) and if that’s all they have in March 2018 then 5 B vs 1B (out of 6B) is still 16%. If they can get to 2B (out of 7B) then that’s 28%.

    But it’s so far out that the numbers are all conjecture and hardly worthy of a “serious prediction”.

  80. I have heard that excuse since fall 2010?

    Is that as long as we’ve been hearing that Apple’s market share is about to collapse? LOL.

  81. I’m simply stating that you guys were wrong before for reasons you now state are obvious Economics 101 and there’s no indications that you’re any more right today that Apple market share is going to suddenly crater.

    Except that it has, just not in the US.

    So you’re about as wrong as the people you’re refuting, and also about as right. As to why Apple is doing so well in the US and comparatively poorly elsewhere (from what I’ve seen, that would be everywhere but the UK), I don’t think anyone has offered a completely adequate explanation. The strong subsidy model in the US would seem to be a natural and unavoidable *part* of that explanation.

  82. @Nigel
    “I’m simply stating that you guys were wrong before for reasons you now state are obvious Economics 101 and there’s no indications that you’re any more right today that Apple market share is going to suddenly crater. ”

    Eric did not expect that the networks would subsidize iPhones to keep their market share artificially high. So, his error seems to be that he was wrong thinking the USA was a competitive market for mobile phones. In hindsight, this might look indeed a little naive. I am sure you saw this coming all the time but had your reasons to remain silent.

    @Nigel
    ” In 5 years it probably won’t be like the “current market” caveat you applied to your “prediction”.”

    Indeed. I expect my prediction to be wrong. But what will be wrong?

    There will not be 5 billion smart phones? Most phones will run iOS, RIM, or WP8/9/10 io Android? They will run FireOS or some other Linux based OS? What do you think? Or there will be a new OS not build on Linux? Would I care if FireOS would disrupt Android from below? Would Eric care?

    Do you really think someone could come up with a brand new OS for phones in the next five years that will drive out the current crop?

    And if Eric was wrong in the past about Apple, how did you fare? I remember a lot of people predicting Apple taking over the global market. Predicting the doom of Android. I remember that you yourself were praising Windows Phone as such a nice phone that could seriously damage Android’s market share.

    @Nigel
    “Given that Apple has already sold 500M iOS devices your guess is…well…probably low.”

    My prediction was about “installed base” of phones(!), not some devices sold historically.

    In five years, few if any of those 500M iOS devices will be in use anymore. Neither will be any of the 1 billion Android phones sold up to this summer (notice “devices” vs “phones”). So I predicted that in 5 years, iPhone will have retreated to the high end market, if they still run an iOS derivative, with 0.5B users max. There will still be (much) more than 5B mobile phone users world wide, and they will all use “Smartphones”. And these Smartphones will run either Android, what I expect, or another Linux system.

  83. Except that it has, just not in the US.

    I hate quoting Apple Insider but since it’s just IDC charts I guess that’s okay.

    http://appleinsider.com/articles/13/01/25/apples-iphone-grew-to-251-global-market-share-in-2012

    I would say that Apple 50% share in the US is home court advantage and that anywhere from 15-35% is probably about par looking at the national averages. A 25% world average is about what one should expect given historical performance. 50% is actually something of an aberration.

    So, no not so much. There are certainly markets they never addressed until recently like India but nowhere have they had majority share and subsequently lost it. Even in the US I don’t think they’ve ever been dominant (as in 70-80% share).

    You can’t say that someone’s market share has cratered/dropped/disrupted/whatever going from zero to 25% or zero to 15% or even zero to 5% in the last 5 years.

  84. Eric did not expect that the networks would subsidize iPhones to keep their market share artificially high. So, his error seems to be that he was wrong thinking the USA was a competitive market for mobile phones. In hindsight, this might look indeed a little naive. I am sure you saw this coming all the time but had your reasons to remain silent.

    So what changed? Smartphones has always been subsidized. As such it’s a level playing field in terms companies competing for subsidies from carriers. I’m sure Motorolla wasn’t crying foul when Razrs were top dog and they had the advantage of a hot product. But the handset market is certainly competitive. Ask all the Android handset makers about Samsung.

    For what reason would Verizon, who passed on the iPhone, who spend lots of money pushing Android, tied their brand to the Droid brand have to “artificially keep iPhone market share high”? If they could have deep sixed the iPhone and AT&T on a pure Android play they would have.

    Perhaps it was because people actually want iPhones and Apple makes a really nice product?

    I don’t think people have all that much loyalty to the fruit company. Folks buy Apple because they make good stuff. Not because of some weirdo conspiracy theory. When Apple stops making good stuff folks will stop buying it.

    Would they sell less without the subsidies? Sure.

    As far as the “silent” thing I’m not the one claiming this is Economics 101. You are. I’m not the one claiming clairvoyant predictive capability. You are. I’m not the one who claimed for two years that Apple was going to crash and burn. You were.

    This wasn’t some “Apple will fail someday” thing. This was an “Apple shareholders should sell their stock because the company revenue is about to crater” thing.

  85. My prediction was about “installed base” of phones(!), not some devices sold historically.

    In five years, few if any of those 500M iOS devices will be in use anymore. Neither will be any of the 1 billion Android phones sold up to this summer (notice “devices” vs “phones”). So I predicted that in 5 years, iPhone will have retreated to the high end market, if they still run an iOS derivative, with 0.5B users max. There will still be (much) more than 5B mobile phone users world wide, and they will all use “Smartphones”. And these Smartphones will run either Android, what I expect, or another Linux system.

    And you pulled these numbers out of which orifice? 0.5B users max based on what?

    And yes, installed base is calculated based on historical sales (to a point in time). And yes iOS devices vs Android activations because that’s what Apple reports and that’s what Google reports.

    “Apple has announced that iOS 6 is now installed on 300 million devices…Apple’s chief executive Tim Cook recently revealed that 500 million iOS devices have been sold thus far. Consequently, 300 million devices running iOS 6 would translate into 60 percent of the overall installed base.”

    http://www.tomshardware.com/news/iOS-6-Install-Rate-Devices-iPad-iPhone,20761.html

    The installed base is nowhere less than 300M today (500M according to the Toms Hardware author) and by 2018 you claim it will be no more than 500M?

    LOL. Like I said, you don’t fail to amuse.

    A curious thing is that Google talks activations but given they track the OS versions of Android devices that accessed Google Play within a 14 day window they could also give a decent indication of the android installed base without much guessing. I assume they can determine the number of unique devices based on udid.

    http://developer.android.com/about/dashboards/index.html

    Why do I get the impression that might be a far less impressive number than 1B device activations by fall of 2013? Wanna bet that the average iOS device has a longer used lifespan than the average Android device.

  86. You mean the one that states apple today has an installed base of 250m and 19% share? Up from 16% in 2011?

    And there’s no indication of methodology. But if you want to believe out of 300M active iOS 6 devices 250M are iPhones that’s fine.

    The one that states that apple will continue to do well and move up the charts along with Samsung?

    Why don’t you email him with your prediction that Apple will only double it’s installed base at best in the next 5 years.

    Whatever sources you want to use your numbers are not “optimistic” but hugely pessimistic.

    If you want to make a contrarian pessimistic prediction that’s fine. Everyone has an opinion and who knows, you may get blind lucky.

    If you want to claim that it’s optimistic and that you’re “serious” folks will laugh at you and not take you seriously at all give you have no data to support such an assertion, insist its based on current market trends which indicate something else and have been equally adamantly wrong for the last two years.

    I don’t believe that apples long term success is assured but for the next five years I believe that have enough product pipeline and inertia to continue to do well. 19% share in five years seems reasonable to me even if you assume they peak in 2-3 at 25% and start a decline back down to 19%.

    That’s mildly pessimistic to me but believable.

    Still Apple will likely introduce a lower priced phone eventually just like they did a 7″ iPad that they claimed they wouldn’t. If so then peaking in a couple years is less likely. Apple has no problems with self cannibilization when it senses its a good time for a lower tier model. They successfully managed the iPod product line that way.

  87. @Nigel

    What does 20% market share in 5 years mean? Say we do get 5 billion Smartphone users in 5 year. Then there should be 1 billion iPhones in use.

    An iPhone costs $600. Even in the rich US, Apple has only 50% market share. So 1 billion iPhones means 2 billion people as rich as the people in the USA . There are not that many Asians Africans or South Americans that make even $600 a month. I think more than 90% of humans prefer to buy a few $100 phones and eat than one $600 iPhone and go hungry.

    And if Apple starts making cheap products, would they really be able to outcompete the rest?

  88. Andy Rubin didn’t jump, he was pushed, by Larry page.

    This is a very similar situation Forestall at Apple. Not sure yet if his going to Google ex is the equivalent of Forestall becoming “advisor to the CEO” or if he really is going to be given a chance to do his own little thing. But Rubin definitely didn’t leave android on his own, and Sundar was his archrival within Google so he definitely isn’t happy about who took over.

    Watch this folks: Android will be redefined, (Android apps on Chrome OS) and Samsung will need to either go their own way (as if they’re not already), as Chrome won’t be put in the same position, now that Google has carrier relationships, and Motorola to build to spec.

    In other words, Google is about to fork android and kill the project.

    “Open always wins.”

    Also note: Google announced this on the eve of the S4 launch.

    Could.Not.Be.More.Clear

  89. @LeRoy
    “In other words, Google is about to fork android and kill the project.”

    I have read this in the last four years many times. And if these authors go on predicting Google will kill Android, one day they will be right. Be it this year or in twenty years.

    If you want to read an analysis from someone who actually thinks before he write, you can read:

    Android plus Chrome OS equals Google’s future operating system
    http://www.zdnet.com/android-plus-chrome-os-equals-googles-future-operating-system-7000012635/

  90. OTOH, If you want to read something that isn’t analysis from someone who is just barely awake at 2:00 am…

    Chaos pen’s poet’s a sword
    But Eris laughs at Discord
    Confusion won’t end – it’s ignored
    Bureaucracy – help me! I’m bored!
    Aftermath all goes to fnord

  91. I had just finished reading the Apocrypha
    Discordia
    , and I was reading the introductions to the Principia Discordia, and I felt the need, for the first time in my life, to attempt to write a poem – a lousy poem, of course – about… nothing.

    Since this post seems to be winding down, and a few folks here might recognize the meaning behind the nothing, I took the liberty of publishing it here.

  92. Hey – don’t let stop this post in its tracks.

    It seems that the OP “Emacs mode” thread has mostly played out, and the “South America” versus “south America” thing is, thankfully, spent, but the

    – Apple is resting on its laurels!
    – And damn expensive laurels they are.
    – If you care about the price, you aren’t part of the target market.

    meme has been running, off and on, since the 1980s – it can’t be over now!

  93. (Android apps on Chrome OS)

    Wouldn’t that be something. Perhaps Fabrice Bellard could be commissioned to write a Dalvik VM that runs inside Chrome the browser? :)

  94. @Jeff Read
    “Perhaps Fabrice Bellard could be commissioned to write a Dalvik VM that runs inside Chrome the browser? :)”

    I assume that is not enough of a challenge to Fabrice. Eric once played with the idea of writing a blog post about his more challenging enterprises. However, it never worked out, it seems.

    1. >I assume that is not enough of a challenge to Fabrice. Eric once played with the idea of writing a blog post about his more challenging enterprises. However, it never worked out, it seems.

      Thanks for the reminder. There are not many hackers that make me think “Holy crap, that guy is a wizard!” Fabrice is one of them.

  95. Thanks for the reminder.

    Do we need to dredge this whole blog for planned posts you have mentioned but never gotten around to / forgot about?

    — FooQuuxman

    p.s. ever hang out on #mises?

    1. >p.s. ever hang out on #mises?

      No – I didn’t know of it until you mentioned it, or if I knew had forgotten.

    1. >While we’re wandering afield, hey Eric what do you think of your shiny new Nexus 4?

      Terrific hardware, especially the display which beats anything I’ve ever seen on an iPhone like a gong. I’m less happy with 4.2.2, as they seem to have sacrificed some usability for visual slickness.

  96. > “Perhaps Fabrice Bellard could be commissioned to write a Dalvik VM that runs inside Chrome the browser? :)”

    You’ll need the following:

    nacl-dalvik: A Native Client implementation of the Dalvik Virtual Machine to allow Android apps that use Dalvik to be rendered as Native Client apps inside Chrome windows
    nacl-android-sdk: A Native Client implementation of the Android SDK APIs to allow full compatibility with all Java-based Android apps
    nacl-android-ndk: A Native Client implementation of the Android NDK APIs to allow additional compatibility with NDK apps (primarily games and other CPU-intensive apps)
    nacl-aapt: Native Client implementation of the Android Asset Package Manager, to enable Native Client-based Android package management (and installation of Android apps inside a NaCl-based compatibility layer).
    Mobile-optimized Aura/Ash form factor (based on the theming strategy of Chrome for Android) to scale Aura/Ash to mobile devices

    Then, on the business end:
    Merge of the Chrome Web Store into the Google Play Store as three categories: Chrome Apps, Chrome Extensions, and Chrome Themes
    Use of Chrome Web Store-style package management in the Google Play Store (i.e. apps can be downloaded/installed directly from the Google Play Store website instead of requiring a local Google Play Store app to be installed)

  97. Terrific hardware, especially the display which beats anything I’ve ever seen on an iPhone like a gong. I’m less happy with 4.2.2, as they seem to have sacrificed some usability for visual slickness.

    I got mine a few days after you got yours. Like you, I’m transitioning from G2 to N4. Overall I am absolutely floored, but I find I miss CM (was running 7.2 on the G2) in a few small ways.

    When you say 4.2.2, what do you compare against? The only 4.2.1 -> 4.2.2 changes I’ve seen are minor positives. I toggle BT and wifi fairly regularly so adding toggle capability to the quick settings buttons is nice, but the way they use the quick settings buttons is overall inconsistent and mildly annoying. CM’s handling is better, but I find just adding a power widget to one of my home screens is nearly as convenient as having CM-style notification quick settings.

  98. @Winter> “These seem rather simple and plain checkbox items for someone who succeeded in booting a Linux kernel plus CLI inside the JavaScript VM, in any browser.”

    yawn. there is a much more complete, ground-up project to build a 386-style emulator in javascript called jslm32. https://github.com/ubercomp/jslm32

  99. @esr> “especially the display which beats anything I’ve ever seen on an iPhone like a gong”
    You’re showing selection bias. Again.

    The Google Nexus 4 might have the larger, higher-resolution display, thrusting its 4.7in (1280×768-pixel) panel in the face of Apple’s smaller 4in (1136×640-pixel) screen, but that difference means the iPhone 5 actually packs more pixels per inch than the Nexus 4. There’s not a lot in it, with just 6ppi separating the 320ppi Nexus 4 and 326ppi iPhone 5, but a difference there is, and it’s in Apple’s favor.

    Both panels are IPS.

  100. @LeRoy:

    yawn. there is a much more complete, ground-up project to build a 386-style emulator in javascript called jslm32. https://github.com/ubercomp/jslm32

    This is a very silly troll. Any idiot who knows how to use a search engine can easily find where the author of jslm32 says “Last year, inspired by Bellard’s jslinux, I too wrote an emulator that can run Linux on the browser. Only I was lazy and emulated the vastly easier LatticeMico32 processor.

  101. Jay, my point is that Eric was claiming his Nexus 4 screen as vastly superior. I side with you, the difference is minor, and at most trivial.

    But hey… look who won the DoD’s mobile device competition to replace its Blackberrys. Left unmentioned is that DoD purchases have a large halo effect on contractors and agencies who don’t have the funds to do the security study and will follow DoD’s lead. . .

    http://m.electronista.com/fullarticles/13/03/20/sequester.holding.up.purchase.intended.to.replace.failing.devices/

  102. @LeRoy
    “but that difference means the iPhone 5 actually packs more pixels per inch than the Nexus 4.”

    This sounds like those people that claim a Hummer is more fuel efficient than a small economy car, per kilogram car weight. The whole point is that the Nexus four has a great *larger* screen and more pixels.

    @LeRoy
    “yawn. there is a much more complete, ground-up project to build a 386-style emulator in javascript called jslm32.”

    You yawn at a person who was the first who created a Linux capable emulator in JavaScript. I wonder what feat would then be worthy of your attention? Running a global distributed cloud supercomputer in “donated browser tabs” of participants?

  103. There’s more to screen appearance than mere resolution, LeRoy. It’s quite subjective. I don’t doubt for a moment that Eric found the Nexus 4’s screen superior to that of any iOS device he’s seen…which may very well *not* include an iPhone 5, given the circles he travels in.

    There’s a reason I tell people never to buy a monitor they haven’t seen in person. What looks good to one may not look so good to another.

  104. The difference between 320 and 326 PPI is equivalent to holding it less than half an inch further away from your face at a range of two feet, which [obviously, considering it has more pixels] still leaves the larger screen visually larger.

  105. LeRoy, if you can tell the difference between 320 and 326 PPI, your eyesight is superhuman.

    I’m no Apple basher, but please get real.

    In which case it would be extremely hard for the Nexus screen to “beats anything I’ve ever seen on an iPhone like a gong”.

    The screens are on par and simply a matter of personal size preference. Personally the thing I don’t like about the Nexus 4 is the lack of LTE (especially if you’re on Verizon) and 16GB limit because I have lots of apps and often use my phone as my camera to take video.

  106. Nigel, as I told LeRoy, there’s more to it than size and resolution…to start with, there’s brightness and contrast. It’s quite subjective.

    I’m not fond of the lack of LTE. The 16 GB max doesn’t bother me as much; I have over 8 GB free on my Nexus 7, and that’s with my 5 GB music library loaded.

  107. Hmm… I’m not into smart phones, so I’m not sure if this is the moment of the cusp, but we seem to have a sea change here…

    We have gone from: “Look how small my phone is!”
    To: “Look how big my phone is!”

    This may seem like a silly, borderline troll, comment, but a lot of folks are into this.

    1. >To: “Look how big my phone is!”

      One of the reasons an N4 beats the shit out of an iPhone for my usage pattern is the larger display. I use it as a portable web device a lot. Carrying around a 7-inch tablet would suck, but a display that is just shy of being too large for one hand suits me fine,

  108. They should just add phone functionality (including a mic and speaker to hold it up to your head) to 7″ tablets and be done with it, and let us sane people go back to 4″ phones.

  109. > I’m not fond of the lack of LTE.

    Wasn’t Eric predicting death of the iPhone a few years back because the iPhone 4 didn’t have LTE, and Android … did?

    1. >Wasn’t Eric predicting death of the iPhone a few years back because the iPhone 4 didn’t have LTE, and Android … did?

      I thought lack of LTE would hurt the iPhone4 pretty badly. This was one of my few actual predictive mistakes, as opposed to the pseudo-mistakes Apple fanboys have invented by misinterpreting and wildly overinterpreting what I wrote.

      At the time I thought LTE was going to roll out a lot faster and be a much larger factor in near-term purchasing decisions than it turned out to be. I’d feel worse about fluffing this if everybody else doing forward projections hadn’t made the identical same mistake. If anybody has an analysis of why LTE stumbled at the gate I’d be interested to hear it.

  110. Re: “Look how big my phone is!”

    I certainly wasn’t criticizing the concept, and of course the exclamation is really:

    “Look how big the display is on my wireless/cell-phone palm-top computer!”

  111. “This was one of my few actual predictive mistakes, as opposed to the pseudo-mistakes Apple fanboys have invented by misinterpreting and wildly overinterpreting what I wrote.”

    You’ve completely avoided writing for 6 months on a topic that you wrote on weekly because you were wrong for a year and a half while you kept claiming you saw evidence of your predictions occurring.

    1. >You’ve completely avoided writing for 6 months on a topic that you wrote on weekly…

      …because I no longer know where to get market-share numbers that I think I can trust, collected by methods that are consistent over time. Go away, cultist. You’re not impressing anyone.

  112. At the time I thought LTE was going to roll out a lot faster and be a much larger factor in near-term purchasing decisions than it turned out to be. I’d feel worse about fluffing this if everybody else doing forward projections hadn’t made the identical same mistake. If anybody has an analysis of why LTE stumbled at the gate I’d be interested to hear it.

    I’ll have a quick go. The technology simply wasn’t ready, supporting pieces weren’t there. Early chipsets WAY too power-hungry. Look at the HTC Thunderbolt, from which HTC still has yet to (and may never) recover. All those Verizon tv commercials of the phone being hit by lighting… and then you think about it and snicker, because that’s the only way to keep it charged.

  113. If anybody has an analysis of why LTE stumbled at the gate I’d be interested to hear it.

    Data caps. The sorts of applications where LTE has a strong advantage over 3G systems tend to, unsurprisingly, suck lots of data, and the non-Sprint customers I know have been trying rather hard to arrange to stream media over WiFi when possible. Additionally, since none of the carriers are using VoLTE, operating dual basebands sucks battery power, so that while subscribers aren’t seeing much advantage from using LTE, they do see the disadvantages.

    (I’m sure the intentional mindcloud around the term “4G”; the fact that many iPhone 4 users consistently referred to their phones as “iPhone 4Gs”; and the haziness over what exactly LTE offers as a specific technology, even among most tech-savvy consumers, didn’t help either.)

  114. @esr
    “…because I no longer know where to get market-share numbers that I think I can trust, collected by methods that are consistent over time.”

    I think the quarterly averages compiled by Tomi Ahonen are the most consistent over the years. He takes the global quarterly numbers of the big analyst houses and averages them. They track all the movements in the market quite faithfully.

    I posted a table of numbers since 2009 earlier above:
    http://esr.ibiblio.org/?p=4854&cpage=1#comment-396284

  115. On slightly unrelated note: Kiln now supports transparently and interchangebly both Git and (original) Mercurial: http://www.joelonsoftware.com/items/2013/03/11.html

    I guess that they use the same mechanism that reposurgeon and git’s remote helpers (like git-hg), namely the fast-export stream… which was originally invented from what I remember for fast conversion of Mozilla CVS repository to Git, which didn’t happen.

    It turned out that Kiln Harmony uses programming API (native Python API for Mercurial and Dulwich implementation of Git in Python; I wonder why not libgit2), which is admittedly a step up of Taylor-ish use of CLI tools to keep Git and Mercurial repositories in sync.

    The linked blog article describes how Kiln keeps Mercurial metadata in Git repository and vice-versa for round-trip equality… inclusing pains to be taken because of repositories from before a standard.

    That is of course completly different use case than using reposurgeon for one-time full conversion from one SCM to other; to make it as if the destination SCM was used from the very beginning.

  116. I have come to realize that I am a Discordian.

    Rather than polluting this thread with more of my weirdness, I would like to invite everyone that is interested to visit the…

    What Would Eris Do?

    website at: http://www.wwed.info

    It is only one page now (just for the folks at “Armed and Dangerous”) but a prettier website with more content will be available sometime soon.

    If our host (that’s Eric, not Eris) does not object, I would like to post a comment on an aging thread when a much better version of the website is available.

    If ESR does object… my apologies for going cosmically off-topic.

    1. >If ESR does object… my apologies for going cosmically off-topic.

      Hail Eris! Objecting would be aneristic of me. Perish the thought. All hail Discordia!

  117. Hail Eris!

    For folks that saw the WWED website before… I think it would be 1:55 pm blog-time,
    I have changed one very important paragraph into three much better ones…

    I believe in Goddess, Eris, our Lady of Chaos, not as a supernatural being that literally exists, but rather as something deep in our minds. This is how any rational neopagan views gods and goddesses. The word “supernatural” is, in a sense, nonsense – anything that exists is part of the universe and its laws. Of course, there is a lot about the universe we don’t know, and this is particularly true of the human brain and mind.

    For some people, being funny is more important than… let me give you an example (from memory): In one of Rex Stout’s Nero Wolf books, Archie was told: “Someday, trying to be funny is going to get your head blown off.” Archie, narrating, thinks to himself, “It was after I left that I realized that it wouldn’t mean it wouldn’t be funny.”

    Discordians are deeply into humor that is both bizarre and makes one see an aspect of our chaotic world in a new way. Humor involving up-tight rule-oriented people being confronted by chaos is also very high on the list.

    All Hail Discordia!

  118. @esr> …because I no longer know where to get market-share numbers that I think I can trust, collected by methods that are consistent over time.

    Doesn’t address that many were calling bullshit while you consistently backed the comscore data, until it no longer supported the assertions you were making. 18 months of believing / endorsing the hype, Eric.

    > Go away, cultist. You’re not impressing anyone.

    except those of us who appreciate great points made in the continuing debate. We’re all impressed.

  119. @esr> At the time I thought LTE was going to roll out a lot faster and be a much larger factor in near-term purchasing decisions than it turned out to be

    It turned out that LTE was waiting for the iPhone.
    Sprint accelerated its LTE rollout to get the iPhone.
    http://allthingsd.com/20120910/ahead-of-new-iphone-launch-sprint-promises-lte-is-coming-to-lots-of-places-at-some-point/

    Even AT&T hurried up ahead of iPhone 5:
    http://www.theverge.com/2012/9/18/3355886/att-lte-live-in-seattle-ahead-of-iphone-5-launch

    Now t-mobile is, too:
    http://opensignal.com/reports/t-mobile-lte-rollout/

    In fact, if you look at the Open Signal coverage maps (which are all generated with Android), you’ll find that LTE is already fairly well deployed in the major metro areas and Interstate Highway corridors.

    It’s just that your phone doesn’t have it.

  120. @R Duke
    “Doesn’t address that many were calling bullshit while you consistently backed the comscore data, until it no longer supported the assertions you were making. 18 months of believing / endorsing the hype, Eric.”

    Last year, 700 million Smartphones were sold globally. 220 million in 4Q2012. According to ComScore numbers, around 50% of Smartphones sold in the USA are iPhones.

    Please, explain to me what useful information you can extract from the ComScore numbers about the global Smartphone market in 2012?

    Do the ComScore tell you *anything* worthwhile about the number of iPhones sold world-wide? Does it explain anything about how many iPhones will be sold this year?

    It is a mystery to me what useful information you can extract from the ComsScore numbers?

    There was a time when ComScore numbers were good indicators of global trends. But that was before the USA providers started to “differentiate” high-end phones on OS.

  121. “…because I no longer know where to get market-share numbers that I think I can trust, collected by methods that are consistent over time.”

    That’s you problem, not mine. There is no shortage of data from numerous sources that is consistent and proves you were wrong for more than a year. Your inability to accept these numerous sources of data or you sudden distaste for Comscore data because it doesn’t agree with your theory that was wrong is none of my concern. Remaining silent and defensive on the topic speaks volumes for the indefensibility of your incorrect theory.

  122. @Spock:

    Do you have anything useful to say? It is clear that negotiating with Apple is similar to negotiating with the mob:

    http://www.dailytech.com/European+Carriers+Tattle+on+Apple+to+EU+for+Anticompetitive+Contracts/article30188.htm

    Nice network you have there. It would be a shame if our phone didn’t run on it.

    Yes, I know this tactic only works because a lot of people actually do want the iPhone. But Apple is leveraging that demand into forcing the carriers to convert more customers to the Apple fold. It’s a zero sum game for everybody except Apple. And when Apple stumbles (and they will) all the carriers will remember. They will salt the earth with the remains of Apple just like they are doing with Nokia after the Microsoft hook-up. Nobody likes a bully. Especially other bullies.

  123. @Spock
    No, I am always like that.

    But, what do the ComScore numbers tell us? Will iPhone do well this year or not? Did they do well last year, compared to Android?

  124. It seems that the OP “Emacs mode” thread has mostly played out, and the “South America” versus “south America” thing is, thankfully, spent, but the . . .

    I came wandering in here to see if there was some announcement that Eric was out-of-commission and am of course relieved that he is alive and well. Seems like y’all might be searching for something significant and controversial to discuss, so I hope it is okay if I throw in my thread jack on the topic of Bitcoin. Drilling down the links, you’ll arrive at my claims on Bitcoin’s fatal flaws and my specification-in-progress for an alternative design. Note Bitcoin’s value has increased from $10 to $70 in the past 3 months.

    1. >Is it reposturgeon or reposurgeon? You seem to use them interchangeably.

      For the forty-million-and-eleventh time…”reposurgeon” is the project, “the reposturgeon” is the project’s mascot.

  125. But, what do the ComScore numbers tell us? Will iPhone do well this year or not? Did they do well last year, compared to Android?

    It tells us the same thing it told us for the last four years: nothing conclusive.

    Which is the point. When this flawed data supported your position you and Eric defended it. Now when it doesn’t and shows the opposite you claim it’s not trustworthy. That’s intellectually dishonest.

    They have not changed ownership nor have they changed methodology. Nor has US market practices (re subsidies) changed vis a vis global practices in any great extent except perhaps slightly in apple’s favor in china.

    There is no current data and never has been in the past that Apple is on the verge of “disruption from below” and about to suffer a huge revenue collapse.

    You guys have been wrong about that since 2009 or 2010 or so.

    But to answer your question: apple did do well this year and last year and the year before in terms of YoY unit sales increases and revenue even in markets nearing saturation or largely dominated by Android or both.

  126. I remember when I commented before years ago, I said that iPhone wasn’t readily available in retail in Philippines, but now I do see at least the iPad stocked and my landlord and her upper middle class friends carry iPhones.

    The vast majority of smartphones I see people using are apparently Android, since they don’t look like iPhones. So I still maintain that Android is winning in terms of volume, and we had this debate before (which I will not re-engage in) about whether marketshare amongst the lower class or revenue share in (some of) the developed countries is more important. Or if either can point to a disruptive end where Apple ends up with say single-digit share again.

    This week I think I saw a dual-simm Samsung Duos on sale for roughly $120, which is two to four weeks months salary here. Dual-simm is very important in this market because it is expensive to text and call between three major carriers. AFAIK there is no dual simm offering from Apple.

    For one perspective, I recently coded a cross-browser extension (Chrome, FF, IE, and Opera), and I will not be bothering to port it to Safari, because my extension loads a script (so I can make real-time bug fixes) and I am told this is never allowed in the iOS App store. That 14% global share will just have to be lost as I have too many projects to work on any way that can generate sales and income.

    I am thinking to start developing apps, and I won’t bother with iOS, because I think the long-term is a larger market of rising incomes for Android, versus a shrinking market of declines incomes for iOS once the USA (is last to) falls into the sovereign debt collapse abyss circa 2017.

    Any way, sorry I don’t have any statistical data to offer to back my opinions.

  127. Apple’s solution to expensive text is iMessage which attempts to shield iOS users from abusive carrier practices.

    Something Apple does and Google does not further highlighting that Apple customers are us users and not carriers, handset maker or advertisers. If carriers crush apple and salt the earth then expect abusive cross network charges and idiotically high SMS costs to remain the norm.

    Frankly many FOSS zealots ignore the huge positives that MS provided by breaking the IBM computer business model and making computing inexpensive. No MS then no commodity hardware and then no Linux. It made MS a lot of money but it was also good for users.

    Likewise folks like Patrick liken apple to the mob without understanding that many cell markets are dominated by carriers without true competiton…which is why they get away charging out the rear for texting, data and even for ringtones. App stores were fragmented, limited and carrier dependent.

    Apple breaking them and turning them into dumb pipes as much as they have makes them a lot of money but is also good for the average user. Something the dominance of android has retarded or regressed.

    Anyway, app devs that ignore either market are simply hurting their own revenue over ideological bias given the cross platform game engines and one cross platform dev platform (xamarin).

  128. Apple’s solution to expensive text is iMessage which attempts to shield iOS users from abusive carrier practices.

    Something Apple does and Google does not further highlighting that Apple customers are us users and not carriers, handset maker or advertisers.

    Erm. If Apple really had a soft spot for end users, it would make iMessage available to non-Apple platforms (e.g., via an XMPP gateway like Facebook does with its messaging). The fact that they don’t (and that, despite explicit promises, they haven’t released protocol specs for FaceTime, which itself could have been implemented via standard protocols) shows quite clearly that what they’re interested in is locking people into an Apple ecosystem for their communications.

    Frankly many FOSS zealots ignore the huge positives that MS provided by breaking the IBM computer business model and making computing inexpensive. No MS then no commodity hardware and then no Linux. It made MS a lot of money but it was also good for users.

    Assumes facts not in evidence. Most glaringly, that had Gary Kildall not been out on the golf course, IBM would have managed to get some sort of exclusive deal for CP/M (of which QDOS was a hackish clone), which was in fact already available on multiple platforms.

    one cross platform dev platform (xamarin).

    So the answer to cross-platform app development for iOS/Objective C and Android/Java is… .NET?!

  129. @Nigel
    “Something Apple does and Google does not further highlighting that Apple customers are US users and not carriers, handset maker or advertisers.”

    I actually misread your comment first like above. But I realized you did not mean customers is the USA but yourself and other buyers of Apple products. However, Google does not sell Android, and Google customers are neither the carriers, nor the handset makers. The incentives for Google are getting people unfettered access to its services.

    So the differences are that Apple becomes richer the more it can get people to pay for their handsets and apps and Google becomes richer the less people have to pay for handsets and internet access and the more they are able to spend on other products.

  130. I suppose I learned the importance of using Occam’s Razor from this site some years ago. I assume the participants enjoy logic puzzles or challenges.

    I hope I haven’t muffed the logic on this one, Any counter-proof that Satoshi Nakamoto did not design a ponzi scheme on purpose?

    I didn’t know that James A. Donald was the first outside developer interacting with the creator of Bitcoin. What a small world is compressed into the participants of this site.

  131. @Nigel
    The availability of a even an open-source cross-platform framework is not sufficient when there are other tsuris costs that impact my degrees-of-freedom.

    Seems you are not using the same time-preference, opportunity-cost equation that I am.

    Since there is only one of me and due to the Mythical Man Month, it makes more sense to forsake incremental near-term revenue if it means doing less work and gaining significantly more long-term revenue.

    What is most important to me is that my work gains traction and isn’t orphaned. Profit is not the scarce good here.

  132. What is most important to me is that my work gains traction and isn’t orphaned. Profit is not the scarce good here.

    Then why did you write:

    That 14% global share will just have to be lost as I have too many projects to work on any way that can generate sales and income.

  133. But I realized you did not mean customers is the USA but yourself and other buyers of Apple products. However, Google does not sell Android, and Google customers are neither the carriers, nor the handset makers. The incentives for Google are getting people unfettered access to its services.

    So the differences are that Apple becomes richer the more it can get people to pay for their handsets and apps and Google becomes richer the less people have to pay for handsets and internet access and the more they are able to spend on other products.

    Yes, I mean that the end users of Apple products are their actual customers unlike Google. The customers for Google are companies that buy ads and not the end users of their customers.

    So the differences are that Apple becomes richer the more it can get people to pay for their handsets and apps and Google becomes richer the more eyeballs and information about those eyeballs they can sell.

    If more eyeballs requires them to let carriers and handset makers put a lot of cruft on their phones then that’s okay even if it negatively impacts the user experience.

  134. Erm. If Apple really had a soft spot for end users, it would make iMessage available to non-Apple platforms (e.g., via an XMPP gateway like Facebook does with its messaging). The fact that they don’t (and that, despite explicit promises, they haven’t released protocol specs for FaceTime, which itself could have been implemented via standard protocols) shows quite clearly that what they’re interested in is locking people into an Apple ecosystem for their communications.

    Apple has a soft spot for their end users first and negotiates that way. I suspect that one of the reasons that carriers didn’t push back as hard on that issue is because it is currently limited to just other iOS devices.

    Facetime was limited to WiFi for a long time. Yes, they never released the spec which is a shame but it is implement on top of open standards: H.264, MPEG AAC, SIP, RTP, Secure RTP, TURN, STUN and NAT. Authentication and hosting is locked down.

    iMessage uses APNS but iChat uses (used?) XMPP. I was a big fan of XMPP but it ended up a little balkanized anyway despite the open standard.

    If you don’t like iMessage there’s always GTalk…until they merge it with Messenger and Hangout and I guess XMPP gets dropped in favor of WebRTC.

    Assumes facts not in evidence. Most glaringly, that had Gary Kildall not been out on the golf course, IBM would have managed to get some sort of exclusive deal for CP/M (of which QDOS was a hackish clone), which was in fact already available on multiple platforms.

    He wasn’t on a golf course, he was delivering software.

    It is how history played out. Every other PC manufacturer, including Apple, favored selling an integrated product. So was IBM. It was in MS’ interest to commoditize hardware since it made money selling software to customers. The same can be said of Google.

    The difference to me is I prefer to be the customer of the product being developed as opposed to being the product being delivered to the actual customer. Businesses tend to try to keep customers happy and less so for the product being sold.

    I think that some folks are discovering that being the actual customer is actually useful after the cancellation of Google Reader.


    So the answer to cross-platform app development for iOS/Objective C and Android/Java is… .NET?!

    Yah, kinda weird but it is what it is. You code your business logic in a platform neutral C# but code the UI against the cocoa or dalvik bindings.

    You can do this in just C++ but it’s probably mildly more annoying.

    It is interesting that Oracle appears to be putting some effort into JavaFX. Perhaps Java as a cross-platform desktop development environment isn’t quite dead yet.

  135. @Nigel:

    What is most important to me is that my work gains traction and isn’t orphaned. Profit is not the scarce good here.

    Then why did you write:

    That 14% global share will just have to be lost as I have too many projects to work on any way that can generate sales and income.

    What logic system makes sales and income mutually exclusive with adoption?

  136. @Justsaying

    None. You had claimed I was using a different metric. I was merely pointing out that you previously provided only provided sales and income as the metrics you were using. If what was MOST important was traction then I would have expected that to have been mentioned first in the prior post and not sales and income.

    From a sales and income perspective iOS appears to have a significant advantage. Local trumps global so obviously whatever the conditions in your local market are most important. Presumably that is the RP.

    14% global share means far less than what your home market is doing followed by the US, EU and Chinese markets (in no particular order).

    That means for US app developers the Comscore numbers are relevant since they indicate broad trends for their home market with some largish error bars.

  137. @Nigel:
    I have to consider sales, because it is ad funded economics demand it. You did guess correctly the extension targets SE-Asia, esp RP. Dating. The extension is likely to be more used by the foreigners though, so I have to consider the western browser shares. And they spend more than RP users. If 14% don’t use the browser extension, they can still use the normal features of the web site that don’t require the extension. So I am not losing all 14%, yet Apple is losing the features (makes Apple look bad). Seems to be an asymmetric loss for Apple. This is why I point out you shouldn’t analyze the vectors in such simplistic terms. Degrees-of-freedom quickly explodes the possibilities you can analyze– you have to just embrace it.

  138. @Nigel
    “If more eyeballs requires them to let carriers and handset makers put a lot of cruft on their phones then that’s okay even if it negatively impacts the user experience.”

    The next 3 billion handsets are low margin as these are from people with a monthly income lower than the price of an iPhone or Galaxy S4. This reasoning means that Apple will not make money from them, so will not sell them handsets. Google will make money from their eyeballs and the hardware adds no cost to Google. So Google will support anyone who sells these people handsets (a billion times a dollar is still a billion dollars).

    It is the same dynamics as with PCs. Microsoft had an incentive to force down hardware prices, as this sold more licenses. So has Google. And Apple stuck at the high end then, as now.

  139. And again, as soon as Eric is challenged on the Comscore data, or Android’s “continued dominance” he bails for a new posting, and most of his readers follow.

    So it has always been, so it shall always be.

  140. > Please, tells, what can we learn from the comscore numbers?
    Only that eric was either dead wrong, or seriously led astray by his own ego.

    Also: T-Mobile ships iPhone (5, 4s, and 4)

    http://arstechnica.com/apple/2013/03/t-mobile-iphone-goes-from-myth-to-reality-580-but-no-contract/

    Waiting to see what the Samsung Galaxy s4 will cost on T-Mobile. The Galaxy S3 is $600, a bit more than the iPhone 5.

    I am ROFLMAO at all the Android fanbois who carped “Apple is expensive”.

  141. > The next 3 billion handsets are low margin as these are from people with a monthly income lower than the price of an iPhone or Galaxy S4.

    The next 3 billion handsets are upgrades to the installed base. Mass consumerism FTW.

  142. @winter actually google is not likely to make money on all those eyeballs given local competiton and lockout.

    A significant percentage of your 3B handsets will be sold in China.

    Want to bet that Apple will make more money in China than Google will?

    The real android winner is Samsung…and depending on the licensing fees perhaps Microsoft.

  143. @LeRoy
    “The Galaxy S3 is $600, a bit more than the iPhone 5.”

    Samsung covers the whole price range of phones. Apple only the high end. If you want to sell big in Asia, Africa, and South America, you have to offer low cost models.

    @Nigel
    “A significant percentage of your 3B handsets will be sold in China. ”

    Less than a third.

    @Nigel
    “Want to bet that Apple will make more money in China than Google will?”

    What does it matter? Even if there is no return from China to Google, there are also no costs involved for Google. Basically, every Android phone sold in China is a free lottery ticket for Google. But I do not see why I should worry whether Google makes more money than Apple?

    It seems to be difficult to grasp, but many of those who come here out of interest for Android, have no interest in the profits of Apple or Google. If you are interested in profits, you might consider visiting a banking or money laundering blog. A technology blog seems less relevant, as a successful technology tends to reduce the profit margins.

  144. This may come out gibberish, as I haven’t slept for 36 hours. Please tell me if I am irrational, delusional, paranoid, or even just wasting your time. Apologies in advance and I won’t try to repeat. But I feel I just have to share what I learned, maybe some of you know this already, but I am really curious what you hackers think about this? Nothing?

    I have discovered that the design of Bitcoin is such that there are no transaction fees to incentivize the P2P miners to include transactions in the block chain. Thus as the geometric debasement reward winds down (halving every 4 years), the only way transactions will move through the system is subsidized. The plan is that the mega-corporations such as Walmart, will offer server farms because they will gain leader advantages. The way Bitcoin works is the difficulty of the hash solution for moving the block chain forward to confirmation transactions, it scales to the level of total hash computational power in the system. So this means if mega-corporations invest $10 billion in ASICs, then no small players can compete to process transactions.

    The Bitcoin developers seem to think this is better because at least it wipes out credit card fees Rather I see that this gives the monopoly to the same entity. Before you had retailers fighting against processors, now they will be one in the same. So basically they can take over our digital life and charge us what ever they want. We will have no choice.

    The design doesn’t have to be this way. It was arbitrarily chosen by an anonymous person Satoshi.

    Please just tell me if this is inappropriate for this blog and I won’t do it again. I am thinking you all value freedom too much. Please set me straight in my thinking. Help me, how to think about this? It is making me depressed.

    http://bitcoin.stackexchange.com/questions/8671/any-counter-proof-that-satoshi-nakamoto-did-not-design-a-ponzi-scheme-on-purpose/8728#comment11454_8728

    http://bitcoin.stackexchange.com/questions/876/how-much-will-transaction-fees-eventually-be/8749#comment11471_8749

  145. @JustSaying
    “The design doesn’t have to be this way. It was arbitrarily chosen by an anonymous person Satoshi.”

    No, this is not gibberish, just uninformed, or partially informed.

    First, Bitcoin’s validity is based on “Proof of Work”
    https://en.bitcoin.it/wiki/Proof_of_work

    You should read up on Proof of Work until you understand why this is fundamental to the whole Bitcoin system. I know of no alternative. Btw, this is also the basis of Bitmessage, an anonymous P2P message system:
    http://bitcoinmagazine.com/bitmessage-a-model-for-a-new-web-2-0/

    The Bitcoin system is “safe” as long as no consortium can revise history fast enough to overtake the real transaction history. Rule of thumb, no entity has half or more of the computing power. Those who sign the transaction history have to be paid, either in new Bitcoins or, when all are mined, in old Bitcoins. That is not a “problem”, but like death, inevitable. See it as a subscription to use the system.

    But Bitcoin is not a employment project to give the “little ones” a source of income. Bitcoin is a payment system that has to be run efficiently and safe.

    Bitcoin has many problems, hoarding, deflation etc. But the Proof of Work system is the one thing that has a solid foundation in cryptographic theory.

  146. What does it matter? Even if there is no return from China to Google, there are also no costs involved for Google. Basically, every Android phone sold in China is a free lottery ticket for Google. But I do not see why I should worry whether Google makes more money than Apple?

    There are costs involved for Google since the investment in Android probably was based on some prediction of profits from China. A market they arguably would have lost anyway but it’s not hard to believe that the worlds largest handset market factored into planning at some point.

    If Google didn’t care then it wouldn’t have (relatively recently) killed the Acer/Alibaba phone. Obviously they do and for good reason. The android ecosystem can be hijacked by Baidu, Alibaba, Naver, Amazon, etc. Even Samsung. If this were to happen in a larger scale then Google will likely move Andy Rubin away from Android and drive toward something else that is more locked down like some kind of Android/Chrome hybrid under someone like Sundar Pichai and cease significant investment in Android as we know it. If that happens you might have a Google Reader moment…

    Oh…hey…

    It seems to be difficult to grasp, but many of those who come here out of interest for Android, have no interest in the profits of Apple or Google. If you are interested in profits, you might consider visiting a banking or money laundering blog. A technology blog seems less relevant, as a successful technology tends to reduce the profit margins.

    If you don’t want to discuss money then why did you write:

    This reasoning means that Apple will not make money from them, so will not sell them handsets. Google will make money from their eyeballs and the hardware adds no cost to Google. So Google will support anyone who sells these people handsets (a billion times a dollar is still a billion dollars).

    I was speaking to the advantages of buying products designed for the actual user rather than using products designed to turn the user into the product. I believe that the user experience is typically superior in the former than the latter.

    I disagree that successful technology reduces profit margin. Successful technology reduces costs. That might or might not lead to lower revenue and profits. Profit margins are whatever the markets will bear.

    In any case, this blog specifically discussed the economic collapse of Apple so crying foul because it didn’t turn out that way and now you think it’s off topic to discuss money and want me to go away is both funny and sad.

  147. I was speaking to the advantages of buying products designed for the actual user rather than using products designed to turn the user into the product. I believe that the user experience is typically superior in the former than the latter.

    Note: When I wrote this I was thinking of Evernote over Google Keep and Dropbox over Google Drive as much as Android vs iOS.

    If Google manages to kill Evernote or Dropbox by dumping free products on the market subsidized by their adword revenue in order to kill the competition that’ll be sad for users. That’d be worse than anything that MS ever did.

  148. @Nigel
    “If you don’t want to discuss money then why did you write:”

    We know Apple is guided by profit margins from the sales of gadgets and services. Google is guided by profit margins on eyeballs. Neither should really care whether the other makes more or less profit than they themselves. Good entrepreneurship is making a profit in your market of choice, not some pissing game.

    @Nigel
    “There are costs involved for Google since the investment in Android probably was based on some prediction of profits from China.”

    Less profit is not nice, but is really not accounted as “Costs” according to the standard definitions. Your reasoning seems to be that Google expected a monopoly on Mobile OS’ and Mobile Advertisement, and therefore, that Google partitioned their investment accordingly.

    I would seriously question any such assertion that Google invested in Android under the assumption of World Monopoly on mobile handset OS’. Not without evidence at least.

  149. @Nigel
    “That’d be worse than anything that MS ever did.”

    Given the criminal record of MS, that is an exaggeration.

  150. @Winter:

    You should read up on Proof of Work until you understand why this is fundamental to the whole Bitcoin system. I know of no alternative.

    I appreciate your reply very much, and I am urging all participants here to please read the following. If it is proven that I am Chicken Little, then you can ignore me in the future.

    I’m rested now.

    I have devised a clever way to do Proof-of-Work with hard-disk space so as to drastically reduce the wasted massively duplicated hashing computational load on the transaction block mining peers. I even addressed someone’s technical inquiries about my alternative design. My design employs the principle of distributed hash table for random round-robin selection of the next peer to process each next block. The Proof-of-Work in my design is the ability to store a large number (e.g. a terabyte or what ever we decide) of shared keys among co-peers to prove work. Thus most of the peers are not continuously computing a hash problem, as in Bitcoin’s design.

    I am all about open source here. I can care less if there is any profit here for me. Any one is free to copy my design. I am genuinely concerned that we are about to step into the abyss of mega-corporate monopolistic control over who has permission to transact. AFAICS the implications are far reaching into all our lives, and even into brick-and-mortar retail.

    The primary problem is not the computational load of Proof-of-Work, although it is a factor as I will explain below. AFAICS, the critical fatal flaw of Bitcoin w.r.t. to our digital freedom, is that Satoshi arbitrarily decided that the transaction block processing peers would not automatically receive a transaction fee, and that the reward for mining blocks would decline geometrically over time– halving every 4 years. Readers should note that the automated scaling difficulty (of solving the hash problem for Proof-of-Work on each 10 minute transaction block) is such that if any entity puts massive mining hardware into the Bitcoin system, then the peers with an order or two less processing power than the monopolistic’s server farm, will almost never be able to solve the hash problem first, and thus never participate in the processing of transactions.

    The key problem is that if you remove the profitability of mining these transaction blocks, then only the monopolists can subsidize and participate. I propose instead that everyone has a hard-disk and we should continue the smallish rate of debasement (and/or system-wide small flat transaction fee) forever. Actually debasement is absolutely required for a monetary system, but I will save that argument for later, as we have a choice and can also solve the problem with transaction fees if so be desired.

    The impact of Satoshi’s design decision is that in the near future, the processing of transactions will need to be free (he even anticipated this). But we know nothing is free, thus the model encourages monopolization. If the second order implications are not clear to any reader, ask me to clarify how I get from there to the conclusion.

    According to the wiki I read and also the religious fervor about “Bitcoin is better because no debasement”, the geometric phaseout of the debasement is cast in stone, non-negotiable. And the Bitcoin proposal to add a market-based transaction fee won’t scale. So AFAIK there is no future for Bitcoin that is different than monopolization of our digital monetary transactions.

    The scariest aspect from my perspective is that we can’t later subvert it technically once it gains market share. For example, I contemplated whether we could later offer to move value from the Bitcoin system to an new open source system without requiring any transaction from the future monopolized mining server farms. It can’t be technically done, because it opens double-spending once on each orthogonal system. Thus the only way to defend against this claimed global financial trojan horse, is to provide a competitive open source system now, that can gain significant market share now.

    The following is my interpretation of market action and is not central to my points above.

    Bitcoin is gaining exponential market inertia right about now, with the price of a digital coin going from $10 to $76 in the past 3 months, yet there is no significant transaction volume. Dormancy is north of the measurable lower-bound of 59%, and given so many anonymous transactions of sending money to oneself via new addresses and FX, then real dormancy is likely above 75%. What gives currency value are either the velocity of exchange from the Quantity Theory of Money, or the monopolistic forces of kings (for gold), government (legal tender) or now ostensibly corporations (for bitcoin). Analogous to facebook, the myopic general public of speculators are funding the loss-leader via speculation of mostly valueless investment.

  151. @JustSaying
    Before you try to come up with a new cryptographic procedure (proof of work is a cryptographic procedure) which is rarely a good idea, you might want to read somewhat more into bitcoin. These are two links to discussion of Bitcoin research on Bruce Schneier’s blog

    https://www.schneier.com/blog/archives/2012/03/bitcoin_securit.html
    https://www.schneier.com/blog/archives/2012/10/analysis_of_how.html

    Note that the proof of work scheme is really not the most urgent problem of Bitcoin, nor the most fundamental one.

    “In addition, we isolated all the large transactions in the system, and discovered that almost all of them are closely related to a single large transaction that took place in November 2010, even though the associated users apparently tried to hide this fact with many strange looking long chains and fork-merge structures in the transaction graph.”

  152. @JustSaying
    About Bitcoin being a Ponzi scheme.

    I think it could have doubled as a ploy to fleece the paranoid and criminal. But this is done like a normal banking operation. It is almost inevitable that setting up a currency makes you rich. The Bitcoin system itself is based on solid cryptography and decent economics. Given the paranoid fear of inflation, a deflationary system actually appeals to some users.

    Bitcoin is a currency and is used in real transactions. It has a productive function. However, a Ponzi scheme does not have a productive function. It simply routes around money.

  153. @Winter:
    My concern has nothing to do with security from the Proof-of-Work.

    I have search far and wide and have found no research on my concern.

    My concern is the economic game theory model, specifically how it impacts who will pay for the Proof-of-Work.

    I am pointing out that the current design results in monopolization.

  154. @Winter:
    I am no longer focusing on whether Bitcom is a ponzi scheme or not, as it is irrelevant to my concern as written above.

  155. @JustSaying
    “I am pointing out that the current design results in monopolization.”

    The whole point of the Proof of Work is that it breaks down if it is monopolized. If one party can mine faster than all others combined, this party can rewrite Bitcoin history and insert/delete transactions at will.

    Anyhow, the scheme is very neat, but I think Bitcoin is irreparable flawed at a fundamental level because it is deflationary.

  156. @Winter:
    Let me clarify that I am not saying that security should be compromised. I am saying that my concern is not in the area of security nor in whether it is ponzi. My concern is the design yields a game theory which results in monopolization of transaction processing by corporations.

    We have that now per se with credit card companies, but the distinction is the game theory points to a unification of processing and retailing, into an even larger artificially induced monopoly with even less degrees-of-freedom (than what we have now).

    Imagine now we have a plurality of payment processors for credit cards. In the future, payment processing will be a unprofitable business controlled by say Walmart globally.

  157. @Winter:

    The whole point of the Proof of Work is that it breaks down if it is monopolized. If one party can mine faster than all others combined, this party can rewrite Bitcoin history and insert/delete transactions at will.

    Exactly. So why is you think it is not bad that Walmart will get this power if Bitcoin succeeds in the marketplace?

  158. @Winter:
    Actually you didn’t say it wouldn’t be evil if Walmart gets such power, but you seem to think there is no threat here?

  159. @JustSaying
    “Actually you didn’t say it wouldn’t be evil if Walmart gets such power, but you seem to think there is no threat here?”

    Even before Wall-mart gets that power, Bitcoin will break down because Wall-mart will be able to change the transaction history. Currency is a matter of trust. No trust, no currency. And if Bitcoin ever gets large, states will step in to regulate it.

    I see that I have been not clear. I think that Bitcoin will fail. Not because it is gamed, or unsafe (both are true). But because it is based on the idea that you can start an unregulated currency in an industrial society. That will never work.

    Case in point: There are barter systems all over the world in every industrial society. They never take over a significant part of the local economy.

  160. @Winter:

    Even before Wall-mart gets that power, Bitcoin will break down because Wall-mart will be able to change the transaction history. Currency is a matter of trust. No trust, no currency. And if Bitcoin ever gets large, states will step in to regulate it.

    Applying Occam’s Razor, if Walmart wants to have the power, they wouldn’t destroy it by corrupting most transactions. Agreed it will be regulated.

    Big government regulation + corporate control = _______ ? (Hint: Fasc….)

    I am concerned about the ability of the corporate takeover of government and the ability to turn off an individual’s ability to buy and sell.

    Cash will be eliminated. Gold and silver dealers are being regulated. Anti-money laundering laws are spreading every where. The USA is bullying every country into controlling the black market.

    I think that Bitcoin will fail. Not because it is gamed, or unsafe (both are true). But because it is based on the idea that you can start an unregulated currency in an industrial society. That will never work.

    Case in point: There are barter systems all over the world in every industrial society. They never take over a significant part of the local economy.

    Why do you assume it will be unregulated?

    The design appears to be to transfer the power to the regulators and corporations (one in the same in my view).

    We could do our part to help make it so, by offering a better open source alternative. I can’t do that by myself. Where is our James A. Donald alias JAD? From the cryptography mailing list, he was the first developer to interact with Satoshi. I am curious to know his thoughts on this matter.

  161. @Winter:
    I should clarify that I agree with you that anything large will be regulated. However, if the technical design of the open source that wins is such that the regulations can’t stop us from transacting, then we’ve won as a human race. Otherwise, we are headed into a Dark Ages.

  162. I have so many typos over my prior posts. Clearly my mind is still in that fog from sleep deprivation over the past several days while I was digging for the complete analysis of Bitcoin (that I now have loaded in my head). When I’m tired, my logic is mostly maintained at a high level, but my english engine sputters. Apologies to readers.

  163. @Winter:
    yes of course thank you ;) Our individual lives can still go forward.

    I just think that if perhaps we provided an alternative design for the P2P currency, at least we in the open source would have done our best, then let the market decide. We might be surprised and enough people might chose to use the design that can’t be subverted so easily by the monopolists. Even a very small minority market-share would mean we still have an alternative transaction market. The sooner we build it, then it will be ready with the global economy goes “tits-up” 2016ish.

    The technical discussion would shift to is that a possible goal.

    Btw, I agree with you that there must be some debasement. The reason is that society will never allow that old capital can enslave new capital. I summarized it here as I quote.

    To understand money, you need to understand the economy, see How an Economy Grows and Why It Doesn’t.

    Bitcoiners will rationalize that gold is a superior money because they claim that no one can debase it. This so stupid. First of all, both inflation and deflation concentrate wealth to the rich. Secondly, because of this concentration of wealth under gold to the rich, the public will never allow it! You get a world war if necessary, but the public will always demand debasement, and there is no technical way to stop it (not even Bitcoin is immune, the government can easily subvert it when they are ready). Besides Bitcoin is even more harsh deflation than gold, as gold’s supply and nominal annual increase are both increasing. Bitcoin’s is geometrically declining.

    And this is the way it should be! A free-market does not mean that it is a merit to encourage non-productivity by erroneously assuming that BLIND capital can invest most efficiently in productivity. The larger the capital one has, the less knowledgeably one can allocate it. This is a fact. This is because innovation is born in the small, and no person is omniscient.

    Storing past productivity and holding the future in chains with your past accomplishments, is the antithesis of prosperity.

    Yes we need some savings, because this represents sacrifice and hard work, which is a merit based system.

    But we also need to debase the capital over time, so that the BLIND and LAZY rich don’t have an unfair advantage to make the innovators slaves to the lack omniscience.

    The philosophical paradigm of Bitcoin, is that the early adopters from first 4 years will have 50% of the money supply. That is insane and will never be allowed in a meritocracy. The free market will never allow this. Yes people should profit on their innovations, but they should not have 50% of all future production. Society will go to world war if necessary to remove those chains, but regulation and corporate-fascist takeover will probably suffice in this case (there isn’t always a distinction).

    Also all you goldbugs needs to read this:

    Does Money Have to be Backed?

  164. @Winter:

    I think that Bitcoin will fail. Not because it is gamed, or unsafe (both are true). But because it is based on the idea that you can start an unregulated currency in an industrial society. That will never work.

    My prior comments had an implied logic, which I want to make explicit. Also I have a new question.

    I think it is possible for Bitcoin to monopolize transaction processing globally, and for this to happen, the mega-corporations will monopolize the computation (“mining”) of the transaction blocks (giving it away for free) in harmony of course with government regulation. In my view, the mega-corporations and government are increasingly in bed together. I see a future where for example USA Homeland security (or the executive branch) blacklists a person’s global id, that person can’t transact. A sort of digital kill switch on life. For the moment, new Bitcoin addresses are anonymous, but I think you and I agree that if it is to become mainstream, then it will be regulated and anonymity will be restricted.

    Winter, I think I remember that you trust the government and regulators to always do the correct thing overall for the society. Presumably because if I remember correctly, you live in Denmark which I read Eric claim has very good governance. So am I correct to assume your view is that there is no threat, because even if Bitcoin becomes mainstream with the corporate monopoly I fear, then in your view citizens will be protected by the government regulation?

    That could explain why I feel more threatened than you ostensibly do.

    Again I will not debate whose opinion is correct on the merits of regulation. I just want to know what your opinion is.

  165. @JustSaying
    “Winter, I think I remember that you trust the government and regulators to always do the correct thing overall for the society.”

    No, I think governments and regulators are inevitable. We just have to make the best of it as the alternative has historically shown to be even worse (very much worse). But I do not think this is a good topic for discussion here.

    @JustSaying
    “Presumably because if I remember correctly, you live in Denmark which I read Eric claim has very good governance.”

    Close, its the Netherlands (380 miles of Copenhagen). “Very good” depends on taste and position, but I would not like to switch with the USA. Denmark seems to be even better though. But here even paying taxes is praised for its efficiency and (relative) ease.

    @JustSaying
    “So am I correct to assume your view is that there is no threat, because even if Bitcoin becomes mainstream with the corporate monopoly I fear, then in your view citizens will be protected by the government regulation?”

    Not quite. Citizens will force governments to protect them, one way or another. Which is rather different.

    Some populations are better at steering their administrations in the right direction than others, I admit. Some populations, as in “the people of …”, are historically bad at it (Russians come to mind).

  166. @Winter:
    I agree populations will demand governments protect them.

    The glass can never be entirely full nor entirely empty of governance, or quality of governance.

    This is really a point about relativity– there is no absolute point of reference. For example, the bitcoin fanboys staunchly defend the “no debasement” attribute of their currency. The price went vertical after the Mar. 18 FinCEN guideline which was perceived as legalizing Bitcoin. Thus proving the government can cause inflation or deflation in Bitcoin at any time that they wish. The value of money is based on confidence. The government can change the confidence at any time, because money is inherently always a socialized institution. The fanboys’ absolute is an illusion.

    Imagine a world of slavery where the rich got richer just for collecting interest on bonds and doing nothing with their mind? That is what the delusional goldbugs would get with no debasement. There is a distinction between preventing excessive inflation (probably good), preventing insider control over and front-running deflation/inflation rate (probably good) and no debasement (certainly bad and unrealistic).

    Ponzi Satoshi carefully crafted the Bitcoin design and whitepaper pitch to match the irrational psychology.

    I still don’t know where this leaves me on the question of should I fear Bitcoin enough to feel compelled to code a better replacement. So far, these P2P currency markets appear to be fanatical and embarrassingly gross time wasters.

  167. @JustSaying
    “Imagine a world of slavery where the rich got richer just for collecting interest on bonds and doing nothing with their mind?”

    I see we agree. If you want to see such a world, just look up RIAA.

  168. @JustSaying
    “I still don’t know where this leaves me on the question of should I fear Bitcoin enough to feel compelled to code a better replacement. ”

    I think it is too late for that. For good or bad, we are stuck with Bitcoin. If Bitcoin fails, it will take a generation for another such experiment will be possible.

  169. @JustSaying
    Btw, did you read “Making Money” by Terry Pratchett?

    I dismisses Bitcoin as much as that Youtube video you link to.

  170. @JustSaying
    “Bitcoin: The Digital Kill Switch”

    A possible definition of power is being able to do a lot of work, i.e., political power ~ electrical power. Thus every “proof of work” system will be vulnerable to concentrations of power.

    What is new?

    What this “Kill Switch” piece re-discovers is that a currency is a social system that is influenced by social structure. The funny thing is that the author is warning against a free market in signing bitcoin transactions.

    And as for gold, that was abolished as a standard for good reasons. It is expensive, cumbersome, and limits economic development. It is also strongly deflationary.

  171. @Winter:
    I don’t think the author is arguing against a free market in P2P currencies, actually the opposite. He was proposing to create a competitor to Bitcoin to fix a flaw that encourages cartelization of the processing. I assert this, because I am the author.

    The article has now been peer reviewed by the discussion there, so it is not true that any P2P can be as easily cartelized as another. Economics of the mining matter. The detail is that 6 billion people can bring a lot of hard-disk power that they already own. That was the key innovation, and the fix to not stop paying miners from debasement.

    Yeah agreed that gold is not suitable as a unit-of-exchange. It is suitable as a store-of-value when there is fear of collapse of government, finance, and chaos.

  172. @Winter:
    I see you are saying arguing against a free market in “signing transactions”, the proposed fix is to remove transaction fees entirely and continue the reward from perpetual debasement instead. This doesn’t remove a free-market, because the miners still compete to obtain that reward. It closes a loop-hole that would have in theory allowed a cartel to use transfer-pricing to bankrupt all competing miners by selling mining below cost at 0.

  173. @JustSaying

    If there exist a free market that is immune to cartels, that would indeed be interesting. We will see.

Leave a Reply to Nigel Cancel reply

Your email address will not be published. Required fields are marked *