The Smartphone Wars: HTC rejoins the good guys

Market pressure works. In response to an outcry from customers, HTC has just annnounced that it will return to its policy of shipping handsets with unlocked bootloaders.

This makes me personally happy because I’ve been a fan of HTC handsets since the G-1; I would have had to stop buying and recommending them if they’d stuck to the lockdown. But the larger reason this story is interesting is because of what it signals about ongoing shifts of power in the Android smartphone market – and throughout consumer electronics, as well.

There are four major stakeholders in the Android smartphone market: the cell carriers, the consumers, the handset makers, and Google. The carriers want smartphones to be locked-down carrier-controlled devices through which customers obediently buy services on carrier terms. All the other stakeholders, by contrast, gain from unlocked devices. Consumers win because they’re not limited to what the carriers choose to provide them; handset makers win because an unlocked phone is more valuable to a consumer than a locked one (and a little less expensive to ship, too); Google wins because its long-term strategy requires it to commoditize the carriers into a passive channel between the customers and Google.

Ever since the G-1 shipped in 2008 control of smartphone handsets has been slipping out of the carriers’ hands, a process well-chronicled on this blog. Google has been fighting its corner (and, on this and most other issues, the consumer’s) by taking control of the Android feature list, most notably via the Froyo announcement in May 2010. The carriers have been fighting back by pressuring the handset makers to ship phones with locked bootloaders. (The carriers’ biggest success on this front is undoubtedly the Motorola Droid 2.) But for a couple of years after the G-1 HTC led the way in resisting this pressure.

So, why did it flip? And why has it now flopped?

The reasons for HTC’s original no-lockdown policy aren’t hard to deduce. As a new entrant in the handset market back in 2008, HTC didn’t want to subtract value from its phones by locking them down – especially when many of its sales were to technically savvy early adopters who explicitly valued the openness of the phone. HTC’s carrier partner T-Mobile, as the #4 in the U.S. market, was polishing a valuable reputation as the carrier that doesn’t screw you in hopes of gaining market share.

As HTC gained visibility and began to woo other carriers, the pressure mounted for it to lock its new models. And I think I know what the carriers’ clinching argument was: movies and streaming video. I would bet serious money that the pitch went like this: the MPAA and the content cartel behind it demands DRM the consumers can’t subvert, and if you don’t give us boot-locked phones to build that on you’ll be out in the cold as your competitors sell locked phones that can stream movies. And HTC flipped.

The fact that HTC has flopped back tells us something very interesting: consumers, just by complaining loudly, were able to exert more pressure on HTC than the carriers and the content cartel. It’s also possible that Google exerted some pressure of its own; the point is, the pro-consumer anti-lockdown coalition won and the carrier/content-cartel alliance lost.

Power shifts like this are self-reinforcing. It’ll be that much more difficult, now, for the carriers to keep Motorola and Samsung and LG and Sony-Ericsson on the reservation – not that anyone but Motorola had been showing any real enthusiasm for locked bootloaders to begin with. The carriers have a weak hand, and it’s getting weaker. Which was, of course, Google’s plan all along.

More interesting, perhaps, is that the strong suggestion that the content cartel (the movie studios and record companies, represented by the MPAA and RIAA) may be losing the Svengali-like ability they used to have to make the consumer-electronics industry cripple its own products and screw consumers. There have been other signs of this recently, notably in the ignominious speed with which Netflix grabbed for a deal with Google after Google announced a movie-streaming tie-up with Amazon. I had been saying for months that this was coming, that Google could encourage unlocked phones and pay only the most perfunctory lip service to DRM in the knowledge that time and market share trends are on Android’s side. HTC’s reversal tends to confirm this.

And of course the MPAA/RIAA is quite right to fear unlocked hardware in the hands of Linux hackers. Netflix tried to keep the cartel happy by refusing to ship a client for Linux desktop machines, but it is almost as exposed now by shipping a client for Android. When the screens on Android tablets reach desktop quality, it will be fully as exposed. And its planners have to know that – but Netflix just can’t leave all that money on the table. Circumstances are driving a wedge between the content cartel and the streaming services, who at the end of the day like DRM only because it appeases the cartel. The streaming services make money by shipping bits, not by locking them up.

Indirectly HTC’s flop is also bad news for Apple. One of the principal advantages of the iPhone and other iOS devices was Apple’s special relationship with the content cartel. The exclusivity – and thus the market value – of that relationship has begun to erode. As the MPAA/RIAA axis weakens, so does Apple’s value proposition. Eventually this is going to put pressure on Apple’s prices and margins.

So, celebrate HTC’s decision. It doesn’t just mean that geeks get to heart HTC again; it means that the entire coalition of villains behind locked hardware and DRM is losing its grip. That’s good news for everyone.

112 comments

  1. I think that consumer electronics manufacturers have not quickly forgotten the lesson of the DAT. It got killed by the unusability that was forced on it by the record companies. Far, far more value was lost by the CES crowd than was gained by the RIAA crowd.

  2. Ever since the G-1 shipped in 2008 control of smartphone handsets has been slipping out of the carriers’ hands

    I’d say “since the first iPhone shipped in 2007”: Apple got concessions from AT&T that no handset maker ever had before: no crapware, etc.

  3. Google has already disabled movie rentals through Android Market to rooted devices “due to requirements related to copyright protection”. Protect the data path or go back to the farm club. If Google is waging war on the MPAA, it’s a fairly weak and clandestine one, as they are publicly very conciliatory to the same.

    The most likely scenario I can imagine is some sort of limited, officially-sanctioned reflashing capability that irrevocably clobbers whatever DRM keys are tied to the device. This enables consumers to root their device XOR play protected content on it. This is precisely what Archos uses for its internet tablets.

    In addition to this my guess is that carriers, the MPAA, etc. are going to become more aggressive about taking your toys away if you do non-business-model-approved things with them. For example, you use a rooted phone to tether without spending the extra $30/month on an approved tethering plan — and suddenly find that your phone can no longer connect to the 3G network or make phone calls. You use a hack you found on XDA-Developers to run Netflix on your non-Netflix-approved phone, and suddenly your Netflix account is frozen. Stuff like that. Game console manufacturers have set the precedent for this — first Sony with Other OS and now Nintendo, which reserves the right to permanently disable your 3DS if they find anything unauthorized running on or connected to it. Smartphones are a lot more like game consoles than they are like PCs.

  4. The fact that HTC has flopped back tells us something very interesting: consumers, just by complaining loudly, were able to exert more pressure on HTC than the carriers and the MPAA. It’s also possible that Google exerted some pressure of its own; the point is, the pro-consumer anti-lockdown coalition won and the carrier/MPAA alliance lost.

    I don’t think we can entirely say this. I think instead this is a “pass the buck” strategy. Given your proposition (which is sensible) that discussions with $(theMan) demanding locked bootloader was the reason for closing (which i think is certainly plausible), and the obvious proximity of their decision to Netflix saying “here’s an android app only for non-rooted phones”, they can now say to whoever “well locking down the data path isn’t my problem anymore”. If Netflix can do it, why can’t you?

    I wonder if screwing with the logic that deactivates an app if the phone is rooted would count as a DMCA violation especially when theres precedent that rooting a phone does not count as a DMCA violation.

  5. I think the lock down has moved, as Jeff says. Yes, you can root your device, but then you can’t play movies on it. It’s not a solution I like because it’s so asymmetrical.

    In order for movie studios to keep high quality copies off the net, they need to protect every single path from the most determined crackers around. Any leak at any point means that the whole cat is out of the bag. It’s very hard to protect anything once you go down the route of a legitimate windows copy running in a VM which is completely controlled by cracker.

    In order to plug all these leaks, MPAA et al are willing to crap on *every single legitimate customer* to try to prevent something that’s impossible to prevent. They used to try to win by not playing – not to distribute digital copies, but that alone just creates a vacuum for people to fill with their dodgy screen grabs and studio leaked tapes.

    I want the big media industries to understand that the only way to win this is to saturate the market with legitimate, easy ways to consume on our terms at reasonable prices. Create the value and I will follow.

    Why would I ever touch limewire when every song I ever want is availble on spotify the moment I want to hear it?
    Why would I bother ferreting through scams and crap copies around pirate bay if a movie I want to see is available instantly from lovefilm?

    I gladly pay monthly subscriptions to both these services. They aren’t as good as I want, where they fall down, other methods must fill the gap.

    There is a price and availability point when it becomes too inefficient to bother with the non-official versions. If I want to root my device, there’s a very small chance that it’s because I want to copy a movie, and a good chance is that I want take a screenshot of my widget setup, or play with a faster rom.

    Make it too hard, and I’ll be hitting the pirate bay because I, frankly, have no alternative anymore – they’re not meeting me where I am.

    1. >I think the lock down has moved, as Jeff says. Yes, you can root your device, but then you can’t play movies on it.

      True, but this is a fundamentally more difficult place to try to draw a line in the sand than locking out customized OS images. Without that lockdown, protecting the secrecy of the DRM keys is very problematic – one trick a modder can play, for example, is to load an image containing a VM that intercepts traps to the hardware. With that weapon in hand, patches to restore the ability to play movies on rooted devices will not be long in coming.

      Similar virtualization tricks can be used to reverse-engineer capability monitoring, and custom images then built to spoof it. Capability monitoring isn’t as easy as people think, anyway. For example, the carriers can’t tell even in principle when you’re tethering, not even with deep packet inspection; they have to guess at it by doing traffic analysis and then hope you’ll admit to your ‘crime’.

      Attempts to treat phones like game consoles would run into serious legal and regulatory problems, even if we left aside market pressure. A whole line of precedent running from the Carterfone decision to the openness rules Google triggered in the Block C auction stands squarely athwart such attempts. Trying to apply the DMCA would be tricky and the attempt would risk blowing up in the content cartel’s face.

      Just look at how the ground has shifted. It used to be that there was no effective counter to the alliance between the Big Media content cartel and the cell carriers. Now, we’re talking as if the cell carriers’ agenda doesn’t matter anymore for the excellent reason that it doesn’t; they’re fighting a doomed rear-guard action against Google, and HTC’s move shows that the content cartel has effectively lost its enforcer(s).

  6. I think Google is in a direct competition with the RIAA/MPAA here too.

    Those behind the RIAA/MPAA are earning their money by keeping artists/producers away from their audience. They are pure middle-men who control distribution, cream off all profits, but provide little else. In a process of disintermediation, artists and their audience try to get into direct contact, removing hte middle-men.

    Apple played along with the studios in iTunes, but is now in a position that they can control digital distribution and become the “most consolidated part” of the distribution channel an get the profit (the studios hate iTunes, always have).

    Google would profit from being an intermediate between producers and consumer just as they do with other bits (eg, YouTube). That is, Google would benefit from a direct channel between producers and consumer. That would allow Google to supply the market place for producers to sell (or distribute) their bits. Just like eBay earns money from allowing people to sell stuff, Google could earn money, and lots of it, by allowing people to distribute bits.

    As Google would cut the backers of the MPAA/RIAA completely out of the loop, their interests clash head on.

  7. A whole line of precedent running from the Carterfone decision to the openness rules Google triggered in the Block C auction stands squarely athwart such attempts.

    Unfortunately, the FCC has not applied the Carterfone decision to wireless carriers beyond requiring carriers to unlock phones after the end of the contract. I don’t see this changing anytime real soon.

  8. I’m a bit surprised by the timing of this. With the merger of ATT and T-Mobile in the works, I would think they would want to lay low.

    1. >I’m a bit surprised by the timing of this. With the merger of ATT and T-Mobile in the works, I would think they would want to lay low.

      You make an interesting point. HTC must have felt some serious heat from the customer base, because it had to be obvious to them that AT&T wasn’t going to like this reversal even one little bit. It will cost AT&T a lot of goodwill to re-reverse – assuming there’s enough time for that to matter before the contract system collapses under a flood of $50 Android SOC phones.

      On the other hand, if AT&T and T-Mobile are being realistic about when that collapse likely to happen, that may explain why T-Mobile could do this without cheesing off AT&T. With ultra-cheap SOC phones set to ship in volume 3Q2011 I don’t see how the contract system is going to outlast 2012, and it may well collapse before the merger goes through. That would imply that AT&T had nothing to lose from this announcement.

  9. > For example, the carriers can’t tell even in principle when you’re tethering, not even with deep packet inspection; they have to guess at it by doing traffic analysis
    > and then hope you’ll admit to your ‘crime’.

    Eric, there’s been research in at least the past 10 years which allows an OS to be fingerprinted based off of the TCP Initial Sequence Numbers alone.
    ISN randomness research. Likewise, there is the Passive OS Fingerprinting Tool. All that a carrier needs to do is prove that you have a connection which doesn’t come from the phone itself. A different ISN pattern, etc, would quite likely be sufficient. You can get around this by having your phone act as a proxy instead of a NAT forwarding gateway (such as with a SOCKS proxy), but that requires additional configuration which most users aren’t going to be willing to do.

    1. >Eric, there’s been research in at least the past 10 years which allows an OS to be fingerprinted based off of the TCP Initial Sequence Numbers alone.

      Shhh! Don’t say that where the carriers can hear! They’re not doing it yet.

      Once they start, as you say, we’ll have to start proxying. There’ll be an app for that in, oh, about six milliseconds once it’s needed.

  10. Shhh! Don’t say that where the carriers can hear! They’re not doing it yet.

    nmap does the fingerprinting by TCP Initial Sequence Numbers, but it doesn’t work very well. Especially if there isn’t at least one open and one closed port. No need to proxy, just use iptables to drop all incoming port connections from the carrier.

  11. A VPN might be cleaner than a proxy, and could probably be camouflaged if necessary, but it would require external servers…

  12. > There’ll be an app for that in, oh, about six milliseconds once it’s needed.

    I think that’s harder than you think. Either you need to do complete stateful packet re-writing on the phone, or you need to have the tethered machine be able to route all outgoing network connections through something like a SOCKS proxy. The first would require changing the packet forwarding mechanisms in the Linux kernel. Not impossible (it already needs to maintain state to map connection to NAT IP address), but I think that will hit another pile of application bugs as I’m certain *something* out there gets the ISN and uses it to horribly stupid ends. Still, probably the easiest long-term solution, but it probably can’t be dropped in as an app.

    Getting a client machine to transparently wrap all outgoing TCP connections over a SOCKS proxy is also possible, but would require extra software and configuration. A lot of individual applications (like web browsers) support this, and there are applications such as runsocks which allow you to use any arbitrary application, but I’m not sure that there is a nice way to make the whole OS do that transparently. I suppose that if there is a demand, it will appear.

  13. @Morgan:

    No need to proxy, just use iptables to drop all incoming port connections from the carrier.

    Yeah, but in this game of cat and mouse, remember that the carrier is your ISP. They can connect from whatever IP address they would care to spoof.

  14. Yeah, but in this game of cat and mouse, remember that the carrier is your ISP. They can connect from whatever IP address they would care to spoof.

    So just drop all incoming TCP connections whatsoever. I can’t imagine why you’d ever want to run a public TCP server on a 3G connection, and if you want to run one for your personal use (say, so you can ssh in), then put it behind OpenVPN running in UDP mode with –tls-auth enabled. That effectively makes the server undetectable to anyone who doesn’t have the HMAC key.

  15. I apologize for being a little confused by the terminology.

    Since HTC won’t require you to unlock future phones, does this mean that you can, for example, install a custom ROM without using some kind of “rooting” escalation hack?

    Would this then mean that you could modify your HTC phone while still having access to Google’s locked down media services (music, movies) and Netflix, or when they say “root” do they mean any modification that requires root, rather than the process of gaining root itself?

    1. >Since HTC won’t require you to unlock future phones, does this mean that you can, for example, install a custom ROM without using some kind of “rooting” escalation hack?

      That’s right.

      >Would this then mean that you could modify your HTC phone while still having access to Google’s locked down media services (music, movies) and Netflix, or when they say “root” do they mean any modification that requires root, rather than the process of gaining root itself?

      The clause after the “or” confuses me. I don’t think we know the practical answer to the first half of your question yet. In theory, yes, this should definitely be possible; how difficult it will be depends on how hard Google and Netflix have worked on concealing the service keys. It won’t be easy for them to have done an effective job, considering that the source code for Android is available.

  16. They can make it up to us by releasing their hardware API docs so Cyanogenmod can finally support things like phone diagnostic menus (Really missing ##DATA# for diagnostics purposes here).

  17. and if you want to run one for your personal use (say, so you can ssh in), then put it behind OpenVPN running in UDP mode with –tls-auth enabled. That effectively makes the server undetectable to anyone who doesn’t have the HMAC key.

    Oooh! I forgot you could do that! Yeah, dropping all incoming TCP port connections would definitely work, in that case. In fact, I wouldn’t be surprised if some Anrdoid devices already drop all incoming TCP port connections to enforce carrier policy.

  18. Microsoft gets $5 for every HTC phone running Android, so Microsoft has received five times more income from Android than from Windows Phone.

    Well, everyone who pirates Windows 7 Ultimate steals $319.99 from Microsoft for every device they install it on, so there! :-P

  19. Two things.

    First, WRT Motorola, I don’t think Verizon has to work hard to push Moto to lock it’s phones. My 2 years working closely with them in the wireless and radio business indicated to me that locking things up tight is par for the course, primarily because they are used to dealing with poorly trained users in critical situations and giving those users too much freedom could cost lives (e.g. military and first responders). Not to mention, it lets them hold the line on their ridiculously high margins.

    Second:
    >You make an interesting point. HTC must have felt some serious heat from the customer base, because it had to be obvious to them that AT&T wasn’t going to like this reversal even one little bit. It will cost AT&T a lot of goodwill to re-reverse…

    This makes me wonder, could it be that HTC made the announcement to help T-Mobile close the deal? AT&T’s premier phone (some Apple product I think :^) is locked up tight as a drum. Verizon’s top products are similarly stitched up. AT&T wants that control as you point out, but swallowing T-Mobile, who’s premium phone is as open as all outdoors probably didn’t look good to the boys on the top floor in Dallas. I’m wondering if T-Mobile didn’t “pressure” HTC to close things up to make that deal happen. If Deutsche Telekom felt it was important to get out of the US Market and get the cash into other investments, they likely would have done so, and HTC, knowing that DT was an important partner in Europe might have consented just to help out and get a little bargaining power in the long run. Now that it’s a done deal, HTC doesn’t have to play that game anymore, so they’ve “flopped”.

    Just a thought, we’ll likely never know.

    1. >Not to mention, it lets them hold the line on their ridiculously high margins.

      If that’s what Motorola thinks it’s doing, I think it would be unwise to put any long bets on them. HTC and the rest of the Asians know how to compete in a high-volume, low-unit-margin market and they are going to take the game to that level, no question. I wouldn’t be too surprised if one of them ends up owning Motorola, or anyway the assets from the fire sale.

  20. put it behind OpenVPN running in UDP mode with –tls-auth enabled. That effectively makes the server undetectable to anyone who doesn’t have the HMAC key.

    Well, I should qualify that slightly. If you’re a MITM, you can still detect its presence by sniffing connection handshakes. But I would expect carriers to find themselves in legal hot water if they attempted this.

  21. > If that’s what Motorola thinks it’s doing, I think it would be unwise to put any long bets on them. HTC and the rest of the Asians know how to compete in a high-volume, low-unit-margin market and they are going to take the game to that level, no question. I wouldn’t be too surprised if one of them ends up owning Motorola, or anyway the assets from the fire sale.

    We have seen that story before in the dumb phone market. It will be v2.0 of the exact same story. Mot was wiped out back then, and they were saying that the phone division was going to spun as a separate company (presumably so that it would be swallowed by some asian player – lenovo/ibm style).

  22. But I would expect carriers to find themselves in legal hot water if they attempted this.

    What legal hot water? Thanks in part to the Department of Homeland Security, the carriers sniff all sorts of shit. Oh, and good luck suing them: thanks to a recent Supreme Court decision they can put a clause in your contract that says you waive all rights to sue class-action and must submit to binding arbitration that has real legal force.

    So no, they will sniff, and get away with it, and you won’t get away with tethering.

  23. I find this interpretation reasonable. GPLv2 section 3 is satisfied if you

    b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange

    Clearly this implies that some amount of delay is acceptable, since it will take some time to process replies to such an offer. Since the GPL doesn’t specify a concrete timeframe anywhere, I’d instead infer a mandate to act in good faith: a request for the source code doesn’t mean that you have to immediately drop everything you’re doing in order to comply with it, but you also can’t deliberately stall in order to obtain a competitive advantage commercially. I can easily envision a corporate behemoth taking 120 days to push something like this through its own legal department while still acting in good faith. In HTC’s case, though, it actually only took them a week. The “90-120 days” seems to have been just an attempt control expectation levels for future releases.

  24. thanks to a recent Supreme Court decision they can put a clause in your contract that says you waive all rights to sue class-action and must submit to binding arbitration that has real legal force.

    Contrary to popular belief, Jeff Read, contracts aren’t a take-it-or-leave-it affair. You can take a pre-printed contract and, upon signing, simply cross out any section you do not like and initial and date it. If the company you have signed the contract with utters no objection, then the cross-out is legally binding.

  25. Meanwhile, Microsoft gets $5 for every HTC phone running Android, according to Citi analyst Walter Pritchard, who released a big report on Microsoft this morning. Microsoft is getting that money thanks to a patent settlement with HTC over intellectual property infringement.

    The net result: Microsoft has made five times more money from Android than from Windows Phone 7.

    Microsoft is suing other Android phone makers, and it’s looking for $7.50 to $12.50 per device, says Pritchard.

    link
    Nokia will be enriched, also

    Thanks for playing.

  26. Jeff, arbitration clauses are not new.

    The Supreme Court decision only affirmed that yes, if they’re in your contract, then yes, if you sign it, you agreed to be bound by it.

    We’re not yet at a point where you can go “I can’t believe the crap in this contract that I didn’t read! Obviously I can just ignore it.”

    [Not that I would be happy if I was the one being forced into arbitration. It would still suck.]

  27. I’m waiting to see Microsoft sue Cyanogen (Steve Kondik) and the other main developers, just for the LOLs.

  28. Eric, there’s been research in at least the past 10 years which allows an OS to be fingerprinted based off of the TCP Initial Sequence Numbers alone.

    What i’m seeing here is that anyone tethering Linux would probably be fine because linux(Android) should look like linux(Desktop). They didn’t test iOS so don’t know what its signature looks like compared to MacOS.

  29. @JonB:

    My point, btw, is that if google is acting within precedent when it says “not anytime soon” then it’s hard to argue that HTC isn’t when they say “give us up to 120 days”.

    What does google’s behavior have to do with HTC? Google obviously ships source to the phone vendors….

  30. “Those behind the RIAA/MPAA are earning their money by keeping artists/producers away from their audience. They are pure middle-men who control distribution, cream off all profits, but provide little else. In a process of disintermediation, artists and their audience try to get into direct contact, removing the middle-men.”

    It’s becoming easier for artists to distribute their products, but now it’s become much harder for the artists to collect money from their audience. People might talk a good game of not wanting to pay the fatcat record producers, but they will still copy songs for free, no matter who is putting them out there.

  31. LS is right.

    These days it is hard for a performer who isn’t a major-label hitmaker to sell more than a few hundred copies of their music.

    Now, granted, this has probably as much to do with licit distribution channels as illicit. If any schlub can sit in their basement, play an instrument, mix a few tracks together with GarageBand and pop it on the iTunes Music Store, that creates a lot more schlubs fighting for a lot smaller audience per schlub, and the power law kicks into play.

    That said, the rabid anti-DRMites have not come up with a workable solution to how to prevent piracy. Until that happens, DRM is here to stay, and as the technology becomes less intrusive, people will be happy paying a few bucks for DRM-locked content rather than attempt to pirate it or break the DRM lock on it. We’re almost at that tipping point now. Once we cross it, DRM will become an essential part of every computing device.

    Of course Linux (save for vendor-supplied Android) users will get left out in the cold for technical or legal reasons. (It is still a DMCA violation in the United States to play protected DVDs using freely available software tools.) C’est la vie.

  32. @JonB, linking to ESR’s defence of Google’s practices wrt. Honeycomb asks “And what’s your point?”

    Google’s only obligation to release source to the Honeycomb “private preview release” was to the select few it distributed the binaries to, which I think is the point Patrick was making. If it screwed them over this, I don’t honestly care that much, because it is between Google and them and they all have lawyers on their staff.

    HTC is witholding the code from people who get the binaries when they buy their devices. This is not OK.

    Is HTC violating its obligations in only a moral sense, or are they in a legal sense? Any commercial communications software/device maker who employs hackers to contribute code to such violated GPL software has the germ of a damages argument. The “precedent” doesn’t count for much, I think, since it has never been tested legally.

    Of course, if you think copyleft is pernicious, you’ll be happy with interpretations of the GPL that say 60 to 90 years is a reasonable grace period.

  33. You say there’s the “germ of a damages claim”.

    Well, if they do deliver in 120 days, then by the time it gets to a judge, the judge is going to say, “Okay, they’re in compliance now. So what’s the issue here, exactly? What did the delay cost you?”

    So, what do you answer the judge? What damages are you going to claim were caused by the delay? If you try to claim statutory infringement damages on the basis that the license requires immediate availability, you can be assured the judge will ask you why the license doesn’t actually specify immediate availability. If you’re going to claim that some delay is reasonable but 120 days isn’t, you’re going to be asked what makes 120 days unreasonable compared to 60 or 30 or 15 or whatever, and why there wasn’t a specific reasonable period put in the license. What exactly is your answer?

  34. @Jef Read
    “That said, the rabid anti-DRMites have not come up with a workable solution to how to prevent piracy.”

    But this assumes artists receive revenue from music sales. What I hear is that labels get all the revenues from music sales and artists get their money from concerts.

    So, for the artists DRM reduces their market of paying audience.

  35. @Steven Ehrbar:

    ey’re in compliance now. So what’s the issue here, exactly? What did the delay cost you?”

    So, what do you answer the judge? What damages are you going to claim were caused by the delay? If you try to claim statutory infringement damages on the basis that the license requires immediate availability, you can be assured the judge will ask you why the license doesn’t actually specify immediate availability. If you’re going to claim that some delay is reasonable but 120 days isn’t, you’re going to be asked what makes 120 days unreasonable compared to 60 or 30 or 15 or whatever, and why there wasn’t a specific reasonable period put in the license. What exactly is your answer?

    Bearing in mind that I’m not the one with my knickers in a twist over this, if I were outraged over this, then my answer would exactly be that (1) the internet being the internet, the only reasonable delay is no delay; and (2) it’s not about damages — it’s about an injunction to keep them from dragging their feet on this over and over.

    I would also try to shame the company in the court documents, which you know would be widely read among geeks. Any ISO 9000 compliant company needs to be able to (a) know exactly what source went into a build and be able to recreate the build with minor modifications if necessary; and (b) be able to set up a reproducible product release procedure that involves placing a tarfile on the internet.

    The direct examination of the company’s experts on how difficult this all is should be priceles.

  36. >That said, the rabid anti-DRMites have not come up with a workable solution to how to prevent piracy.

    Neither have the pro-DRMites. The entire idea is fatally flawed — to work, they need to successfully protect the entire data path, not just on the device of the would-be pirate, but on every device everywhere.

    Guess what? If I had a need to, I’d have the tools and the expertise to strip down a screen, and intercept the signal between the electronics that handle the DRM and the actual display element. *So long as there is a market, the content will always be available on TPB and the likes.* DRM does not in any way stop or mitigate this. It’s completely worthless as an anti-piracy device. It’s a scam on the content owners, perpetrated by those who take paychecks to “protect” their content.

    Once the content is legally available at a price and effort low enough to make people buy it, people will. DRM has absolutely nothing to do with this.

  37. Hello Eric,
    I want to write an article about the technology around White Spaces.
    Does it participate to the unused spectrum debate ?

    And seeing this article :
    (via TechReview) Microsoft Wants to Rule the White Spaces http://bit.ly/mR2He6

    Do you think Microsoft tries here to breach telcos monopoly like Google tries with Android ?
    Or are there reasons to detect an evil move here ?

    ESR says: I don’t know enough about this yet. I’ll have to do some research.

  38. @Thierry Lhôte
    “Do you think Microsoft tries here to breach telcos monopoly like Google tries with Android ? Or are there reasons to detect an evil move here ?”

    MS have a war chest. They will waste every cent in it trying to buy themselves a new (mobile) monopoly.

    Until their share holders see the money flowing down the drain. Then MS will pay out their money to the share holders. Or what will be left by then.

  39. Antti-Ville Tuunainen:

    DRM has done its job. It’s made piracy more diffcult and more overtly illegal; at the same time purchasing content has gotten vastly easier (in large part thanks to Apple). It takes an overt act to circumvent DRM, and that overt act goes a long way towards establishing mens rea on the part of the infringer.

    A lot of government security standards work the same way. Oftentimes it’s terribly inconvenient to build a system that makes it completely impossible to see something you’re not supposed to, so they build it with a lesser protection that you have to really try to break to see the secret stuff. Then they can nail you for trying.

  40. @Jeff Read:

    > DRM has done its job.

    This would be the case if the rate of non-DRM uploads to the net is less than the rate at which new high quality DRM media is being generated by the content/copyright owners. You have to compare the rate at which the two processes are happening to conclude who is/isn’t winning.

  41. It certainly appears that Baen Books has quietly done quite well by selling Ebooks with no DRM beyond a polite request that we don’t casually pass them around.

    Eric Flint started it, Jim Baen went along to see where it might go. Last I heard they *only* get about the same income from ebooks and Canada sales.
    Their theory — Most people prefer to be honest. If you offer your product for sale in a convenient manner and at a fair price, most people will pay, rather than steal. Most of those who steal wouldn’t have bought anyway.

    Another thing they do right. You can read the first quarter or third of a book online before you lay out any money. I imagine that almost all the people who reach the end of the sample buy the book. Barnes and Nobel frequently gives less than the first chapter. I think they are missing a serious opportunity.

    Works for me. I have purchased somewhere around 650-700 ebooks from Baen. Somewhere in the neighborhood of $3500 over 12 years. If they had tried to do some sort of obnoxious DRM the number would be much smaller.

    Bottom line — I can’t see where DRM is required to make a profit on content, and is probaby a net negative for profitable content sales.
    Control of content sales is a different matter.

  42. This would be the case if the rate of non-DRM uploads to the net is less than the rate at which new high quality DRM media is being generated by the content/copyright owners.

    One word: Netflix.

    Netflix streaming has far surpassed the pirate channels to become the single biggest bandwidth consumer in the United States.

    DRM’d out the wazoo. Probably will be till the day it expires. You have to get a plugin from Microsoft to view the damn streams.

  43. @Jef Read
    But even netflix is an example of DRM used to lock out new producers from the distribution channels.

    Both music and movies are produced in such massive quantities that they far outstrip demand (yuotube?).

    However, distribution channels are all throttled and choice limited to a few publishers.
    And the fiercest actions are against legal offerings that threaten the control (mp3.com and youtube).

  44. I would also try to shame the company in the court documents, which you know would be widely read among geeks. Any ISO 9000 compliant company needs to be able to (a) know exactly what source went into a build and be able to recreate the build with minor modifications if necessary; and (b) be able to set up a reproducible product release procedure that involves placing a tarfile on the internet.

    It would be nice if they added some transparency around why the delays but i highly doubt that the delays are related to either of these things. Even without ISO 9000 it’s inconceivable that there wouldn’t be a central build tree somewhere that they could package using tar in about 1 minute. Vastly more likely in my mind is that company policy is that “legal”(or compliance or HR or whatever that department might be called in HTC) have to run some checks before they release the code. The two things that spring to mind are “make sure there’s nothing we’ll regret in the comments” and “make sure we haven’t accidentally added something proprietary (HTC or other)”. The second one in particular jumps out because it accounts for the massive time allowance (120 days) and the minimal reality (7 days). My only reservation is it really should have been done BEFORE release.

    The main reason the attitude of “HTC is horrible, you bring the pitchforks i’ll bring the tar and feathers” exasperates me is that the one thing that parts of the Open Source movement have problems doing is calming the fuck down. HTC is our ally here and yet members of the crowd are basically throwing down on them.

    Does the delay suck? Yeah sure especially when having the code would have made breaking the bootloader easier. But the hyperbole and emotive content in some of the blog posts and posts here is just retarded and it doesn’t help hte cause at all. You can still heart HTC while wishing they got their ass in gear over source release. 2-4 months isn’t the end of the world.

    1. >the one thing that parts of the Open Source movement have problems doing is calming the fuck down.

      Tell it, brother. On those unusual occasions that I flex my open-source community-leader muscles, the most common reason is that people need to be told to calm the fuck down.

  45. @Jeff Read:

    One word: Netflix.

    Netflix streaming has far surpassed the pirate channels to become the single biggest bandwidth consumer in the United States.

    That is not content ‘generation’. That is content ‘distribution’.

    ‘Generation’ has to do with the fundamental ability to generate art/content at a rate faster than the rate at which non-DRM versions of yesterday’s or yesteryear’s art appear.

    A high quality (non-DRM) mp3 audio file lives on forever. Far longer than expected lifetime of Netflix. There are no theoretical limitations on the ability to copy, upload or download that material -> open nature of the internet. This is unlike say “physical media” like cassetes in the old days where copying, distribution etc are all constrained by the physical medium carrying the material and your ability to manufacture, distribute, market it.

  46. @Patrick Maupin:

    I’m not a lawyer, and I may well be improperly analyzing this in terms of my understanding of contracts from business law surveys rather than proper lawyer-type licensing law, but . . . I can’t where the judge’s answer will be any more pro-plaintiff than “Okay, at best, you’ve got a material breach here, not a fundamental breach. So you can get specific performance and damages. Since they already released the source code before the case got here, ordering specific performance is irrelevant. Since you can’t point to any damages, you can’t get any damages. So that leaves us only with the injunction request. Explain to me how the injury to your rights justifies an injunction to require them to act faster than within 120 days in future situations.”

    Which is to say, the plaintiff is not going to have the opportunity to embarrass HTC, because the case will probably not get far enough that HTC actually has to explain anything.

  47. @Steven Ehrbar:

    Did you mean “immaterial breach”?

    If the plaintiff asks HTC if they plan to release in a timely fashion in the future, and HTC’s weasel-words effectively say “no,” then I would think that the plaintiff could explain to the judge that this was an “anticipatory breach.”

    But IANAL either…

  48. @JonB:

    Agreed. Throwing a fit at HTC over the delay in the release of GPL’d source is counterproductive. Those doing so aren’t helping.

    Another further thing to consider is that the vast majority of the code that makes up Android isn’t GPL anyway: it’s Apache licensed or BSD licensed (same thing, really). Only the kernel and I think one or two other packages are GPL.

    Rumor on the street is that HTC is considering their own mobile OS. I’ll bet it wouldn’t take much for them to fork Android with a FreeBSD kernel.

  49. @Patrick Maupin:

    No, I didn’t mean immaterial. I was giving the absolutely most pro-plaintiff ruling I could see possible, and so postulating a ruling of material breach. A material breach authorizes a demand of specific performance; an immaterial breach does not.

  50. @Steve Ehrbar:

    I think licenses are not quite the same as contracts. Here’s an interesting exposition on some case law:

    http://www.ipmall.org/hosted_resources/IDEA/33_IDEA/33-2_IDEA_225_Jones.pdf

    If the plaintiff could show that the breach resulted in the unlicensed distribution of copies, then statutory damages may apply. We’ve already seen how far the RIAA can run with those, so I’m sure if you found the right judge, you could get a windfall.

  51. @Morgan:

    Rumor on the street is that HTC is considering their own mobile OS. I’ll bet it wouldn’t take much for them to fork Android
    with a FreeBSD kernel.

    It will be interesting to see the desire for differentiation play out against the desire for lowered development costs.

  52. Interesting article and comments. I have to admit that until I read this discussion, I was barely aware of DRM. Perhaps the perspective of a consumer would be helpful. After all, it’s the behavior of the casual consumer who’ll determine what technology will be adopted. Sorry for the length of the comment. Once I started, I couldn’t stop.

    Just a little background of what informs my perspective. I used to be more tech savvy than I am now. Unfortunately for the last decade I’ve been concentrating on other problems than keeping up with technology. :( I am generally able and willing to follow simple instructions to optimize the technology that I use so that it’s “good enough” for my needs. As long as technology works seamlessly, I can focus my attention on other things that have preoccupied me. I target my adoption of technology to be a little behind the curve. I avoided buying mp3 players, digital cameras, etc. because I’m not willing to pay a premium just to be an early adopter. My ideal is One Device to Rule Them All. My general price point for an electronic device is generally around $300-400. I’m consciously looking for a completely mobile lifestyle.

    I had heard of DRM but did not really know what it was. DRM enabled seemed to be a signal that I could enjoy media content that I was willing to pay for to either support the content provider or to avoid the risk of virus or malware. From a casual consumer perspective, my main goal is to be able to enjoy media in a convenient* way at a price point** that makes sense for me. Because of streaming services, e.g. Netflix or Hulu and single purchase services, e.g. kindle or itunes, I’ve moved beyond the tv, dvds, and cds. My decision to purchase a laptop will be affected by whether I can use these services. If I can get the same convenience using a tablet to access those services, then I’ll eventually stop using a laptop as my primary media device. If I still need a laptop, then my decision to migrate to Linux or to ChromeOS will be affected by how well it functions as a media device. So far I haven’t migrated away from Windows 7 because it’s worked well enough that I haven’t been socked in the face with the need to migrate. I’m willing to be convinced. I remember Linux as requiring more time than I had to learn and figure out. If I’m not able to use media I purchased on Linux, then that would be a deal breaker.

    Ironically, I read this post right after I went to a used bookstore to sell about 5 boxes of books that I’ve accumulated through the years. If I could have somehow have had access to electronic versions of physical books which I had already purchased, I would have gotten rid of the remaining two boxes. I’m a fairly mobile person so my ideal would be to have my media content in a cloud service so I don’t have to worry about losing content if I forget to make back ups or happen to lose them (which when moving or traveling a lot can be -very- easy to do).

    This post also explained why development of ROMS for my Epic 4G seemed so slow. When it comes time to replace my android, I’ll give consideration to manufacturers such as HTC where the development is not so painfully slow. Just for background, I moved to Android last fall after deciding to replace my Palm Centro (the charger died and it seemed time to move on).I’d seen a friend’s Android and it seemed to be a cheaper and more convenient option than an iphone. Plus I liked the fact that the ability to customize my own phone was built into the idea of the device as opposed to my perception of the iphone as a device that I couldn’t control as compared to the Android. So the the blocks for choosing iphone were: I’d have had to pay a heft penalty to break my contract and move to AT&T, it’s a more expensive phone especially after the carrier subsidy for the android device, my perception of iphone is that I would not have as much control right down to the kernel as I would with Android.

    To my surprise, once I rooted my phone, I found that I could use wifi tether to make it my primary internet option. Ironically I had bricked my phone recently in an attempt to go back to rooted stock ROM from a customized ROM that was too unstable for my tastes. I have been able to get by with my back up plan of Virgin Mobile’s no contract broadband service combined the ability to make free phone calls with google voice and chat. In the future, I’ll be looking at cell carriers more as internet providers than for anything else. I don’t really use the phone to make calls a lot. My android was mainly used for google maps, web browsing when I’m on the go, as an internet hub for my laptop, and as a digital camera which I can then use to upload pictures to share with my friends on facebook. I suspect that by the end of this year, I’ll be going to a phone + tablet model. Tablet only would be nice, but I’m not familiar enough with tablets to know if I can conveniently carry it on my person to places where a phone is good but a laptop is too much, e.g. when I go to a concert.

    *Convenient means that ideally I can share purchased content among my media devices. At this time my media devices are my laptop and my android. At this time a tv is an unnecessary inconvenience since I can fulfill my entertainment needs without an extra piece of immobile equipment. Heck event the laptop is starting to be an annoyance as a carry-able device given the options for smartphones and tablets.

    **a fair price point is different for every person. I was never someone who purchased hundreds of cds and dvds because that was too expensive for me. I would pay $10 for a cd at a band’s show or $10 for an album on itunes for an artist that I really enjoy, $7.99/month for Netflix (although I may not renew my subscription because Hulu’s free service meets my needs. Previously I had both Netflix and Hulu’s paid service. I’m not willing to pay $1.99 per episode for Doctor Who, even though I really enjoy the show. That’s the point where I would go to pirate bay if I could. Not sure what the price point would be to purchase a movie. It doesn’t make as much sense for me to do that if I know I can stream it as part of a subscription service. As for books, Kindle prices are generally too expensive for me to purchase books willy nilly. Frankly I was surprised that it cost the same to purchase a Kindle version as a printed version. I don’t have a Kindle, but the Kindle app came with my Android and is surprisingly convenient. I’d be very happy if there were a subscription service for ebooks somewhere.

  53. Sorry for the diarrhea of the mouth. I didn’t realize it was that long. That’s what I get for procrastinating before moving. :p

  54. Ehrbar, you are confusing contract law principles with the principles of copyright law and licensing. Here with copyright law and licensing, the analysis is one of infringement from copying outside of the license terms.

  55. “Bottom line — I can’t see where DRM is required to make a profit on content, and is probaby a net negative for profitable content sales.”

    It isn’t just a profit…providers want the maximum profit possible.

    Its not just DRM. One example that comes to my geriatric mind is TV cable systems back in the ’70s. The cable operators were already bringing all their channels into your house with their cable feed. TV manufacturers produced ‘cable ready’ sets so you could get some splitters and cable of your own and put several sets in your home. The cable operators threw a fit. Here they were not doing any extra work for you, yet they demanded that your rent additional ‘boxes’ from them that were completely unneeded. It was blatant and stupid, yet they wouldn’t back down. They felt it was your duty to line their pockets.

  56. @Shenpen:

    Not much. The $5 per phone agreement is, allegedly, to pay Microsoft for those still-undisclosed patents that Microsoft claims the Linux kernel violates. This agreement is similar to the Novell agreement, and HTC likely agreed to it to help shore up its defense against Apple.

    Until Microsoft actually sues someone with enough of a spine (read: patent portfolio) to fight them, Microsoft is going to continue its FUD invasion against Linux.

  57. @Morgan Greywolf & Shenpen
    “Until Microsoft actually sues someone with enough of a spine (read: patent portfolio) to fight them, Microsoft is going to continue its FUD invasion against Linux.”

    MS just confirms claims that it is a parasitic company. MS do not make products people want to buy and resorts to extorting other companies that do make products that sell.

    Read the response of Barnes and Nobles when sued by MS over those patents:

    Barnes & Noble Charges Microsoft with Misusing Patents to Further an Anticompetive Scheme Against Android
    http://www.groklaw.net/article.php?story=20110427052238659&query=barnes+and+noble

    …Microsoft has asserted patents that extend only to arbitrary, outmoded, or non-essential design features, but uses these patents to demand that every manufacturer of an Android-based mobile device take a license from Microsoft and pay exorbitant licensing fees or face protracted and expensive patent infringement litigation. The asserted patents do not have a lawful scope sufficient to control the AndroidTM Operating System as Microsoft is attempting to do, and Microsoft’s misuse of these patents directly harms both competition for and consumers of all eReaders, smartphones, tablet computers and other mobile electronic devices….

    Barnes & Noble: Microsoft wanted ‘shockingly high’ fees for Android in Nook
    http://www.geekwire.com/2011/barnes-noble-microsoft-shockingly-high-fees-android-nook

    Insisting that an NDA was necessary, Microsoft discussed the alleged infringement on a high level basis only. Microsoft nevertheless maintained that it possessed patents sufficient to dominate and entirely preclude the use of the Android Operating System by the Nook. Microsoft demanded an exorbitant royalty (on a per device basis) for a license to its patent portfolio for the Nook device and at the end of the meeting Microsoft stated that it would demand an even higher per device royalty for any device that acted “more like a computer” as opposed to an eReader. …

  58. Until Microsoft actually sues someone with enough of a spine (read: patent portfolio) to fight them, Microsoft is going to continue its FUD invasion against Linux.

    It wouldn’t surprise me if one of a) HTC engineers cut a few corners in implementing some of the (closed i’d assume) bits on their phones which walked on MS Patent technology that they knew about because of the ongoing WM work or b) their lawyers suggested that it’d be a chancy fight to prove that the above wasn’t the case. Hence why the settlement.

  59. @Steven: > understanding of contracts

    Obviously, contract law is inapplicable since there is no contract in this case; as Patrick and SPQR observe, licenses are not contracts because there is no agreement. The violation is that every sale of HTC’s phone is a copyright violation by HTC of the GPLed software. Remedies under US contract law are injunctions and damages, both usually sought together.

    The case is tricky, which is why I say “germ of a damages argument”, but consider such a possible argument: If a group of telecommunications companies combine their code in this way, they are a little like such consortia as the H.264, one informally knitted out of copyright law rather than formally through a consortium agreement. The damages are that HTC takes new IP from other group contributors, and shares its IP when it is less fresh. This injures the other group members competitiveness, in just the same way that a consortium member would if it violated mutual openness clauses.

    I gather that when courts cannot determine actual damages, they generally fall back on rather steep statutory damages. So the plaintiff need only defeat the suggestion that there are no damages to have a serious case.

    How steep? Well the Copyright Act specifies a sum of between $750 and $150 000 for wilful infringement: http://chnm.gmu.edu/digitalhistory/links/cached/chapter7/link7.24b.statutorycopyrightdamages.html

    If I think HTC’s chances of suffering more than an injunction are not high, defending this case is a bit like running through a minefield with slippers on. I should think they would settle out of court.

  60. @Joe Presley:

    I don’t think any apologies are necessary. That was a well-written factual anecdote, which is sometimes hard to come by.

  61. So the plaintiff need only defeat the suggestion that there are no damages to have a serious case.

    No. The plaintiff should never make the suggestion that there no actual damages, only the suggestion that actual damages would be nearly impossible to determine.

    Copyright infringement is a tort. In a tort, legal damages are determined by law. In the case of copyright law, statutory damages apply to all copyrights that are registered with USPTO; if not registered, the plaintiff may not be entitled to statutory damages, but he or she is still entitled to sue for actual damages.

  62. Meanwhile, the Nokia implosion continues to snowball:

    Zdnet summary of quarterly conference call

    They have inventory piling up in Asia, and cheap Android phones are killing margins in Europe; they’ll be lucky to break even this quarter, and are withdrawing their estimates (!) for the rest of the year.

    They still expect to ship a Win7 phone late this year – at that rate, they’d better, and it had better sell.

  63. A rather fact-starved piece. But we know how MS works. If you want to ship MS Windows, you cannot use good hardware. Your hardware must be crippled not to compete with more expensive offerings of MS.

    Microsoft Hardware Rules ‘Troublesome’ for Makers, Acer Says
    http://www.bloomberg.com/news/2011-05-31/microsoft-hardware-rules-troublesome-for-makers-acer-says.html

    Microsoft Corp. (MSFT) is putting “troublesome” restrictions on makers of processors used to run the coming Windows tablet-computer operating system, Acer Inc. (2353) Chairman and Chief Executive Officer J.T. Wang said.

    “They’re really controlling the whole thing, the whole process,” Wang said at the Computex trade show in Taipei without identifying the restrictions. Chip suppliers and PC makers “all feel it’s very troublesome,” he said.

    1. >The introduction of the Verizon iPhone has stopped Android’s growth:

      But Nielsen itself says these stories are bogus, derived from comparing two surveys by different companies over one month and three months including March. You’re grasping at straws, fanboy.

  64. > But Nielsen itself says these stories are bogus, derived from comparing two surveys by different companies over one month and three months including March.

    [citation needed]

    ESR says: here it is.

  65. I don’t have time to analyze the data right now, but one thing I took away from one of the articles about it is the possible bifurcation of the Android space:

    Although iPhone owners are more likely than Android aficionados to download apps, stream music or watch videos, individual Android users who do that kind of stuff tend to gobble up significantly more data.

    The accompanying charts could be the best proxy evidence to date to support the theory that Android will gain both the most sophisticated users and the most cost-sensitive users…

  66. @Life As We Know It:

    The introduction of the Verizon iPhone has stopped Android’s growth

    I had a chance to glance at the raw data. First, the March data, then the April data.

    Even the most ignorant, bogus, and cursory analysis of the data doesn’t support your iPhone 4V theory. Instead, it supports the somewhat more ludicrous proposition that the launch of the RIM Playbook stopped both google and Apple dead in their tracks.

  67. Somewhat offtopic, the manufacturer of steam powered carriages may be bought by the buggy whip maker:

    Microsoft (MSFT 24.54, -0.47) is back in the headlines after technology blog Boy Genius reported the tech giant was set to purchase Nokia’s (NOK 6.68, -0.34) phone business for $19 billion. The report has since been denied by a spokesman for Nokia, calling it “100% baseless.” Regardless, both stocks are seeing heavy volume on the session.

  68. @SPQR:

    MSFT buying NOK (at today’s valuation, $25B) makes a certain amount of sense. Microsoft has a bunch of cash overseas that would be best spent on overseas acquisitions, rather than repatriating it and paying huge taxes. That’s one reason it paid for Skype, it did so with cash it had already sitting in Europe.

    It’s also the only way that MSFT would ever make huge piles of money off Windows Mobile.
    http://www.businessinsider.com/microsoft-nokia-windows-phone-2011-5

    And now that they own Skype outright, they could force a Skype phone. It would be a huge bet, and assure MSFT of a market for WM7 phones.

    Don’t forget, if they can find a way around the patent exhaustion issues (having licensed to HTC, Samsung, etc.) they could turn around and sue the carriers for infringement, making Android phones a helluva lot more expensive in the process.

    1. >And now that they own Skype outright, [Microsoft] could force a Skype phone.

      The cell carriers would revolt at this. They hate Skype – because, though they may recognize Android as a long-term threat to commoditize them, they see Skype as a prompt and immediate threat to commoditize them. They can be seduced by the lure of lower NRE and time-to-market into fooling themselves that they can tame and harness Android, but they can’t fool themselves about Skype.

      If Microsoft were to try this, not only would they get no takeup or co-marketing from the carriers, the carriers would probably increase their investment in Android as a countermove. So, er, maybe I should hope Balllmer is stupid enough to try it.

  69. @esr> The cell carriers would revolt at this. They hate Skype – because, though they may recognize Android as a long-term threat to commoditize them, they see Skype as a prompt and immediate threat to commoditize them.

    OK, Eric.

    Let’s all ignore the elephant in the room named “Google Voice”.

    1. >Let’s all ignore the elephant in the room named “Google Voice”.

      No, let’s not. Google did a clever thing; for incoming calls, users must have an established US telephone service to activate a Google Voice phone number. This isn’t technically necessary and Skype doesn’t require it – it was a way of co-opting the carriers.

  70. Here’s the difference between the single-vendor approach (Apple) and the multi-vendor, loose consortium clusterfuck approach (Android):

    There is no reliable way to reboot a PC.

    Meanwhile, all Macs have had software-mediated reboot capability that worked the first time, every time, since 1984 to the present day.

  71. I don’t think it does either, but as an off topic anecdote, my boss was absolutely horrified the first time he saw me tell my linux work station to reboot without first manually closing every application. Being primarily a mac user, (and to my experience, this is mostly true for windows) I’m used to being able to tell the computer to restart and having the OS ask all the applications to quit nicely, and only pausing for ones that require my intervention (would you like to save?). Apparently this isn’t true for linux in all cases.

    Eric, any thoughts on why this is so? Having not “been there” at the time, I can only assume there is some good* historical reason for this, and it’s never bothered anyone enough to change it.

    1. >Eric, any thoughts on why this is so? Having not “been there” at the time, I can only assume there is some good* historical reason for this, and it’s never bothered anyone enough to change it.

      Older versions of Unix could get very confused if you hard-rebooted without first muttering an incantation to flush certain in-memory data structures to disk. This is still theoretically possible with modern journaling file systems but very, very rare.

  72. > Google did a clever thing; for incoming calls, users must have an established US telephone service to activate a Google Voice phone number.

    Oh, really?
    http://slashdot.org/story/11/03/08/1955200/Google-Voice-Discovered-Allowing-Pure-VoIP-Calls
    http://www.androidpolice.com/2011/03/07/goodbye-voice-minutes-first-signs-of-google-voice-voip-emerge/

    and this just in: Motorala signals defection from open Android app store process:
    http://www.pcworld.com/article/229251/motorola_ceo_open_android_store_leads_to_quality_issues.html

    (tee hee)

    1. >and this just in: Motorala signals defection from open Android app store process:

      You are trolling. That article neither says nor implies this. You have just wasted my time; repeating this can get you banned.

  73. That makes sense, but it wasn’t quite what I was talking about. I mean using the shutdown / restart commands from within Gnome, KDE or xfce, not simply hitting the reset button.

    1. >That makes sense, but it wasn’t quite what I was talking about. I mean using the shutdown / restart commands from within Gnome, KDE or xfce, not simply hitting the reset button.

      I don’t understand your question, then. Obviously, a software-initiated shutdown can do whatever is necessary to ensure a clean disk state before triggering the hardware shutdown.

  74. > This is still theoretically possible with modern journaling file systems but very, very rare.

    It’s actually got little to do with journaling, and much more to do with ensuring that any given write to the disk leaves the filesystem in a consistent (or at least recoverable) state.

    But you didn’t really answer his question. MacOS (X) has a somewhat more sophisticated mechanism for notifying applications of things like a pending halt than Unix’s signal(2). Thus, on a Mac, an application can hold-up (or even cancel) a restart, and apps on MacOS have been written to advantage themselves of this. Unix apps… don’t / can’t.

    The difference is one of focus. Unix is a server OS, though some attempt to make it a desktop. MacOS (and Windows) are desktop-focused, and tend to fail when used for any but trivial server use.

    True, Linux is the core of Android, but there is a thick layer of non-unix code (written in Java) between Android apps and the kernel. Android apps never see so much as a libc interface.

    (it’s funny, but iOS apps are q lot closer to running on unix than are Android apps. Google’s Chrome OS is similar in that Chrome apps never see libc, either.)

  75. Sorry…Posted in haste.
    I will not be offended if you remove both my comments.

    Both URLs sometimes work, but I do get an intermittent:
    Error 503 Service Unavailable

    Service Unavailable
    Guru Meditation:

    XID: 540485857

    Varnish cache server

    1. >Looks like Samsung has seen the light as well:

      The economics of this is interesting. I’ll probably blog about it later today.

  76. Yes, and? They sank a certain amount of NRE into that, apparently. Particularly since Aunt May and Uncle Ben bought the thing assuming Sense would be on it – if an upgrade strips it, that would be a problem.

    What I found more interesting in Android news this week was that Virgin Mobile evidently made Moto strip BLUR out of VM’s latest Moto Triumph unit… (http://www.pcmag.com/article2/0,2817,2386729,00.asp was the first article I could lay my lunch hooks on).

    (Incidentally, unless Gingerbread comes with the contact-merging capabilities that Sense on my TDR-4G does and my Transformer does not, make mine Sense.)

    1. >Particularly since Aunt May and Uncle Ben bought the thing assuming Sense would be on it – if an upgrade strips it, that would be a problem.

      A phone with (say) security bugs that never get patched out could be a much worse problem. Probably is, since carrier skins are usually a branding exercise that subtracts value from the user’s point of view. I don’t think your argument holds a lot of water.

      But not(Andy Rubin)’s hyperventilating isn’t justified either, because the Desire is fully supported by CyanogenMod. People who want it will have an upgrade path. HTC would have been smart to include a pointer to CyanogenMod in their announcment – implying, if you think about it, let another reason boot-locking is not really in handset makers’ interests.

  77. Careful about generalizing users when talking about hardware vendor skins. I am a fairly technical user, and I like a lot of the sense UI, particularly in the contact management arena. I know I am giving up rapid access to OS updates, but it is a trade off, not a “always negative”.

Leave a comment

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