And now, for a bit of comic relief, let us examine the state of Windows Phone 7. It launched just over three months ago in a cloud of hubris – Microsoft’s first-ship party featured pallbearers lugging a huge mockup of an iPhone (and the rest of what you need to know about that moment was that the video was caught by an Android phone). Just how is Redmond’s bid to escape irrelevance doing?
Not well enough for Microsoft to want to disclose sales figures, apparently. While Google reports 200K Android phones a day are shipping, Microsoft is reduced to gamely insisting that it is confident Windows 7 will eventually succeed. At least one of its channel partners is, shall we say, less sanguine.
James Choi, a strategist at handset maker LG, says “From an industry perspective we had a high expectation, but from a consumer point of view the visibility is less than we expected”. Choi holds out some hope that WP7 will make better sales as it is ported to less expensive handsets, but that hope itself suggests the software is hobbled by bloat and poor performance.
Developers aren’t showing up in droves either. Microsoft claims 6K apps in its store, a poor showing when Android and Apple are both well above 100K.
In truth, WP7’s prospects were never very good. It has neither Apple’s reputation for UI and polish to build on nor Android’s low-cost and open-source appeal to hardware partners, and was launched years after both competitors were well established. The disastrous failure of the Kin phones gave Microsoft a black eye in the youth market, and Blackberry had already captured high-end corporate customers with deluxe Microsoft Outlook support before Microsoft itself got there to contest the territory. There was early speculation that Nokia (now helmed by a Microsoft alumnus) might jump to WP7 rather than Android or its faltering MeeGo project, but that prospect has faded in the face of continuing silence from both companies.
Microsoft claims 93% of its customers are satisfied with WP7, but one can’t help asking asking “93% of how many?” This is a product still looking for a market, motivated by Microsoft’s belief that it needs a play in the smartphone space rather than either customer pull or any kind of innovative value-add.
I predicted that WP7 would be a bust, swamped by Android and iOS. This was not a projection that required a lot of nerve even before launch; today, three months in and with nothing in the way of visible market penetration to show for all of Microsoft’s hype, it’s a complete no-brainer. No rabbit got pulled out of any hat, and network effects are pushing against WP7 rather than for it.
It would take an event at least as dramatic as Nokia betting the company on WP7 to revive this product. But I think nothing like that will happen, and that WP7’s affect on the smartphone market will amount to nothing more than statistical noise.
You may call it comic relief, but I cringe, much as I do at shows like “Three’s Company” which are supposed to be funny, but which in reality are just pathetic and embarrassing.
It’s nice to see the network effect working *against* Microsoft for a change.
> I predicted that WP7 would be a bust, swamped by Android and iOS.
Weren’t you just accusing me of being chicken-shit for making the exact same prediction?
“It would take an event at least as dramatic as Nokia betting the company on WP7 to revive this product.”
It is easy to dismiss this as very unlikely. But I keep wondering what Nokia *is* going to do. They looked like the 800 lb gorilla not long ago. I just can’t buy that they are still clinging to Symbian or MeeGo as any sort of viable strategy. But nothing comes forth. The silence seems deafening.
Microsoft will merge the handset biz of Nokia into the Borg.
Nokia’s shareholders will get a nice pump.
Windows Mobile will obtain instant market share and market access via re-branding.
WP7’s prospects are not good, but it is too early to write it off.
There is no doubt that MS has the financial resources to force WP7 into a relevant position in the market. And they also have a history of being late to the party and wearing funny clothes when they walk in the door. But…in those situations where MS achieves clarity of purpose, they can be a brutally efficient competitor.
The millstone around WP7’s neck is that indications are (so far) that clarity of purpose has not been achieved. MS’s efforts are an extension of the three year slide in which their actions show that they haven’t reconciled what they want to do in the mobile market and what makes financial sense to do.
I speculate that the root problem is that there is no viable scenario in which WP7 directly generates enough profits to feed a baby in a household the size of MS’s. Even in a future where they captured 30% of a smartphone market grown to 600 million units per year, that would be 180 million licenses per year. Assuming a net of $6 per license you get 1.08 billion in profit. That doesn’t amount to much in the MS scheme of things. If they took 90% of the same market and tripled the net, they would be getting close to numbers that rise to significance.
But these are rosy scenarios that will be costly to achieve.
It’s also hard to see how a successful WP7 would have tangential benefits that would spill over to other existing product lines. At least not in numbers that are large enough to matter.
That’s why WP7 will likely fail. MS is a software company that makes its living selling high dollar products to lots of people. The lack of articulable financial gain will limit the amount of resources they will commit to the effort. UNLESS…clarity of purpose is achieved.
If Ballmer happens to arrive at work some morning with bruises on his head from whacking it repeatedly all night and repeating “It’s all about mobile, stupid!” over and over and over, the game could change.
The Nokia idea won’t happen until it is too late in the game to affect the outcome for either company.
>The Nokia idea won’t happen until it is too late in the game to affect the outcome for either company.
I said previously that I thought Microsoft and Nokia have about a four-month window from now before anything they do together would be irrelevant; what I base that on is a projection that Android’s market share will go from its current 25% (already #1) to upwards of 35%. That’s a conservative projection, too; it wouldn’t astonish me if Google bettered that by ten points. At >35%, with Apple and Blackberry and HP and everyone else in the smartphone market scrapping over <65%, the network effects will suck all the oxygen (read "developer attention") away from Android and whatever the second-place platform is, probably iOS. Life will suck for minority players, whether WP7 or MeeGo or QNX.
The straight-line implication is that it’s already too late for a Microsoft-Nokia tie-up to do any good. You don’t do M&A or even a technology alliance in six months from a standing start, that’s all the amount of time they’ll have had since Elop took over at Nokia, and there haven’t been any investor rumors that sound like they’ve already started. Conclusion: they missed the window, if it was ever there to begin with.
When did you make it?
ESR was making that prediction at the same time WM7 was getting announced.
Nokia’s problem isn’t that they’re losing the smartphone market.
Their problem is that there are signs that the smartphone market is going to swallow the dumbphone market, where they’ve always been(and still are) the 800lb gorilla.
You know, it’s kind of sad to see Microsoft turning into IBM.
A big company that once looked invulnerable, hip deep in money and talent….. that can’t seem to get out of it’s own way.
Wonder how long it will be before Windows itself is dead? And once that goes, what’s left?
> Microsoft will merge the handset biz of Nokia into the Borg.
That would be interesting. Although it runs counter to Microsoft’s usual MO of extracting rents for software and letting the hardware vendors duke it out for scraps, it may have come to Microsoft’s attention that things have come full circle — these days some of the companies making the most profit actually make things. (Well, not directly, of course, but the actual laborers who make things haven’t been profitable in a very long time.) Obviously, Microsoft has a bit of experience here with the XBox and mice, etc.
If Microsoft did choose to jump into hardware in a big way, they could do worse than buying Nokia. An exclusive WinMobile 7 with a snazzy handset design might be just what it takes to get Nokia back on track. Although Nokia’s profits are suffering, they are still masters at pumping out cheap handsets at razor-thin margins, and have the sort of volume Apple can only dream of.
Actually, if Microsoft reads the tea leaves and realizes that Nokia is the only vendor at all interested in its smartphone offering, I could easily see them swapping the mobile division and a chunk of cash for a lot of Nokia stock. They won’t have the balls or stomach for an outright purchase.
>Microsoft will merge the handset biz of Nokia into the Borg.
If so, they better move fast. I think another two quarters of Android growth will move the market to a place where not even a Microsoft/Nokia tie-up can get any traction at all.
But the reasons this would be an idiotic idea for Nokia remain potent. For them, WP7 would bring all the disadvantages of someone else controlling their software (and skimming off the profits) in exchange for a Microsoft brand that is weak in this market, especially after the disaster that was Kin.
It would be a poor technical fit, too. What good would Nokia’s ability to manufacture low-end handsets in volume be when LG reports that WP7 is high-end handsets only?
Your prediction is that “WP7?s affect on the smartphone market will amount to nothing more than statistical noise.”
While you might be right [of course], I have a different prediction. I believe that the then current Microsoft Windows Phone phone [11?, 9?] of the “7” family will have a meaningful share [> 10%] of the global smartphone market 5 years from today.
I’ll check back with you then, right or wrong.
I think most of the basic core competencies are the same.
“There is no doubt that MS has the financial resources to force WP7 into a relevant position in the market. And they also have a history of being late to the party and wearing funny clothes when they walk in the door. But…in those situations where MS achieves clarity of purpose, they can be a brutally efficient competitor.”
I’m trying to think of an example where Microsoft entered a market and became a “brutally” effective competitor, **for a market where they were unable to directly leverage their desktop monopoly.** (Not to be confused with leverage the *cash flow* from the monopoly.)
I just can’t think of one. Office? Heavily based on the existing Windows monopoly. Xbox? Moderately successful, but no threat to the Playstation 2.
What did you have in mind as an example of MS as “brutally efficient competitor”?
I have avoided adding my two cents on the smartphone debate. I don’t own one; in fact, I resisted getting even the most basic phone for the longest time. I have handled both the iPhone and an HTC EVO. The iPhone maybe edges out the EVO on usability and the ‘it just works’ factor. However, my EVO friend recently had an idea for an Android app and came to me to write the code, so here over the last couple of weeks I’ve jumped into the world of mobile Java development (I intend to write on this later). The development tools ‘just work’. Getting Eclipse up and running with the right plugins and SDKs was simple, and I worked through the first few hello-world examples in my first sitting. I had a working toy program in short order. On the other hand, I have only studied Objective-C, but it looks too much like a mutant variety of C++, which I loathe. I cannot imagine having gotten as far with iOS/ObjC/Cocoa as I have with the Android/Java. On the gripping hand, what tools are available to Windows 7 app writers? According to Wikipedia, XNA and Silverlight, with VS as the IDE. Really?! How many wretched Visual Basic apps could come of that? Market share aside, only die-hard Windows fanbois would put up with that. The Windows phone is a non-starter. I reserve judgment on whether Apple resorts to its old MO of delivering the so-called BMW of hardware/software/UI or actually takes off this time, but I have a sneaky suspicion it’s going to be ARMs and Linux as far as the eye can see.
Do not underestimate the sheer amount of cash that Microsoft can throw at a problem, if it so chooses. The Playstation 2 had tons of network effects with PS1 backwards compatability and totally kicked the Xbox’s ass. MS lost a hundreds of millions of dollars, maybe more. The 360 is outselling it and, if memory serves, is actually making a profit these days.
The console market is not the phone market, of course. But if all WP7 and its successors do is toss out 3-4 good phones per year for the next ten years, they WILL end up with marketshare. And the cost of doing so for MS is not even very high, and 5-10% of a large market is still a large amount of money.
>Do not underestimate the sheer amount of cash that Microsoft can throw at a problem, if it so chooses.
I think the previous commenter who pointed out that Microsoft needs to achieve “clarity of purpose” first was right – especially given that Microsoft’s cash hoard is not what it once was. They’ve burned up a lot of it on stock buybacks to keep the share price from tanking.
I had a brainstorm just now. If Microsoft wants to buy market share, the smart move wouldn’t be to buy Nokia…it would be to buy Research In Motion. They have a customer base and line of business that Microsoft understands; the game would be (1) Buy RIM, (2) Strangle QNX and replace it with WP7, (3) Profit?
Ok, long time lurker, first time poster.
If Microsoft wants to buy market share, the smart move wouldn’t be to buy Nokia…it would be to buy Research In Motion.
I doubt Microsoft would dip its toe back into the antitrust waters. Even if it could get approval from US authorities, the EU, etc., it would likely take too much time, money, and bad PR. That said, it would be its best bet to gain market share.
Windows Phone is an amazing case study. To their credit, Microsoft was one of the first in the market for smartphones, but never put out a product that got people even remotely excited.
> especially given that Microsoft’s cash hoard is not what it once was. They’ve burned up a lot of it on stock buybacks to keep the share price from tanking.
Microsoft is currently sitting on $43.25B in cash. Last Sep it had $36.8 billion. In 2004 it was sitting on $56 billion in cash.
Microsoft has also (as you state, but didn’t provide any figures) spent $76 billion repurchasing its own shares since the 2006 financial year, and it’s now partway through a $40 billion buyback that will run until 2013. Yes, $106B is a lot of money, but they’re mostly paying for this out of proceeds (running the business), not by hitting the bank account.
While the deltas here represent a lot of money, I think it’s inaccurate to imply that Microsoft doesn’t have the cash to gobble up Nokia (or RIM), should it choose to do so. If for no other reason that normally these things aren’t cash transactions.
Nokia only made 1.2BEUR in profit last year. (3.3BEUR in devices and services, but the Nokia-Siemens unit lost a wad.) Still, Nokia’s management could get might attracted to Microsoft ‘investing’ a couple billion, which would improve the bottom line, at least for a while.
>Microsoft is currently sitting on $43.25B in cash. Last Sep it had $36.8 billion.
Ah, that’s useful information, thank you – Microsoft has repaired its cash position significantly since I reviewed its financials in 2008. You’re right, that’s enough of a war chest to fund a bid for either RIM or Nokia. More facts like this and less flaming would actually put you in good standing here.
MS does indeed sit on a lot of cash. But what is always amazing me is how little quality they can get for their cash. It seems love (for quality and design) cannot be bought for money,
MS has also been hemorrhaging brilliance for years now. It used to be that brilliant young developers would aspire to work for MS. But we now only hear about people leaving in droves.
And the balance sheets show that Windows and Office licenses are the cash cows, the rest is losing money or barely profitable. There is no phone market where MS will be able to make enough money to make it worthwhile to spend a lot. Their only fear is that a smartphone OS will move up the chain into laptops and desktops. If that OS would be Android/Linux, it could accelerate the collapse of the Windows&Office franchise. Actually, the moment you can dock your smartphone to use it as you main computer, MS will collapse.
I see Windows Phone only as a strategy to prevent other options from getting into the tablet and mobile computing markets.
>I see Windows Phone only as a strategy to prevent other options from getting into the tablet and mobile computing markets.
I agree, that’s the agenda with respect to which WP7 would make the most sense if the goal were even remotely possible. But that horse is already out of the metaphorical barn; there’s no way WP7 is going to magic iPad and the flood of cheap Android tablets out of existence.
There is only one word I disagree with in your posting – it’s “effect”, not “affect” in the last sentence.
I kind of figured the hype on WP7 had run its course, just from watching Castle on ABC.
The week with the alien abduction mystery, I noticed that Richard Castle had ditched his iPhone for a Samsung WP7 unit. Of course, my immediate thought was, “How much did Microsoft pay for that to happen?”
However, in more recent episodes, he’s gone back to the iPhone. I gather from this that the bubble has burst.
So long, WP7. Don’t let the door hit you in the ass on the way out.
“I had a brainstorm just now. If Microsoft wants to buy market share, the smart move wouldn’t be to buy Nokia…it would be to buy Research In Motion. They have a customer base and line of business that Microsoft understands; the game would be (1) Buy RIM, (2) Strangle QNX and replace it with WP7, (3) Profit?”
That sounds even more off-the-wall and riskier than buying Nokia. RIM’s customers are business types, who don’t NEED their phones to be flashy, they need them to Just Work. I know my brother, a longtime Crackberry addict, would be quite upset if they jerked the rug out from under his phone and forced him into WP7. Even if they figured out how to skin WP7 and morph it into I-Can’t-Believe-It’s-Not-Blackberry-OS, I think the overall reliability issue would trip them up. They’d tank what market share RIM has left in the process.
>I know my brother, a longtime Crackberry addict, would be quite upset if they jerked the rug out from under his phone and forced him into WP7.
No doubt. But the rug was about to be jerked from under the present Blackberry OS anyway; like Symbian, the design is obsolete and inflexible and needs to be replaced. Reports are that RIM went through a bruising internal fight over whether to jump to Android or roll its own; the pro-Android faction lost, and RIM acquired QNX. So the messy process of skinning something into I-Can’t-Believe-It’s-Not-Blackberry-OS was already on rails for the near term, making it as least bad a time as any for Microsoft to buy in and insert WP7.
Someone pointed out upthread that an attempt at a RIM purchase would risk falling afoul of antitrust issues. True, but so would an acquisition of Nokia, and the EU’s competition watchdogs have been more willing to block Microsoft’s game than their American counterparts. Cultural incompatibility is the most frequent cost of failure in mergers and acquisitions, making a Canadian company look like a better bet than a Finnish one.
By every metric I can see, RIM looks like a more plausible acquisition target than Nokia.
II wonder if Intel/Nokia are looking at Apache/Java friendly Android interpreters at all to sit on top of Meego. Crazy enough that it may just work?
“Choi holds out some hope that WP7 will make better sales as it is ported to less expensive handsets, but that hope itself suggests the software is hobbled by bloat and poor performance. ”
Why? Wouldn’t bloated, underperforming software run even worse under less hardware? I don’t get this affirmation.
>Why? Wouldn’t bloated, underperforming software run even worse under less hardware? I don’t get this affirmation.
His implied theory is that WP7 will slim down as it’s ported. Either that or he’s counting on hardware improvements to support the bloat – which on a PC would be a good safe bet, but in this market is going to run headlong into battery-life constraints.
“I agree, that’s the agenda with respect to which WP7 would make the most sense if the goal were even remotely possible.”
On the other hand, MS’ stocks are kept afloat on the assumption of monopoly rents. If they cannot make their main investors believe they are working at keeping at bay any competition from smartphone OS’ for the next years, their stock price would fall through the floor.
As long as they can keep up a brave face at their stock holders that WP7 will be a valid player in the market, things will pan out.
Their own people see the writing on the wall and are deserting the ship. But investors can be lead around by “analysts”.
My personal hope is that Microsoft is able to carve out a comfortable third place for Windows Mobile. For various reasons the world of OEM manufacturers, like my own metal cutting industry, is dominated by windows. Basically the kalidoscope of hardware and software in the 1980s was supplanted by windows programming by the late 90s. With Hardware manufacturers offering Windows drivers for their products.
Since our software life cycles streches across decades (the software I maintain has been in continuous development since 1985). And we are bound to the hardware we support to control machinary. This means we tend to move slower than other types of applications and currently have a lot of our software in Windows.
In many cases (but not all) it is cheaper and quicker for many of us to develop application for Windows Mobile as Microsoft makes it easy for us to reuse our code between the different platform. Or at least used to be able too. It is still technically possible but I fear Microsoft will be going down the wrong road in eyeing Apple’s road rather than Androids. The Apple model of business really hurts OEM trying to use their software and hardware and for a brief moment in the early 90s there was some work done using Apple computers. But Apple was so focused on the consumer while Microsoft wanted Windows everywhere Windows won out in the OEM world.
It is my hope that Android beat the tar out of iOS and Micorsoft secures a comfortable third place by returning to the idea of trying to get windows everywhere. If Microsoft goes down the route of Apple they are going to kill themselves in the OEM world which will give Linux, and Linux based environment a huge advantage.
Before in the 90s the tools weren’t there for Linux but now in the 2010s it growing leaps and bounds. If a OEM vendor switches they are not going to return given long term freedom and stability Linux and the alternative gives.
I wonder how many of you have actually used WP7 for more than 5 minutes in a store. I have. Two of my friends are MS diehards and I got to play with their phones for extended periods of time. The interface is clean, elegant and the underlying concepts are the most innovative ideas i’ve seen since the original Iphone. Yes, it has limitations: no copy/paste, no multitasking, limited sync abilities. The underlying SDK is limited in some very significant ways. And it is about three years late to the party. But MS has a solid base here and it can improve it quickly if they decide to throw resources at it. If they achieve”clarity of purpose”, as others have put it.
I also suspect you overestimate the power of network effects in this case. Successful apps have been ported to all major mobile platforms. Yes, developers of FartApp #34 will focus iOS and maybe Android , due to limited resources. But customers won’t really care if their day to day apps are available on their platform of choice. Facebook, Twitter, Foursquare, a few games… hell, Angry Birds has been ported on pretty much any modern mobile platform (WP7 is coming out soon).
IMHO, MS and RIM will probably not happen. The products are just too different and RIM users are fanatical about their interface. Plus MS might have their own ideas about enterprise integration. If it does happen, it will be more of an “acquire and shutdown” type of operation.
MS and Nokia: we’ll probably see a bunch of high end handsets sporting WP7, but Nokia’s strong point is now low end devices. (yes, that means they’re screwed)
Looking at news articles today, it appears Bloomberg got quarterly sales numbers for WM7. Most notable quote in that article?
Wait, why is Sullivan trying to convince us that there’s a better indicator for sales than actual sales figures? Is it perhaps that we can already publicly see WM7’s sales flattening? (Per official press release here, MW7 sold 1.5M phones in its’ first 6 weeks; now 3 full months in, they have sold 2.2M total. That’s a really steep decline in sales!)
Even with these numbers, your final statement about “statistical noise” stands… after all, we know how many smartphones in total were sold last quarter, and that 2.2M is a tiny drop in the ocean.
>Looking at news articles today, it appears Bloomberg got quarterly sales numbers for WM7.
No they didn’t. Comparison with other sources revealed that the Bloomberg figure is the number of WP7 licenses sold to handset makers and/or carriers, not the number of finished WP7 phones sold to actual customers. MS has been channel-stuffing – the actual number of phones sold is something less than quoted, with the balance as some partner’s unsold inventory (or even just so many permission slips to put WP7 on phones that don’t yet exist).
More significant in that article is the estimate from the notorious Microsoftophiles at Gartner that MS’s smartphone share has dropped from 7.9% to 2.8% in the last year. That’s not just bad, it’s disastrous – it’s the kind of numbers that’ll have channel partners quietly bailing out.
“Wait, why is Sullivan trying to convince us that there’s a better indicator for sales than actual sales figures?”
Because if investors loose faith in MS’ monopoly, they will dump the stock.
The question is, who pays a independent stock analyst who publishes his/her analysis on-line?
I know Microsoft has to distract from flattening sales. [Your answer seems to presuppose a MS monopoly in the smartphone market; however I know you mean the smartphone threat to their desktop OS monopoly.] I meant the question to be entirely a rhetorical lead-in to the following question; would adding an em-dash have clarified that? (“…actual sales figures?–is it perhaps…”)
History would suggest the opposite will happen. Slim and Microsoft are not two words you see in the same sentence.
Just an “off the wall” suggestion as a possible (desperate) Microsoft strategy:
They launch a “pay as you go” cellphone company like Tracfone or Virgin Mobile. They offer cheap smartphones running Windows Mobile, and contract with Sprint (or someone else) to provide the infrastructure.
Maybe they’ll go for the “Hail Mary pass”……
Unless you’re talking about people with a metric assload of money. (lCarlos)
> Just an “off the wall” suggestion as a possible (desperate) Microsoft strategy:
> They launch a “pay as you go” cellphone company…
An interesting idea… on the other hand, Sprint’s market cap is only $13 billion…
You know, it’s kind of sad to see Microsoft turning into IBM.
That would be a best case scenario. IBM does good research and is still producing innovation in hardware and manufacturing, Microsoft not so much. I have an admittedly strange bias towards innovation in hardware and manufacturing, it seems somehow more fundamental than rearranging bit clouds. I have a similar weak spot for agriculture as the basis of civilization.
“That would be a best case scenario. IBM does good research and is still producing innovation in hardware and manufacturing”
IBM even at its worst was a far more diverse and innovative company. Still, it gives hope that any Bumble can be reformed with a bit of dentistry and a stout choke collar.
It’s probably a no-brainer to declare that we’re now in the post-Microsoft era. Strange I have no stomach for celebration. Perhaps it is already old news.
>Strange I have no stomach for celebration.
I sure do. May Microsoft’s defeat be total and their humiliation utter, leaving wreckage so ugly that closed source is forever equated to failure.
Well, I do think their new game playing interface is cool. If they came out with an XBox 360 compatiable phone I think they really could clean up. Nobody is really going after the gaming market yet – and that’s a great way to build brand loyalty.
Breaking news: AT&T CEO: We’ll push Android phones
>Strange I have no stomach for celebration.
I’m not celebrating yet, but I certainly will. I have to support others’ use of it, and have been dreaming of the day that is no longer true.
>Breaking news: AT&T CEO: We’ll push Android phones: http://www.physorg.com/news/2011-01-att-ceo-android.html
Yow. This article is stuffed full of interesting revelations. Merits a blog post of its own, I think.
“If they came out with an XBox 360 compatiable phone I think they really could clean up. Nobody is really going after the gaming market yet –and that’s a great way to bulld brand loyalty.”
Phones work on ARM. So they need to rewrite their stuff to ARM.
Can they still do that?
>Can they still do that? [Port to ARM.]
Microsoft has already run controlled demos of Windows on ARM, so the answer is almost certainly yes. It’s the kind of task that CS professors call “uninteresting” – solvable by throwing enough cash at it. They can do it; the question is whether they can do it in time.
If I were a Microsoft planner, my worst nightmare would be a situation in which mobile devices and netbooks – all cored with ARM chips – stole the PC’s mass-market monopoly before I could field a response. Finishing the ARM port has to be strategic priority #1 if they’re going to keep the Windows./Office/Outlook lock on the market.
No. Objective-C is C with the message-sending bits of Smalltalk grafted on. An entirely different (and far less crufty) beast than C++.
> Objective-C is C with the message-sending bits of Smalltalk grafted on.
For some reason when I read this sentence to myself it sounded like this:
Objective-C is C with the message-sending bits of Smalltalk crufted on.
Netbooks are finished. The iPad delivered the coup de grace to this solution looking for a problem about a year ago. We’re just waiting for the death throes to subside.
Microsoft has the resources to easily make Windows for ARM the dominant non-smartphone mobile platform. They are often late to the game, but they usually end up winning anyway. Remember The Road Ahead and its conspicuous lack of much mention of the internet? Didn’t stop Microsoft from controlling the Web five years on.
Jeff Read Says:
“They are often late to the game, but they usually end up winning anyway. Remember The Road Ahead and its conspicuous lack of much mention of the internet? Didn’t stop Microsoft from controlling the Web five years on.”
Microsoft won the Web? Even the dominance of IE hardly allowed them to control it. And that was very temporary. And that all came about because of their lock-in of the desktop. They have no such lever this time.
If WP7 dies, it’ll be sad (though I’m not sure Kevin isn’t right about long-term prospects being less than immediate doom).
Sad because I’ve actually had hands-on time with one (no SIM, but that only matters when you want to make a phone call), and it’s got a good and honestly innovative UX, and I always approve of that.
I’m actually not quite sure why sales numbers are so bad for them, assuming they are (which seems very, very likely from MS not crowing about them and being so evasive). Bad marketing, perhaps, and doubtless some people scarred for life by Windows Mobile*. But actually having seen and touched one, they’re actually kinda good.
(* Having used WM/CE devices as handhelds in two separate “PocketPC” implementations, I was dubious as hell about WMP7. But by God they actually managed something that wasn’t a pile of crap.)
(And I’m with Jeff Read above. Netbooks aren’t important – and netbooks aren’t primarily ARM cored. It’s possible that will change, but I’m exceedingly dubious.)
Netbooks are finished. The iPad delivered the coup de grace to this solution looking for a problem about a year ago. We’re just waiting for the death throes to subside.
Now, now, netbooks did solve a problem: a simple computer with a long battery life that you could take with you everywhere.
Prior to netbooks, laptops were definitely not portable. They weighed too much to actually carry with you everywhere, or their battery life was horrendous, or both. Usually both. Even a laptop weighing 5lbs is too much to carry around consistently – if you don’t agree with that, it means that your car carried your laptop more than you did.
Obviously, what netbook consumers really wanted was a tablet, but up until the iPad, nobody made a tablet that was good enough to replace a laptop.
winter: No ARM device has anything like the power to be “XBox 360 compatible”, especially in emulation – which it’d have to be, since the X360 is a PowerPC architecture.
Sure, they could trivially (like esr’s “uninteresting”, indicating it would just take some time and money, rather than anything actually difficult) port the X360 OS to ARM.
But nobody’s going to port existing games, even if there was appropriate hardware, which there isn’t.
Note that Sony’s not trying to make the PSP2 (just announced) “compatible” with the PS3. It’s not even compatible with the PS2, though it can do PS1 games in emulation. And the PSP2 has a modern two-core ARM Cortex A9…
>Nobody is really going after the gaming market yet – and that’s a great way to build brand loyalty.
That statement’s about a day late. Just yesterday, Sony announced their PlayStation Suite, for Android.
> That statement’s about a day late. Just yesterday, Sony announced their PlayStation Suite, for Android.
Woo Hoo! Thanks for the info dtsund. I feel vindicated.
@esr: You’re probably right, the Microsoft/Nokia window has passed, especially from the M&A perspective.
@Cathy: Your statement “for a market where they were unable to directly leverage their desktop monopoly” is absolutely correct and I should have included that caveat in my post. MS has not been successful in generating major profit from anything not leveraged off the Windows franchise.
However, the Windows franchise does give MS a strategic advantage over its competitors should they choose to exploit it. They have the means and opportunity to add features to Windows/Windows Phone that would enable the two to play together better than someone else’s mobile OS might. Use of undocumented functions/API’s, etc.
Yes, they run the risk of crossing the anti-trust line, but they’ve shown the willingness to do that in the past when they really want to. Paying the anti-trust fines after accomplishing your objective is just part of the cost of doing business. You just have to be careful not to cross the line so far that someone tries to stuff structural changes down your throat. MS has successfully danced on the edge of that cliff in the past.
Earlier I used the terms means and opportunity. What’s still missing is motive, or “clarity of purpose”. Without it, WP7 will be (as esr put it) “statistical noise”.
Here’s another scenario in which WP7 could avoid oblivion (without MS needing a “clarity of purpose” epiphany)…
IMHO, Google and the wireless carriers are mortal enemies. Or at least that’s the wireless carrier point of view. They put up with Google/Android because it sells phones. They are under no illusions about where they fit in Google’s long term strategy.
As context, take recent developments with Google Voice. Google is slowly becoming Skype with lawyers, guns, and money. Google has been moving incrementally on this. I suspect their strategy is to not make any overtly hostile moves until Android is ubiquitous enough to not get hurt in the aftermath. (Personally, I wish they’d hurry up)
If Google makes additional move(s) that are perceived by the carriers as a threat to their current business model, they could get downright friendly toward WP7. If that happened soon enough, and MS made some of the necessary improvements to WP7 quickly enough, it could change the outcome.
Not a strategy I’d bet my business on, but not out of the realm of possibility.
@Kevin: “However, the Windows franchise does give MS a strategic advantage over its competitors should they choose to exploit it. They have the means and opportunity to add features to Windows/Windows Phone that would enable the two to play together better than someone else’s mobile OS might. Use of undocumented functions/API’s, etc. ”
Of course, but this doesn’t work in every category. They have failed to leverage it in the Xbox space, for example. And how much longer will it be before the Windows cash cow disappears?
We’ve all talked and thought extensively about how Apple, or Linux, or [fill in the blank] was going to out-compete MS with a better product and take over the desktop. But this ignores the bigger threat to Microsoft: their biggest competitor for some time now has been *their own installed Windows base.*
How many users voluntarily go out and upgrade to the latest Windows, especially at retail price? How many companies are eager to upgrade? (My last several employers — Windows shops all — have been very, very slow to upgrade. Heck, the Office on my work computer is 2003, not 2007.) It’s difficult to get users, corporate or personal, to upgrade even at minimal cost; it’s nearly impossible to get them to upgrade at a profit.
The *only* thing keeping the MS monopoly in place in personal computers is the lack of OEM installs of other alternatives. The corporate space is very conservative and doesn’t want anything new…including a new MS product.
>The *only* thing keeping the MS monopoly in place in personal computers is the lack of OEM installs of other alternatives.
>The corporate space is very conservative and doesn’t want anything new…including a new MS product.
While I agree that the corporate space is not looking to upgrade, there is one other thing that helps keep the MS lock in corporate in place: Microsoft Office.
The company I work for is in the process of slowly migrating away from windows to OSS solutions. The back end is solved, and most of the front end is easy, if just time consuming. Office however is the biggest anchor we’re facing. Too many people’s work flow uses various office documents, and too many of those documents use certain features or bits in MS Office that just don’t exist in the alternatives. Even being lucky enough to have employees willing and able to experiment and attempt the migration, we’re finding that the ideal of moving to OO for everyone is dying a death of 1,000 paper cuts. There’s no one thing, but all the little annoyances add up and add up quickly.
And I can think of at least 3 major (F500) corporate environments where where Windows XP is still the only desktop OS approved for use.
You are pretty much not fit to participate in post-industrial Western society unless you have and can use Word and Excel. Just to get a job, your résumé must be submitted in Word format in many places. Correspondence and documentation often takes the form of Word documents and — sophisticated bug tracking systems notwithstanding — management will often ask you to maintain an Excel spreadsheet of work progress. Then there’s the fact that Exchange and Outlook are pretty much the accepted solution for electronic communication, scheduling, and meetings.
And this, kiddos, is why proprietary software ain’t going away any time soon.
Don’t underestimate corporate indecision and infighting. The top levels of Nokia under Kallasvuo really have been unable to set a new course. Elop is still new and I doubt that he has any better ideas. I actually think that it would still be possible to pull off Meego and get significant share of the market with it, but it would take something like Jobs-level hype and reality distortion, which is not exactly Nokia’s strength.
From the mobile developer point of view, lack of multitasking in WM7 is horrible, horrible, horrible. A real killer for any “smart” application.
Example: forget porting any decent VoIP client.
Marian Kechlibar Says:
> Example: forget porting any decent VoIP client.
From the carriers’ point of view, this is a feature.
Jobs traded exclusivity for his vision, and his vision cannot tolerate too much typical carrier ugliness.
ESR’s insight is that Google’s Android Strategy gives a voice to the consumers, and the marketplace may be more successful than any philosopher-king.
>Do not underestimate the sheer amount of cash that Microsoft can throw at a problem, if it so chooses.
They threw a lot of money at Longhorn, and still failed to ship it. Do not overestimate the power of cash when it’s wielded by incompetent management.
> On the other hand, I have only studied Objective-C, but it looks too much like a mutant variety of C++, which I loathe.
You couldn’t be more wrong. Objective-C is a very small delta from C: a handful of new keywords and the message sending syntax. Any competent C programmer can learn Obj-C in a day, and be an Obj-C language lawyer in a week.
> Any competent C programmer can learn Obj-C in a day, and be an Obj-C language lawyer in a week.
Then it’s not the syntax that’s the learning cliff, it’s all the new libraries. It does sound cool though, but if I’m correct and it’s never made it out of Apple land it’s even less useful to me than C#, which has at least made it to Linux. I work on Aix, HP, Linux and Sun boxes. I love cool new and old stuff. Apple should work harder at making Obj-C have a community on those machines if they want more developers. Similarly, Microsoft should work to put C# on the non-Linux Unix boxes – including Macs. Right now C, Perl, Python, Lua and Java all beat Obj-C and C# for me because they all run just about every where.
@Some Guy: The problem isn’t learning Objective C, it’s that you have to learn an entire framework (Cocoa) to write an iPhone app. Consider the difference between coding a simple mod_python application and a Django app.
With the Android SDK, the framework is much, much simpler and you have access to a decent subset of the Java class libraries. If you can already code Java, you can write Android apps quickly.
@Tom DeGisi: Same here — I’ve done systems engineering and administration on all four of those platforms –though at my new job I’ll be concentrating on Solaris and Linux, with a bit of NetApp storage thrown into the mix.
That’s kind of a lousy comparison. Of course if you know Java, working with something based on java is easier than learning a while new framework. But that applies equally to the other side, if you’re a mac programmer using ObjC, it’s much easier to pick up iPhone programming than to learn Java if you don’t already know it.
Now to be fair, ObcC/AppKit suffer badly from poor “getting started” documentation, though I couldn’t say whether or not that’s the same for Java, as I picked that up via classes rather than from online documentation.
>It does sound cool though, but if I’m correct and it’s never made it out of Apple land it’s even less useful to me than C#,
>which has at least made it to Linux. I work on Aix, HP, Linux and Sun boxes. I love cool new and old stuff. Apple should
>work harder at making Obj-C have a community on those machines if they want more developers.
There have been some movements to bring AppKit to other platforms http://www.cocotron.org/ and http://www.gnustep.org/ are the two biggest ones though I can’t say how complete they are as I’ve never used them. Yes, the frameworks are awesome and it would be nice if Apple ported them some, but honestly this goes back to the question I asked the last time this came up which is whether a framework like Apple’s AppKit could even survive in the mainstream OSS community, or would internal politics shred it and fork it too much to be useful?
Long ago, I did considerable development with the original Obj-C pre-compiler: StepStone’s. On 386 Dos machines with a memory extender library. This was before Next ever happened, and I never did learn that framework, not being the Apple type. But the language itself (with the basic StepStone class library that came with the compiler) was very sweet to work with, and the code ran quite fast, even with the messaging overhead. At around the same time, I looked at C++ several times at various points in it’s development as a language, and always concluded that I had no interest in a painful bondage-and-discipline language that pretended to be *the* successor to the elegant C. Constantly being in a fight with the compiler to get it to let you do what you need to is no fun. The 2 languages are as different as night and day.
@Cathy Re “Checkmate.”
I didn’t think we were on opposite sides of a chess board. In fact, it seems that we are on the same side.
I am an Android user, and a supporter of what I believe Google’s long term objectives are. I believe that the consumer ultimately benefits if there is true competition in the marketplace. The Windows/Office franchise has IMHO held back innovation in ways we can only speculate about. My company still uses XP (except in the Engineering department) and Office 2003. I stick with Access 2000 at home since I support a rather large database app for a non profit I work with, and converting it to Access 2007 or newer is a chore I simply don’t have time for right now. I upgraded to W7x64 not because I like it but because the 3d CAD program I use (at work and at home) requires it. When I upgraded my PC to W7 I ended up having to buy my wife a new PC since her XP machine and my W7 machine never did communicate on our home network after that. Bottom line is I have a history of being dragged (kicking and screaming) into upgrades I don’t need or want, have no gain in functionality, and are driven only by someone else’s idea of planned obsolescence.
Back to the idea of competition. It boggles my mind that, when I upgraded my home PC, I installed a ten year old application (MS Office 2000 Developer Edition) on a brand new PC with the latest and greatest hardware, and then installed years worth of patches and updates. Why? Because the newest software doesn’t do anything I need that the ten year old version doesn’t do, at least not to the extent that I am willing to fork out the cash to buy it. Either I am a Luddite or there has been a lack of innovation in the office productivity software arena.
The harsh reality is that the Windows franchise reached what I call “critical mass relevance” a long time ago. Critical mass relevance is achieved when your product becomes so relevant in the marketplace that everything else becomes irrelevant. Products that achieve critical mass relevance retain it until some new technology comes along that destroys their relevancy. MS survived 98 and Vista with little or no threat to the franchise. Anyone else who introduced products like that (without the backstop of critical mass relevancy) would have gone bankrupt.
As much as I (emotionally) want to see Android kick everyone’s arse around the block, I also do not want to see any smartphone OS achieve critical mass relevancy as Windows did. Despite my antipathy toward MS, I kind of hope that WP7 gains some traction as a reasonable competitor to Android. There are no viable alternatives. iOS, Symbian, Blackberry, WebOS, etc. will be fighting it out for 25% or less of the marketplace imho. The other 75% will belong to the non-proprietary alternatives.
I have tried to point out some plausible scenarios in which WP7 could avoid oblivion. I doubt if any of them will take place but in a left-handed way I can only hope.
Competition makes everyone stronger and benefits the consumer.
Actually, IIRC, StepStone’s Objective-C compiler translated Objective-C to K&R C and compiled the result. And 386 DOS extenders were always a bit dodgy, if you ask me.
>Apple should work harder at making Obj-C have a community on those machines if they want more developers.
Apple has no shortage of developers. This isn’t 1996.
>the language itself (with the basic StepStone class library that came with the compiler) was very sweet to work with, and the code ran quite fast, even with the messaging overhead.
That’s why NeXT picked it. C++ was never much of a contender; the internal debate was between Smalltalk, Objective-C, and Postscript (NeWS style). The 68030 hardware wasn’t really snappy enough for Smalltalk, Postscript required too much of a mind-shift for most programmers, and Obj-C delivered a lot of the benefits of Smalltalk.
> Apple has no shortage of developers. This isn’t 1996.
How about one more developer. Me. I need things that work where I do. And frankly I’m sure I’m not the only one.
Yes. I said pre-compiler. I forget what protected-mode C compiler we used, maybe Watcom? It was not MSC. Of course, many C++ implementations then were pre-compilers also, including the standard AT&T one. This was long before the Obj-C front-end to gcc was written. The 3rd party DOS extender was reasonable, as long as you followed some rules. (Better than the de-facto one that succeeded it, i.e. Windows.) If you wanted to access something in real mode, you had to write a real-mode routine to do that (a primitive, really), and then go through some hoops to switch modes and call it. But the bulk of the code ran fine in protected mode – it was much too large to fit in 640k, which is why we went that route in the first place.
> frankly I’m sure I’m not the only one.
Hey, I’m sure there are dozens of you. Now, why should Apple put their time and money into making life better for people who are using other platforms? How does Cocoa on Linux help them sell more Macs and iPhones?
I dunno — ask Apple. They certainly do this when they perceive it aligns with their interests. Web browsers, Unix kernels, printer drivers, C compilers, …
Yes, ubiquity is a tool that can be used for lock-in. But lock-in is uninteresting without the ability to seek rents. If you use an open source OS, with open source tools on top of it, you don’t really have that much to worry about. If you convince your friends to do the same, then you don’t have to worry about network effects making it painful to exchange information with them.
MS buying RIM? Very plausible. Note that there are a lot of personal connections between the two companies, which are probably the two largest employers of Waterloo grads out there. And RIM’s entire business model is built on leveraging MS products in the phone space.
iPads killed the Netbook? Gimme a break. The iPad finally put the stake into the Tablet PC, Android 3.0/Gingerbread and WebOS will handle the funeral itself. The Netbook on the other hand has become the defacto standard laptop. iPad is just too much money for what it delivers, running twice the price of a Netbook for a worse computing experience (it is however a MUCH nicer media reader/player solution). And Netbooks were never a solution in search of a problem. They solved three problems simultaneously, ie cost, battery life and portability.
> Now, why should Apple put their time and money into making life better for people who are using other platforms?
> Yes, ubiquity is a tool that can be used for lock-in.
We could argue about whether of not this is a good strategy for Apple. But all I’m really saying is that there is no reason for me to learn Objective-C, cool as it may be. So I put a wish out on the Internet for Apple to fix that. Sometimes my wishes come true, like when I said some phone manufacturer should go for the game market. No sooner did I comment than I was informed Sony announced a Playstation / Android development kit. Sweet. I must admit I was half hoping you would tell me about this wonderful cross-platform Objective-C based kit that Apple had released to encourage iPhone apps. But not all my Internet wishes come true so fast.
I think in the case of a technology like Cocoa/ObjectiveC, it’s more the case that ubiquity is a tool that can avoid lock-in. If developers are using Apple’s stuff instead of soupdujour.NET, then they aren’t locked-in to MSFT, and are available to Apple.
> If developers are using Apple’s stuff instead of soupdujour.NET, then they aren’t locked-in to MSFT, and are available to Apple.
Yeah. C# is the hot language headhunters want right now and .NET the hot environment. Before that it was Java. It has never been Objective-C. I wish it was going to be Perl 6, or Python or Erlang next but that’s not an Internet wish I think will even come close to being true. It’s true that Java didn’t save Sun and .NET may not save Apple, but Java did get a lot of developers employed and .NET is now doing the same. From where I sit successfully promoting developer employment is really supporting developers. And Android, by using Java, is helping a lot more Java developers stay employed.
And, BTW, this does lead to a way for MSFT to get developers on board. Come up with a .NET dev kit that produces apps which are really good on Windows Phone 7 and pretty good on iPhone and Android or even on Phone 7, RIM and Palm. Just brainstorming here. Now rip it apart.
Correction: .NET might not save Microsoft. Not Apple.
Objective-C became hot when Apple enabled apps on the iPhone. Its TIOBE rating shot through the roof starting in about 1Q 2009.
The odd thing is, in late 2010 Sony had announced an open source toolkit called SNAP based on Objective-C and GNUstep, then promptly iced it. Rumor was that Apple had released the attack lawyers, but the announcement of “PlayStation Suite” seems to suggest they said “Aw, fuck it, Android is already here, open source, and it works” and built what was to become SNAP on top of that.
I don’t think they’ll put the name “PlayStation” on something without wrapping it in a few nice thick layers of DRM and charging out the wazoo for devkits though. I could be wrong — I hope I’m wrong — but this is Sony.
> Objective-C became hot when Apple enabled apps on the iPhone.
Haven’t heard any buzz here, but then I’m not that tuned in.
Assuming you trust the TIOBE rating (and i’m assuming you do since you quoted it), it became “hotter” not hot.
Java is still takes the “hot” ranking at 17.77%. and hasn’t been below 15% since 2002.
Obj-c has been languishing at effectively 0% until 2009 where it started its climb to the 3% it’s at today (and still climbing).
C# is currently double obj-c’s number.
Obj-c is still climbing so it will be interesting to see where it gets to. Note however that it didn’t have the highest rise in 2010 and has never had the highest rise in a year (has never been awarded the “language of the year” award). At this stage however TIOBE seems to kind of support Tom’s assertion that Java was the old hotness, my read of it says it’s the current hotness too (especially if you include some of the C stuff being for JNI) but c# is still kicking obj-C’s butt.
Yes but the pre-conditions are not equal.
Java programmers come from all kinds of areas, and all kinds of systems.
Mac programmers using ObjC are probably doing UI for Macs (bald faced assumption: if you’re doing drivers for macs is probably in C/Asm like everyone else). I’d argue that you could remove the “Mac programmers” part and still say the vast majority of ObjC programmers are doing UI for Macs.
The point i’m making here is that the precondition to android being easy to learn to develop for is almost (but not quite) “Being a programmer”. The precondition for iOS being easy to learn to develop for is “be a UI programmer for Mac”. These two groupings are far from being equivalent.
Not to mention that the fad of the last 10 or so years was to teach Java to first year comp sci students. I.E. the current generation of programmers will (probably) all know java. This is apparantly starting to switch over to python.
> This is apparantly starting to switch over to python.
If true, another one of my Internet wishes (see above) may become true. Color me surprised! Thanks, JonB.
I’m praying for true democracy in Iran, Tunisia and Egypt. Maybe I’m on a roll! :)
Python makes a very good training language. Sure you’re not going to teach memory management or low level concepts in it but high level concepts translate very well indeed in a nice, easy, play friendly fashion.
>Python makes a very good training language.
I find it quite effective for production use.
> Come up with a .NET dev kit that produces apps which are really good on Windows Phone 7 and pretty good on iPhone and Android or even on Phone 7, RIM and Palm.
Oh yeah! They can call it the “LCDKit”, and all the third-rate coding monkeys who are too lazy to learn the native environment can make apps that suck rocks just like all the Qt turds! What a win!
 not to be confused with Liquid Crystal Display technology
Oops, I forgot. The LCD Kit already exists. They call it “mono touch”.
The only thing I ever knock python for is I wish its concurrency was more like Erlang. But it’s a small thing and Python’s decisions along those lines are at least sensible.
If you want Erlang, you know where to find it. ;)
My only serious complaint with Python is performance. Of course, if you throw enough hardware at it, performance is at least reasonable. There are options for improving performance, but I don’t think I like any of them yet.
Some comments on possible M&As:
Nokia has a market cap of $39B; and has $6.5B of cash in excess of debt.
RIM has a market cap of $31B; and has $1.8B of cash in excess of debt.
MSFT has $43B in cash.
So MSFT could acquire either Nokia or RIM, but it would use nearly all of their cash reserves. That’s not quite a bet the company risk, but it’s still pretty close. To paraphrase Daffy Duck, “they can only do it once”.
>My only serious complaint with Python is performance.
That problem will soon be solved: http://codespeak.net/pypy/dist/pypy/doc/index.html
Two words: Stock swap.
Sugar for the OLPC is all Python. And it does work on an XO, which is low power in most meanings of the word.
My yardstick is the time won during writing against the time lost in running the program. It is pretty daft to spend an extra month writing optimized C when you gain only a week pure run time over the useful life of the application.
As I said, I don’t like any of the options for improving performance. :) PyPy is incompatible with CPython.
@ Kevin: “Re: Checkmate. I didn’t think we were on opposite sides of a chess board. In fact, it seems that we are on the same side.”
Microsoft and I are on opposite sides of a chess board, and have been since at least 1988.
For years they were winning. I became excited about alternate platforms, particularly Amiga, only to watch them fail to achieve critical mass and become squashed. Finally Linux appeared in 1993, and I installed SLS 1.0 and Win95 in a dual-boot configuration. This late me live in the Unix world that I already preferred (after years of Sun 3/60’s, VAXes running Ultrix, and the like) while still having a toe in the Windows world where all the applications lived.
Over time, I’ve been able to do more and more on Linux and have less need of Windows. Except, of course, at my employers, who have generally been Windows offices.
I don’t care whether Microsoft continues to exist or not, but I very much want their stranglehold broken.
It’s also slow as balls. I mean pathologically, almost perversely slow. I don’t think I remember the default environment for a computer being such a painfully small fraction of the computer’s full speed since the TI-99/4A BASIC which ran on two interpreter layers stacked on top of each other.
Of course a lot of that isn’t Python; it’s bad coding. Go ahead, look at the Sugar code for the “icon throbber” effect that appears when you start a new activity. Heehee, it’s Daily WTF worthy!
Not all time has the same value! Of course there’s the obvious case: If a life support system misses its window by a few milli- or microseconds, that tiny amount of time becomes far more valuable than the time it takes to make the code efficient. If you’re running a very large Web site, and you can hlave the time it takes to respond to an HTTP request, you can roughly double the number of users on your site with the same hardware. Depending on how UI-focused you are, the time saved by avoiding a GC cycle in the middle of your UI loop (*cough*Android*cough*) in order to reduce jitter can well offset the time it takes to write your UI code with careful use of deterministic memory management rather than relying on the GC.
My favorite example of a mission-critical, real-time system with severe consequences for missing hard deadlines is payroll.
“Not all time has the same value! Of course there’s the obvious case: If a life support system misses its window by a few milli- or microseconds, that tiny amount of time becomes far more valuable than the time it takes to make the code efficient.”
And how is hard real time relevant to the current discussion? Does coding in any other language make this problem go away?
If my application has 10,000 users, 1 second shaved off their time to complete a task is worth 3 hours of my time for each time it would apply. So, indeed, I would take care shaving off that second. The same for a web site where I can attract more visitors if I half the delay.
And all those discussions about “language X sucks” is not more informative than “red cars suck, buy a blue one”. The language user must be productive, and she probably could not care less about what makes you productive