I posted this on slashdot, but old posts on slashdot can be difficult to find, and I just know I'm going to want to point to this in a few years and say, "See, see! I predicted that!" So I'm posting it here too. Note that I of course don't expect most of the exact details to come out; it's the overall general flow of the thing that seems likely to be dead-on accurate. So here it is, my projection for the development timeline for the next major version of Microsoft Windows:
2007 Q1 Vista released; work on Vienna begins.
2007 Q4 Microsoft announces Vienna will contain innovative new filesystem
2008 Q2 Microsoft projects release date for 'Vienna' as late 2010 or early 2011
2008 Q3 Microsoft announces Vienna will revolutionize the internet desktop
2009 Q2 Microsoft announces Vienna's filesystem will make search irrelevant
2009 Q4 Microsoft projects release date for Vienna as second half of 2011
2010 Q1 Microsoft announces Vienna will be inherently more secure than Vista
2010 Q2 Microsoft announces Vienna's new API will make developers' jobs easy
2010 Q4 Microsoft announces Vienna will have built-in internet telephony (VOIP)
2011 Q2 Microsoft projects release date for Vienna in early 2012
2011 Q3 Microsoft announces Vienna will work with next-generation security hardware
2012 Q1 Microsoft announces partnership with wireless internet provider to enhance Vienna's
internet telephony, allowing users to go "unplugged"
2012 Q2 Microsoft projects Vienna release date pushed back to 2013
2012 Q3 Microsoft announces Vienna's wireless internet telephony will make cellphones obsolete
2013 Q1 Microsoft announces Vienna's wireless internet telephony will be more secure than cellphones
2013 Q3 Microsoft announces Vienna kernel will be most secure OS kernel ever
2013 Q4 Microsoft projects Vienna release date in early 2014
2014 Q1 Microsoft announces the new filesystem may not be ready for RTM but will ship
just after Vienna in a service pack
2014 Q2 Microsoft announces Vienna public beta will be forthcoming later in the year
2014 Q3 Microsoft announces the new developer API will be spun off as a separate project from Vienna
2014 Q4 Microsoft promises Vienna release no later than 2015 Q2
2015 Q1 Deal with wireless internet company falls through
2015 Q2 Microsoft announces innovative filesystem will be in release after Vienna
2015 Q2 Microsoft announces Vienna will still feature "unplugged" internet telephony,
but user will have choice of third-party wireless providers
2015 Q3 Microsoft releases limited beta of Vienna to select individuals and companies
2015 Q3 Reviews of Vienna start coming out; reviewers note internet telephony not present
2015 Q4 Microsoft announces final product name for Vienna will be Windows Fiesta
2015 Q4 Microsoft confirms internet telephony will not be ready to ship with first release
2016 Q1 Microsoft releases public beta of Fiesta to a wider audience
2016 Q2 Microsoft announces final release date for Fiesta in November; nobody believes it
2016 October Microsoft announces Windows Fiesta will be available to select customers in
November, retail version will ship in January
2016 November Microsoft announces Fiesta now available to select customers
2017 January Microsoft actually releases Windows Fiesta
I suppose I could be wrong. In theory. Time will tell.
Father, in Whom We Live
14 hours ago
19 comments:
I guess I'll use comments to keep track of events as they unfold. This will make it easier later to compare the timeline as predicted to the actual events themselves.
2007 Feb 01
Bill Gates predicts next big release in 2010-2011. In case the article becomes unavailable at some point, here's the relevant part: the interview asks if MS will "be back with the next big one" in 2010-2011, and the first sentence of Gates' reply is 'Absolutely.' It is conceivable that he's talking more about (what is currently being called) Vista R2 than about Vienna, but the wording "next big one" seems to belie this.
If my interpretation is correct, this puts Vienna five quarters ahead of my timeline so far, because my timeline doesn't have the Vienna release date being projected as 2010-2011 until 2008Q1, and it happened already in 2007Q1.
Vienna is more than a year ahead of schedule! Yay!
2007 Feb 09:
Ben Fathi, corporate vice president of development with Microsoft's Windows Core Operating System Division, says late 2009, but when asked whether Vienna is the codename, he cannot disclose that.
My interpretation here would be that Microsoft currently wants us to expect Vienna in that timeframe, but they're hedging, figuring it may take longer, in which case they'll bring out a recodenamed Vista R2 (presumably little more than a service pack) at that point and spend a little longer on Vienna. This lets them promise things for Vienna that they know they can't get done by late 2009, without directly telling the public outright lies that they can't even swallow internally. (Some companies lie to the public without remorse, but it is my considered opinion that Microsoft, although they certainly spin things, do mostly believe their own press releases, at the time.)
Speaking of promises they won't be able to keep, Fathi avoids making any specific ones, although he starts to hint: "We're going to look at a fundamental piece of enabling technology." He gives two possibilities for what that might be (one of which is an extremely vague but grandiose UI improvement, and the other has to do with virtualization), but fundamentally he says twice in as many sentences that he does not, at this point, know what the "fundamental piece of enabling technology" will be. Further announcements are promised. I'm counting on that; it'll allow me to track progress against my timeline.
All told, this is roughly the sort of announcement I expected, and predicted in my timeline, for 2007 Q1. Basically it says that with Vista out, work on the next version has begun. So maybe we're not as far ahead of schedule as Gates' prediction on the first implied. We could be right on schedule, in which case, expect Microsoft to make the first big feature announcement for Vienna later this year, and if it's an innovative new filesystem, I'll look like a prophet.
I've run across a reasonable timeline someone else has maintained of the past history of Windows development, which may be a useful reference for comparison, although it's not detailed to the feature-announcement level.
Also, Microsoft seems to be officially taking the position that for now they aren't ready to say any more about Vienna (other than that work has started). John Pospisil seems to think that Fathi's comments may be viewed by Microsoft's PR department as Too Much Information, but I suspect it's more likely community speculation they're trying to discourage. All Fathi said, basically, is that they've started work, and this new quelling statement even goes so far as to confirm that much. I think it's third-party speculation Microsoft was reacting to, rather than Fathi.
Raju Shanbhag is an example of someone who has already speculated quite wildly about Vienna features, suggesting among other things voice and touchscreen input support (doesn't Vista *have* voice input support already?), integration of search (doesn't Vista have that too?), email (and that?), social networking tools, and the Live service, the ability to access your PC from anywhere (which XP has in the form of RDP, or I have completely misunderstood what he's suggesting), something big having to do with "Web 2.0", and an assortment of other things. As if all that meaningless speculation weren't bad enough, he calls Vista "old wine in a new bottle" and Vienna "a whole new world of computing". It seems obvious to me that this is the sort of thing Microsoft was seeking to discourage with their "focus on Vista" press release. Who wants to buy old wine when a whole new world is on the way?
So anyway, we shouldn't expect any more information from Microsoft about Vienna for a few months. That's fine: my timeline doesn't predict any until Q4, so things are still on schedule.
How do you know all of this, you are going almost 10 years in advance
I don't actually *know* any of this, in the sense of knowing the future in detail. I'm simply extrapolating based on patterns that have been observed in the past. In other words, I'm assuming that Microsoft will handle Vienna in much the same way they have handled previous versions. As stated, I don't expect my details to come out totally accurate, but I do expect the general flow of things to fit the familiar pattern in much the same way.
There is also new information to add. an article on slashdot claims that Microsoft has announced that Vista and Longhorn Server (which they're now calling Server 2008) will be the last releases capable of booting on 32-bit CPUs. The cited article is unavailable at the moment, presumably due to a traffic surge, so I'll be back to comment further once I can get a look at it.
The article that slashdot was citing, over on apcmag, has been updated now with a clarification: the "last 32-bit version" comment applies specifically to the server line. The statement was not meant to be talking about the desktop side (Vista), that was an inference someone made reading between the lines, and Microsoft has not yet determined whether Vienna will be 64-bit-only. (Of course Vienna will be available in 64-bit form, but so is Vista and so was XP, so that's not much of an announcement.)
So we're back on schedule, waiting for some announcement about Vienna in probably the fourth quarter, possibly having to do with the filesystem, or some similar or related component. Further bulletins as events warrant.
Ars Technica carried an article a week or so back reporting on some comments an MS exec made about Vienna. (It was also discussed on slashdot.)
Specifically, the claim made was that Windows Vienna will be "fundamentally redesigned" to better take advantage of multiple processor cores. If you read the article, it's clear he's talking about WAY more than current dual-core and quad-core technologies. Indeed, the numbers given are clearly pie in the sky, at least for anything resembling consumer-grade hardware.
Still, "fundamentally redesigned" is exactly the kind of hype I'm was looking for in the first big feature announcement in my timeline.
It's a kernel feature, it's far and away too technical for any ordinary end user to ever care about, it's probably not going to be a major determining factor for performance for most power users either, and it's not clear exactly what kind of changes would need to be made to make it happen (other than that they would presumably be rather significant), but it sure has the potential to get the tech press excited. If we continue to see announcements about multi-core over the coming months, it could be that this is the new logical successor to OFS/RFS/WinFS. That would put us a full quarter ahead of my timeline, incidentally.
OTOH, if this is just something the one exec is excited about, and we don't see any further announcements about it, then it's probably irrelevant to the timeline, in which case we're still waiting for the first big announcement this fall, some comparable kernel-level Big Feature.
Time will tell.
There's an article over on C|Net claiming that Microsoft is now referring to Vienna as "Windows 7", and that they plan to ship it "within roughly three years". (There is also a discussion of the article over on slashdot.)
The Windows 7 name, if it sticks, will be this version's third name. (It was once called Blackcomb, back when Vista was called Longhorn.) OTOH, this might not end up being merely a product codename (though the article is clear that it IS currently exactly that). If 2K and XP are NT 5.0 and 5.5, respectively, and Vista is version 6, then this could signal a return to normal version numbers, which I can tell you for sure would be a relief to consumers, who have had a hard time keeping straight which version comes after which ever since the marketing department decided to call NT 5 "Windows 2000", making it sound like part of the 95/98 product line. Then again, "roughly three years" is plenty of time for the marketing department to change its mind about the final name, if they've even considered the possibility of keeping the Windows 7 name at release time, which they may not have done. So the product easily could end up with at least four names by release time. I'm personally still pulling for "Windows Fiesta" :-)
Now, about the timeframe: the inclusion of the word "roughly" almost certainly signifies that the programmers know very well, and have told the managers, that this timeframe is unattainable. One supposes the managers are optimistically trying to talk them down. So the suits are telling themselves (and the public) three years, but they included the word roughly as a hedge in case the programmers, who are smarter than the managers, turn out to actually know what they're talking about. We're currently sitting in 2007Q3, so three years on the nose would be 2010Q3. The word roughly makes it more like the second half of 2010 optimistically, or realistically somewhat later. My timeline called for this announcement (release date of late 2010 or early 2011) in 2008Q2, roughly three quarters from now, placing us the better part of a year ahead of schedule, if the previous technical announcement (multicore) indeed turns out to be Microsoft's Next Big Technical Hype, replacing the filesystem -- which we won't know for sure until it gets mentioned again, but it seems likely.
The same article also projects Vista SP1 to come out before the end of 2007. I'm going slightly off-topic for this thread here, but this, too, is probably an optimistic estimate. I suspect Microsoft is feeling the need to do do something to address the (valid) concern that Vista is new and relatively untested, since, as in the past, early adoption figures aren't meeting their opimistic estimates (go figure). Getting a service pack out the door will help bolster Vista's image. I know I personally don't want anything to do with Vista until a couple of service packs come out, and I'm probably not the only IT guy thinking along those lines. Perhaps Microsoft has realized this. (There have always been people at MS who understood such things, but it's a question of whether the RIGHT people have figured it out, which is another matter.)
It's also, bringing things back around to topic, helpful to get a couple of service packs out the door if you want people to believe the next version is coming in roughly three years. I mean, if they can't do a service pack in a year, how are they going to do the next major version in just three years?
I should say a brief word here about the presentation that Eric Traut gave at the U of Illinois, which was discussed on slashdot. First off, I'm not sure MinWin fits any of the slots on my timeline, especially since nothing was said about how it would improve the consumer product. It's pretty much a purely technical thing, and technical things only figure on the timeline if they're the subject of significant hype. I'm not seeing that with MinWin, or at least not yet. Barring some future announcement about it, I'm guessing it doesn't mean anything for the timeline.
It is also worth nothing however that a good amount of the presentation had to do with hypervisors, which are one of the enabling technologies Fathi mentioned in February as things Microsoft might look at for the next Windows release. Nothing was promised then, and nothing has been promised now, so I still don't really know where this fits on the timeline, if it does at all. Perhaps at some point in the future we'll have a real announcement about hypervisors, but the next really big tech announcement isn't due until 2010 (represented on the timeline by VOIP). Before that we're looking for a flagrantly non-technical announcement (2008 Q3), a reaffirmation of the first big tech announcement (2009 Q2), a revision of the projected release date (2009 Q4), something about security (2010 Q1), and something for developers (2010 Q2). Of course, the timeline is a rough projection, and things may come a little out of order ;-)
Incidentally, FWIW, I was right about the Vista SP1 reldate: 2007 was optimistic, and they're now saying first half of 2008. I'm guessing they may actually get it out the door in that timeframe (it is just a service pack after all), but I wouldn't bet money on it.
s/worth nothing/worth noting/;
[Off-topic advertistement deleted]
A news story that's been making the rounds lately, which is primarily focused on a legal matter, nonetheless also contains a piece of "information" that may be construed as pertinent here: Seven is apparently "due for release in late 2009 or early 2010." This completely bucks my timeline's predictions. I didn't think they'd even have the chutzpah to suggest such an early date. (Granted, it's not Microsoft, directly, making this claim.)
If we look at how far that projected date is from present, it's somewhere in the neighborhood of 6-7 quarters out. My timeline has a 7-8 quarter projection in 2009Q4 and a 3-4 quarter projection in 2011Q2. Interpolating, a 6-7 quarter figure would come in round about 2010 on my timeline. It's 2008, so that arguably sets us two years ahead of the timeline, in terms of release-date projections.
However, the intervening announcements have not yet been forthcoming. That could mean one of two things: either the release-date projection in this story is an anomoly (which, given that it didn't even come from the horse's mouth, is entirely possible) or else Microsoft has completely altered strategy and is really trying to stay mum about possible future features until closer to release time.
What we have already had, surprisingly, is a lot of talk about IE8. I suppose that could arguably stand in for the "internet desktop" announcement on the timeline, but on the other hand the IE people, to my knowledge, have not so much as mentioned the next version of Windows, and furthermore they have specifically said that IE8 is supported on not just Vista but also XP SP2. Granted this is a beta, but why beta test on a platform you have no intention of releasing for. It doesn't look like IE8 will be Seven-only, so I'm inclined to say that IE announcements don't count toward the Seven timeline.
Which leaves us, as best I can tell, somewhere in 2008 on the timeline, roughly right on target.
Vista SP1, incidentally, is in beta and looks to be on target for a 2008Q2 release. In fact they're saying it'll be avaliable for download this month, and via Automatic Update and OEM preload in April. That's close enough that it probably won't slip much. It's also close enough that they probably won't say anything more about Seven at this point until the Vista SP has come out, because they wouldn't want to steal its thunder.
According to some guy going by the name of Dev Corvin, Microsoft is planning to drop binary backward compatibility in Seven and use emulation (like Apple's Classic).
I've written a more detailed discussion of this issue in a separate post, but I wanted to make a note of it in this thread as well. Information about Seven has been thin on the ground so far (a strong indication, IMO, that we're very early in the dev cycle still), and this has the potential, if it's for real, to be quite significant.
More information has emerged, this time mostly about dates. Executive summary: Seven development may be running significantly ahead of my timeline. With an optimistic interpretation of available information, it could be as much as six quarters ahead of schedule.
Well, now they've put a specific month on their projection, and I'm completely convinced that actual development is well ahead of my timeline. Fewer feature announcements, less wasted time, sooner release than I predicted. In other words, I was off, way off.
windows fiesta?
2015 vienna beta?
2013 microsoft announces new release for vienna?
what is this?
are you kidding?
I wasn't kidding, but I was wrong, and I think now I understand why.
Take a look at this recent article on PC Pro News. I quote:
Sinofsky said Microsoft's decision [...] will benefit those who have already migrated to Vista [...] citing the company's decision to change the underlying Windows version number to 6.1 rather than 7 as a sign of its intention to smooth the upgrade path. "If it works on Windows Vista, it'll work in Windows 7. The move from Vista to Windows 7 we expect to be seamless."
Does that sound like Blackcomb to you? This is an incremental, minor-version release, which we can probably best think of as Longhorn release 2 (where Vista is Longhorn release 1).
Windows Seven is the next release after Vista, but it's not Blackcomb. Blackcomb has been set on the back burner again, presumably pushed back to a subsequent release.
Which, incidentally, makes the recent announcement about Azure suddenly make sense. That news is an uncannily good match for 2008Q3 on my timeline. We still haven't heard anything about a revolutionary new filesystem, though.
Right, so Seven is in beta now and, if you listen to the buzz (e.g., here, here, here), it apparently has very few new features, unless you count smooth operation, perf improvements, hardware support, and bug fixes, the sort of thing you normally expect to see in Automatic Updates and/or service packs. To me this seems to confirm that Seven is essentially a second release of Longhorn, and that the next major version, Blackcomb (which this timeline was intended to be about) will follow at some later date.
Post a Comment