How will Windows 8.1 Update 1 be deployed?

Discussion in 'Windows 8' started by ian82, Mar 2, 2014.

  1. ian82

    ian82 MDL Expert

    Mar 7, 2012
    1,147
    302
    60
    Will it be like it was the case from Windows 8 > 8.1 where it was actually a whole new installation and the old Windows 8 was put in a Windows.OLD folder or will it be a simple update via Windows Update where the user loses nothing?
     
    Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
  2. CODYQX4

    CODYQX4 MDL Developer

    Sep 4, 2009
    4,812
    45,366
    150
    Sounds like it would come from WU, like a Service Pack, in the sense it adds features and changes the way the OS works a bit.

    Hopefully MSDN releases new ISOs for everything, I prefer that over integration as I can validate the hash at any time.
     
    Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
  3. ian82

    ian82 MDL Expert

    Mar 7, 2012
    1,147
    302
    60
    Ok so I shouldn't worry so much that I just did a clean installation of Windows 8.1 and would have to do another clean install it's just an update.

    Coz from 8 to 8.1 it sounds like an update but it was a whole OS reinstallation because the old Windows folder was renamed to Windows.old
     
    Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
  4. arseny92

    arseny92 MDL Secret Weapon

    Sep 22, 2009
    570
    1,260
    30
    #4 arseny92, Mar 2, 2014
    Last edited: Mar 2, 2014
    A major update, such as Windows 8 > 8.1 required a full OS upgrade, due to 8.1 actually being an entirely new system - therefore being a full OS upgrade (through UpgradeMatrix), while also shifting the major build from 9200 to 9600.
    A minor update, such as Windows 8.1 > Update 1, won't require a full upgrade, as it is going to be a normal Windows Update GDR update, similarly how Service Packs worked in the past - only the build delta will be rounded to 9600.17xxx.


    Either channel will work, when it'll be released: you can install the full clean build from an ISO containing it slipstreamed, or you can apply an update to an existing system to achieve the same result.
     
  5. ian82

    ian82 MDL Expert

    Mar 7, 2012
    1,147
    302
    60
    Very very informative post and now I get the clear picture! thank you very much for taking the time to reply man!

    Cheers
     
    Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
  6. FaiKee

    FaiKee Misinformation spreader

    Jul 24, 2009
    4,305
    5,803
    150
    MS doesn't round up deltas, win7 SP1 was 7601.17514. :D

    And win8.1 Update1 is actually a new version, DISM reports 9600.17019 instead of 9600.16384, MS just have no room for a new RTM number because win9 had been started on build 9622 two years ago. :p
     
  7. sevenacids

    sevenacids MDL Addicted

    Aug 17, 2012
    634
    188
    30
    I wonder why they didn't continue on the scheme to increment the build number by one for such releases like they did back in the Vista/7 days... Vista RTM: 6000, Vista SP1: 6001, Vista SP2: 6002 and 7 RTM: 7600, 7 SP1: 7601. Translated to Windows 8, 8.1 should rather have been 9201 and Update 1 maybe 9202. That always leads me to the question why they bumped up the minor version of the kernel for 8.1 from 6.2 to 6.3 as normally such a thing only happens when there are bigger changes to it but I don't believe there were any drastical changes made between 8 and 8.1 that justified that step. And, what's also confusing in 8.1 is that it doesn't reflect the fact that it's a new Windows release in the branding on System dialogs - it still "presents" itself as "Windows 8", it's only a little line that reports the actual version as being "Windows 8.1".

    It's all only numbers, sure, but sometimes it just feels like they make these all up along the way instead of having a consistent rule or scheme for versioning Windows. Seems that to be cool and "in", you obviously can't release service packs anymore, so what's been SP1 in the past is now replaced with an entirely new OS version release and what's been SP2 is an update to that new version... but wait... how exactly does it work now? Is Windows 8.1 Update 1 more like a SP1 to 8.1 or a SP2 if you count in 8.0? But wait... you cannot just install that "SP2" on Windows 8.0, you have to install "SP1" (8.1) first. You see how hard it became to categorize these things nowadays.

    It really smells like they didn't have any strategy after Windows 8 was released how they are going to deliver updates in the future and now are testing out different approaches, just to throw them aboard again with the next release anyways.
     
  8. tN0

    tN0 MDL Member

    Apr 17, 2011
    120
    25
    10
    Will MS offer Update 1 via Windows Update also for users of Windows 8.0?

    If you look at the recent market share numbers it is interesting that a lot of users haven't yet installed the free upgrade to 8.1 trough the Windows Store. Why? And maybe Microsoft plans Update 1 to solve this?
     
  9. fkar

    fkar MDL Recognized Advisor

    Nov 3, 2012
    1,577
    1,400
    60
    Update 1 is applicable for Windows 8.1 only.!
     
  10. FaiKee

    FaiKee Misinformation spreader

    Jul 24, 2009
    4,305
    5,803
    150
    MS is trying like hell to make people "realize" that "An Update is not a SP", so although they have the nos. 9601~9615 reserved for win8, they'd rather stick with 9600 - they don't want people to say "See? This is a SP, it even follows the SP numerology!" :D

    ATM I don't believe it's possible, win8->win8.1 path is thru Store, unless MS develops a KB for this path. :p
     
  11. abbodi1406

    abbodi1406 MDL KB0000001

    Feb 19, 2011
    8,089
    26,966
    270
    Win8 users will get updated edition of store upgrade .esd file

    don't know if it will be whole new build (like leaked 9600.16596) or RTM with update1 slipstreamed (like they use now in the current store upgrade)
     
  12. arseny92

    arseny92 MDL Secret Weapon

    Sep 22, 2009
    570
    1,260
    30
    #12 arseny92, Mar 3, 2014
    Last edited: Mar 3, 2014
    I referred that the delta/qfe adds up to 17xxx (from RTM 16384, I don't take the leaks here to account), so the above still apply.
    Win9 is developing in different branches, and builds numbers don't have to follow WinMain.

    To the view of a regular user, maybe, since MS just marketed 8.1 as being an "update" via Store while it is not. It is a system upgrade.

    It is beyond the fix engineering (KBs) because 8.1 was a system upgrade to 8.0 and updates for 9600 aren't applicable to 9200, because no applicability references.
     
  13. CODYQX4

    CODYQX4 MDL Developer

    Sep 4, 2009
    4,812
    45,366
    150
    I don't think MS will ever do a Windows Service Pack ever again.

    W7 needs one, instead they did update rollups. They have to extend support if they call it a "Service Pack", so they will call it anything else. 8.1 is a glorified SP claiming to be an OS upgrade, and Update 1 is like a mini-SP (or platform update).

    Not doing a SP means 8.0 takes a bullet to the head relatively quickly, even 8.1 possibly.

    MS better not blow their chance to come out with a new OS and call it Windows Cloud 9 :biggrin:
     
    Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
  14. s1ave77

    s1ave77 MDL Guide Dog/Dev

    Aug 15, 2012
    15,339
    21,343
    340
    Yep, would assume Win XP generated a lot of grey hair due to that ... :D.
     
    Stop hovering to collapse... Click to collapse... Hover to expand... Click to expand...
  15. sevenacids

    sevenacids MDL Addicted

    Aug 17, 2012
    634
    188
    30
    #15 sevenacids, Mar 3, 2014
    Last edited: Mar 3, 2014
    Right, but that leads one to ask (again): What is so bad about service packs? Okay, maybe the name is not appropriate anymore because now features are added (again, it has also happened "ages" :rolleyes: ago in the past as we know), but what if they just continued on that route and simply renamed them to something like "update packs", or simply "Windows 8 Update 1, Windows 8 Update 2, etc." like in Visual Studio? The fact that this didn't happen is kind of another "proof" that there seems to be no consistent update strategy for Windows in place right now... or maybe it just got started with Windows 8.1 Update 1, the MSU way... we'll know with Update 2, I guess (if there will be one).

    I'm not sure if the "average user" is really interested in this kind of numerology (more likely they are in astrology :D), and who gives a ... about a build number when the version tag (-> 8.1) doesn't match the kernel version number anyways. Long live Windows 6.3! (Okay, maybe full-screen Metro on non-touch not so long). The last Windows to reflect the kernel version was... NT4. Gee, I'm getting old.

    Gosh, with all these service packs/updates every few months... if you ask me, it kind of eliminates the need for a "major release" of Windows ever again. Just look at OS X. Or Visual Studio, for example: The 2013 version could also have been an update to the 2012 release. There are some nice features in it, some of them "major", but if you compare them side by side the changes are not that big. Plus: A lot of the stuff was also already available on 2012 via Power Tools. Of course, you cannot sell updates if you make them available for free... got it? ;) And that's legitimate up to certain level; I'd rather pay for a software product every two to three years instead of getting it for free and being nagged with ads... which could be the case for the mysterious Windows 8.1 with Bing SKU in the pipeline. Ad-ware Windows(?)
     
  16. burfadel

    burfadel MDL EXE>MSP/CAB

    Aug 19, 2009
    2,627
    3,843
    90
    Microsoft have changed the way they label and release updates, it seems they're still working out the best way of doing it. The reason for the change is a change in the way they market the OS to consumers.

    In terms of the actual base OS, this Windows progresses (excluding server OS)
    Vista --> 7 --> 8 --(8.1, 8.1 Update 1)

    Of course, there have been a lot of changes since Vista, but fundamentally it is based of Vista. Vista itself is a revolution OS, 7, 8, and 8.1 are evolution OS's. I hope that makes sense, but even Vista has it's basis in earlier OS's (in this case, Server 2003).

    I put Windows 8.1 in brackets because it really isn't a new OS per se, under the old marketing strategy it would have been labelled Windows 8 Service Pack 1. Likewise, Windows 8.1 Update 1 would have been labelled Windows 8 Service Pack 2. An obvious flaw in calling Windows 8.1 Update 1 as 'Update 1' is that it most definitely isn't the first update in the conventional sense of the term 'update', the conventional term covering security, general, and hotfix releases. It is also correctly 'update 1', but in a manner which is difficult to explain without creating too much confusion. They could have at least called it Service Pack 1. The term 'Update 1' also has more implication that there will be further 'updates', in the same sense as 'update 1', whereas Service Pack 1 doesn't necessarily imply that there are future updates, just that there is a possibility of one (instead of just calling it Windows 8. Service Pack). The version number of Windows is also purely up to what Microsoft decide to call it. In the old system, Windows 8.1 Update 1 (Service Pack 1) version number should be 6.3.9601.

    I should point out that although Windows 8.1 wasn't a new OS in the classical sense, it was more 'inclusive' than what you would typically find in a service pack, so in reality it was a service pack/quasi new-OS. From a marketing point of view the number change was justified, especially since people thought they were getting a new OS as a 'free upgrade!) to Windows 8. Microsoft don't charge for service packs :p.

    Update 1 also implies that there won't be new ISO's released, which of course I hope isn't true.
     
  17. Enigma256

    Enigma256 MDL Senior Member

    Jan 17, 2011
    354
    300
    10
    #17 Enigma256, Mar 4, 2014
    Last edited: Mar 4, 2014
    Because it was stupid to begin with. Sit down, young one, and have a history lesson:

    In the old days, the build number (the third group) was always whatever it happened to be when RTM was hit. They rounded it up to a "nice" number only for the consumer versions of Windows. So NT4, 2000, and 2003 were 1381, 2195, and 3790, respectively, while XP, being a consumer OS, was given a nice, pretty, sinister round number, 2600 (this is a number of very great significance to old-school hackers, as it was the frequency of the tone that can be used to hack the telephone system into giving a free long-distance call; in honor of this role, it also became the title of a well-known quarterly hacker magazine that discussed exploits, security, etc., which since this was back before the WWW even existed, was a very important resource for hackers... in a way, XP had a very ironic build number namesake... and yes, many years later, someone at Microsoft did eventually confirm in a TechNet article that 2600 was indeed in homage to its role in hacking and not just a weird coincidence).

    The OS will live with that build number for the rest of its life, and only increment the fourth group, the subbuild, even through SPs. So we never saw 2601, 2602, or 2603.

    With Vista, they decided, for some idiotic reason, to reserve the lower 4 bits of the build for the service pack. This caused problems for their round-number system, because if they wanted a number that looked round in base 10 and that also had the lower 4 bits zeroed out (i.e., the last digit, when written in hex, must be zero, so 2600, while round in base 10, is not round in hex because it's A28, or SP8 if the lower-4-bits system was applied to it), the smallest increment could only be 400. According to Raymond Chen (a Microsoft employee whose blog is famous in the Windows programming community), they initially planned 8888 for Windows 8, but someone forgot to tell the build number planner about the reserved lower 4 bits, so they couldn't use 8888 (8800 would've been fine, however), but by the time this unofficial policy was communicated, they had already incremented to 8888, and they couldn't go back down to 8800 or it'll mess up some of their internal version-tracking or something like that, so they had to go up to the next available increment, 9200. (This also means that, if they continue following this policy, Windows 9 will, at a minimum, be 10000, not 9xxx.)

    Now, the reason I assert that this is stupid is that "lower 4 bits" is not something that a human does because we don't look at the world in hexadecimal--it's something useful only in programming, and Microsoft has worked for years to get programmers to detect Windows versions the right way--that is, through API calls--instead of bad and will-potentially-break-in-the-future ways like scraping info. When there is already a perfectly good Win32 API to determine the service pack level, why the f*** did they think it would be a good idea to offer a programmers a shortcut to do it by a quick bitwise mask of the build number? That's just asking for long-term compatibility problems (it means that if they change this mechanism in the future, programs written to assume such a thing might break on a newer version of Windows that doesn't use it).

    Whoever got the bright idea to increment the build group for the SP obviously didn't consult with the group in charge of Windows compatibility (they're the ones in charge of making all those compatibility shims to patch up things like programs that misdetect Windows versions and the such) because I'm sure they would've gotten an earful otherwise. I, for one, am happy to see that 8.1U1 does not do this and am hopeful that this means the end of this silly practice.
     
  18. burfadel

    burfadel MDL EXE>MSP/CAB

    Aug 19, 2009
    2,627
    3,843
    90
    If versioning it 6.3.9601 breaks compatibility with programs expecting version 6.3.9600, then it's bad programming on the affected program's side, not the OS. In reality, the '9600' of Windows 8.1 and the '8400' of Windows 8, and so on, are arbitrary and superfluous, since this number doesn't denote the OS version, the 6.x (6.0 - Vista, 6.1 - Windows 7, 6.2 - Windows 8, 6.3 - Windows 8.1) does. Therefore, the only real function on the '9600' etc., from the point of view of the end user, is to signify the OS iteration. For instance, Windows 7 SP1 is a new iteration, it's a fresh start. You can't install Windows 7 updates on Windows 7 SP1 (not including the odd feature pack), the addition of '1' to this four number block is justified. If Microsoft don't change this number with update 1, what point does it serve? If programs want to check Windows version numbers for compatibility, they really only need to rely on the '6.3' etc, not whether the 9600 has changed or not. If there is the potential for the program to be affected by major changes to the OS (like a service or update pack), and therefore it does check this four digit block, wouldn't it therefore be necessary that Microsoft actually do change the number?

    Of course, I realise that the 9600 forms part of their testing number scheme, but that's not the point :). In a conventional numbering sense, going from 8400 --> 9600 is just what they chose to do, seeing as the real number of any consequence is the change from 6.2 --> 6.3. For Windows 9 (or 8.2, whatever the next iteration will be in this number system) there is no reason why they couldn't just start off at 0000 again, there really isn't any reason not too. Bad programming in third party applications isn't an excuse :).

    The next revolution OS (which Microsoft are supposedly working on) should get a new first number, if this is Windows 9 they will probably jump from the Vista based numbering system of 6.x to a Windows 9 numbering system 9.x (meaning no Windows 7.x, 8.x numbering etc). It would be silly to call this next gen Windows 7(.x)!
     
  19. Enigma256

    Enigma256 MDL Senior Member

    Jan 17, 2011
    354
    300
    10
    #19 Enigma256, Mar 5, 2014
    Last edited: Mar 5, 2014
    It's the other way around. That is, a program expecting to determine the SP by looking at the build number instead of doing it the right way by invoking the proper API to report the SP level. So it would be a program expecting to see 9601 when it only gets 9600.

    Many people take that attitude, that the OS should just let misbehaving programs out to dry. This is terribly naive.

    1) The end-user uses a new version of Windows and finds that their favorite program is broken. Who are they going to blame? "Stupid Microsoft and their stupid Windows, it broke Acme Corp's Super-Awesome Widget!" Just look around Internet forums! How many threads do you see where people blame the OS and how many threads do you see people say, "Acme Corp is run by incompetent fools, because their Super-Awesome Widget had shoddy programming that made bad assumptions and had an error that, previously hidden, has become uncovered by improvements in the new version of Windows. Shame on them!"

    2) Sometimes, waiting for the program to be fixed isn't possible. Maybe it's a paid program and the user doesn't want to pay for an upgrade. Maybe the new version is radically different and is not suitable. Maybe the program has been abandoned and nobody is maintaining it any more (goodness knows that's happened to me many times). Maybe the program is a critical custom line-of-business software that was developed in-house, and the employee who wrote it quit the company years ago and nobody on the current staff can figure it out.

    This is why there is a whole group at Microsoft dedicated to compatibility. There's a compatibility layer in Windows that, through the use of shims, can go as far as reenable buggy behavior for specific processes, and that's why every month or two, you'll see in Windows Update something that is said to "improve compatibility"--that's an update of this compatibility engine and/or the huge database of specific programs and the way in which each needs to be fixed.

    The other way is the Apple Way, which involves them giving everyone the finger and saying, "Screw it, I don't care if this breaks a whole bunch of things on the market, because we can do whatever the f*** we want, and, oh yea, our market share is also small so there aren't enough people to complain anyway." Frankly, I like the way Microsoft does things. (The compat. layer is separated from the rest of the OS, so it's not like they're putting these compatibility hacks into the core of the OS.)