Will the project that port the original Start Menu from 9834

[Camwinny]

Will the project that port the original Start Menu from 9834 to 8.1 be carried out? It seems attractive.

[gtgamer468]

Haven't you heard about the Start9 project by Splitwirez (who is a member of this forum)? He's using a shell implement that will include features like the Start Menu from builds 9833 thru 9901 with working Live Tiles and a Sidebar from a Longhorn build. Currently, it's available but to use it, you will need to code. I haven't try doing it because I don't know a thing about coding but you want the apis, the site is start9.menu.

[Splitwirez]

Haven't you heard about the Start9 project by Splitwirez (who is a member of this forum)? He's using a shell implement that will include features like the Start Menu from builds 9833 thru 9901 with working Live Tiles and a Sidebar from a Longhorn build. Currently, it's available but to use it, you will need to code. I haven't try doing it because I don't know a thing about coding but you want the apis, the site is start9.menu.

...close.

Start9 is a modular shell, the idea being that you'd install chunks of shell (modules) for it. A Taskbar might be one Module, a Start Menu another, but there's no constraint that says you have to have that bar which has a button to open a menu on the left. It's freeform, so you could put together, for example, something along the lines of an iOS-style shell on your Windows computer. Why you'd do that in particular I don't know, but it serves as an example for how far I hope to allow people to deviate with Start9...albeit not a very good example. That's not to say you can't put together a more traditional Windows-like shell with it though, and indeed I expect that many users will do just that, hence the whole 98xx menu thing.

We're doing this with an API of sorts, which will allow developers to create modules of their own, which can work with other modules, even by other developers, as well as a separate toolkit to streamline the process of incorporating Windows shell functionality across different versions of the OS (everything 7+), including undocumented features. Other customization tools have all had to separately research Windows' inner workings to access a lot of this functionality, but the idea is that Start9 will allow all modules to share a common codebase for that stuff, which is updated separately as Windows itself is updated. This all means far less maintenance for module developers.

In any case, though, the project is far from complete. To say that it's "available, but requires coding knowledge" isn't quite accurate - the source code is available, yes, but I wouldn't call that "available for use"...more just "we're open to external contributions". We don't have any releases out yet. That having been said, we are working towards a Developer Preview release, although there is no planned release date, currently.

[Gustav Vasa]

I think that they have included it in RT 8.1 with update 3. But, why didn't Microsoft port it to regular 8.1 as well? I think it can be ported to regular 8.1 following the same process that Microsoft did with implementing that Start Menu in RT 8.1 update 3, but RT is ARM and some things might be different.

[Camwinny]

I think that they have included it in RT 8.1 with update 3. But, why didn't Microsoft port it to regular 8.1 as well? I think it can be ported to regular 8.1 following the same process that Microsoft did with implementing that Start Menu in RT 8.1 update 3, but RT is ARM and some things might be different.

I have the update file(.cab) KB3033055, and it didn't work.

[AlphaBeta]

I think that they have included it in RT 8.1 with update 3. But, why didn't Microsoft port it to regular 8.1 as well? I think it can be ported to regular 8.1 following the same process that Microsoft did with implementing that Start Menu in RT 8.1 update 3, but RT is ARM and some things might be different.

Remember that one of the heavily propagated new features of Windows 10 was the new Start menu. If they had readded a Start menu in a Windows 8.1 update, quite possibly it would have sabotaged their upgrade campaign.

Windows RT was already a dead platform back then, so they decided to port the 98xx start menu, knowing well that it wouldn't make any impact on Windows 10 upgrades at all.

I have the update file(.cab) KB3033055, and it didn't work.

I wonder why.

[awesomeguy]

This will probably seem extremely ignorant. Is there a possibility to extract the Windows RT update and find the corresponding files in a build w/o timebomb and port those files to Windows 8.1 x86/64?

[Wheatley]

This will probably seem extremely ignorant. Is there a possibility to extract the Windows RT update and find the corresponding files in a build w/o timebomb and port those files to Windows 8.1 x86/64?

Instead of writing that you'll seem ignorant, you could have just searched the internet a little bit to discover why Windows RT is special. It's simply not possible to replace x86 binaries with ARM ones.

[awesomeguy]

I know that, but I am saying that you could search for the x86 equivalent of the files contained in the Windows RT update in the 98xx builds with no timebomb.

[yourepicfailure]

Tried. Certificates forbid running 9834's executables under RTM 8.1.

[Wheatley]

I know that, but I am saying that you could search for the x86 equivalent of the files contained in the Windows RT update in the 98xx builds with no timebomb.

Oh, sorry, I misunderstood you. Even if you managed to bypass the certificate issues yourepicfailure mentioned and get explorer and all its dependencies ported over, you'd end up creating a very fragile frankenbuild, which just wouldn't be worth it.

[yourepicfailure]

To add on, 9834's explorer relies on some specific theme engine apis that are not present under 8.1. Mostly for the new start menu. You'd have to also port the theming engine as a whole from 9834 as well. Too in-depth with system files and at that point just running 9834 itself would be more stable if the frankenbuild ran at all.

[Splitwirez]

To add on, 9834's explorer relies on some specific theme engine apis that are not present under 8.1. Mostly for the new start menu. You'd have to also port the theming engine as a whole from 9834 as well. Too in-depth with system files and at that point just running 9834 itself would be more stable if the frankenbuild ran at all.

The new Start Menu plays just fine with (most) Visual Styles created for Windows 8.1 on any 98xx build, no modifications to the msstyles are needed. All of the msstyles-compliant regions of the 98xx menu use parts which either already existed in Windows 8.1, or are not necessary for the menu to function as intended. Does this completely disprove your point? No, far from it. But it does leave me curious about these "new theme engine APIs" you've mentioned...

EDIT: (btw this image predates Start9 by around a year, if not more, in case anyone was wondering about that)

[yourepicfailure]

I stand slightly corrected.
Let me elaborate. While the apis called have not changed, the apis themselves have. For instance, 9600.1 themes lack information needed for dwm dwm and explorer to call through uxtheme to create the start menu. This is evidenced when a computer restart is made with the 9600.1 theme applied. The start menu will not open.


However, switch back to the default aero, open the start menu (dwm loads the information to ram) it then works when you switch back to the 9600.1 theme. However it is black until you perform a color change. 9600.1 themes lack default color information to set for the new theme based colorization.


This is because under 9834, ImmersiveUI colorization(exception of start screen) is set by values provided by uxtheme from the msstyles, instead of a separate values set through settings or charms (9600.1). The reason it works after you set a color through personalizing is because new color information is generated that replaces the null value in the user profile generated due to lack of information.

In addition, uxtheme provides different information from the theme as well, such as for creation of superbar flyouts. Using 9600.1 theme breaks them.


These problems also arise when using 9600.1 default Aero.
Just in case, 9600.1 stands for Windows Server 2012 R2/Windows 8.1 With Update 1.

I know not if these are isolated to just 9834, but if one wants to use 9834 resources these issues must be accounted for.
Also, there are other apis required in other dlls other than the theming engine such as user32, userenv, ntdll, etc.

EDIT: Corrected slightly incorrect information and added some other detail.

[Splitwirez]

Interesting...I'd always wondered why the menu seems to break after restarting explorer until the default style is applied. Guess I was on the right track in deciding to re-implement it after all, lol...even if that isn't the main focus anymore.

I'd love to know, though...how did you figure all of this out?

[yourepicfailure]

Long story short, a tweaking adventure.

[Gustav Vasa]

Remember that one of the heavily propagated new features of Windows 10 was the new Start menu. If they had readded a Start menu in a Windows 8.1 update, quite possibly it would have sabotaged their upgrade campaign.

Windows RT was already a dead platform back then, so they decided to port the 98xx start menu, knowing well that it wouldn't make any impact on Windows 10 upgrades at all.

You are right, the propaganda about Windows 10 focused a lot on the Start menu, which is probably the reason why Windows 8.1 didn't get the Start Menu from the 98xx builds. Anyway I think that's a pity since the 98xx Start Menu fits perfectly in Windows 8.1, maybe someone in the future tweaks that Start Menu into regular Windows 8.1 even if it's a challenge as it's difficult, but it may be a "tweaking adventure" as someone has already said earlier. I've heard that there's a project called "Start9" which wants to do a similar thing but I'm not sure about it.

[yourepicfailure]

Actually the initial plan for the new start menu was to implement it as part of an "update 2" for 8.1/Server 2012R2. Instead, Microsoft later realized it would be more commercially and reputably viable by emphasizing it further as a new feature of a new version of windows. As such the "propaganda" was a core component of this plan. In the end this "update 2" only surfaced for RT.