Microsoft Expands Access to Windows Source Code

Brain Stew writes "According to eWeek, MVPs living within thecountries that have signed up with Microsoft's Windows Source code program can now see it for free (limited source code of course). 'Microsoft Corp. has expanded the Source Licensing Program under which its Most Valued Professionals get access to the source code for the Windows operating system. The Redmond, Wash., company said on Monday that all the MVPs within the Microsoft platforms community and living within the 27 eligible countries worldwide will now be able to access Windows source code at no cost. '"

Comments

Re:This is somewhat good

[irokitt]

From what I've seen of OSS, the solution is not many, many eyes, but a core handful of experienced eyes that have experience and training.

Point needs to be made, however, that these guys who get free access are not here to "fix Windows" as much as they are there to write applications that require close cooperation with the OS (think antivirus or DRM applications). So the chances of them finding a bug and fixing it are slim, because they won't be looking for them.

Compile It?

[John Hasler]

Do they get to compile it and run it, or do they have to take Microsoft's word that the binaries that they are running were built from the source that they are seeing?

actual source?

[Doc Ruby]

If developers want to be sure that the "source" code they're looking at is the actual code their own app calls, can they compile it and link it to the still-secret Microsoft code, then run the whole thing in a debugger?

It's an open secret that Microsoft's own apps, notably SQL-Server, call a "secret Windows API" that isn't documented. That API is said to be faster to code for (time to market) and execute at runtime (performance), giving Microsoft apps advantages in competing with their rivals. Is there a way to use this new code access to discover whether Microsoft apps are calling a "shadow" API, rather than the code made public?

Re:actual source?

[M.C. Hampster]

It's an open secret that Microsoft's own apps, notably SQL-Server, call a "secret Windows API" that isn't documented. That API is said to be faster to code for (time to market) and execute at runtime (performance), giving Microsoft apps advantages in competing with their rivals.

Oh really? Do you have some citation for us? How was this "secret" API call discovered since people don't have the source code to SQL Server. And what exactly does this secret API perform? It must be some sort of duplication of some existing API in order for it to be "faster to code for" and execute faster at runtime, right? Where is this "secret" API located? Which library is it linked in to?

This talk about secret API's in Windows to make MS programs work better is such FUD. Has it been done in the past? Perhaps. But I'd like to see some sort of proof that current MS applications, like Office and SQL Server use some sort of mythical secret API.

Will it do any good...

[Beast in Black]

While at first glance this may seem like a good thing (at least for MS), i'm left wondering whether it will actually do any good...the MVPs who gain this access seem to be part of a rather closed community, being voted to their status by a bunch of people from other peer microsoft communities/groups.

Take a look at the MVP FAQ: http://mvp.support.microsoft.com/default.aspx?scid =fh;EN-US;mvpfaqs

Although some might say that sharing the code out, even among a closed community, might conceivably lead to improvements, from MS's track record with their multifarious products (some of which had oodles of people involved), there is no palpable confidence (at least in my mind) that it will get any better.

And in any case, even though microsoft shares the code out to the MVPs, there is nothing in the article that states that the MVPs will be allowed to modify the code...rather, the article explicitly states that they will "help" the developers. So even if some sagacious MVP does somehow manage to make a tiny improvement (unlikely, i know, but let's just suppose it for the sake of argument), wanna bet that he'd probably have to move heaven and earth to get someone who counts at MS to recognize this?

Also, as someon posted earlier, there is a good chance of the code getting leaked, even if MS uses the strongarm tactics that it is capable of to get the leaks plugged as fast as possible. What would happen then would be anyone's guess...

Anyway, here a link to the Windows 2000 source code http://www.albinoblacksheep.com/text/source.php. (if it's been already posted elsewhere on this site, beg pardon, i did indeed search, so my search skills are lacking...)

Wow, just as i hit SUBMIT earlier slashdot went down...is the big M already guuning for /.?

They want the source out as far as possible

[Anonymous Coward]

As someone who works for MS in their Mac business unit (that's why I'm anon). I'd like to point out that they're gradually increasing the exposure to selected parts of the Windows source code to people in other departments - us in the Mac unit have had access to the source code for 4 months now, but there's not enough code to compile a working system so it is only of use if you're interested in how certain sections of the OS work (if they happen to be released). So, even as an MS employee we can't get our hands on the whole thing.

Because the source code is not complete and Windows is implemented a lot differently to Linux and MacOS X then some of us in the Mac business unit believe that they'd not care if the source was leaked - in fact a couple of us (me not included) think they actually want this to happen. In a way it makes sense - I mean if a wine developer seen some of the Windows source code (or even *suspected* of seeing it) then MS, in theory, could tie them up for ages with legal action. Personally, I think they'd have to be a lot more desperate to do that as it'd generate a lot of bad publicity for them. So I don't think they'd sue just develop a lot of negative spin around the fact open source people steal other peoples code and ideas.

So, to be safe, anyone who doesn't work at MS should resist the temptation to look at the code even if you're doing so legally. Of course, it's easy for them to point their fingers at open source contributors, but it's harder to track down stolen code in closed source software. I can't say if any GPL code theft goes on at MS (officially we're all warned against it and us Mac developers pride ourselves on writing good quality original code), but it'd be so easy for a lazy programmer to steal some code from Mozilla or Apache and of course we're all free to persue the open code to get ideas from.

Speaking of web browsers we used to have the best web browser for MacOS at one time, until management killed the project (officially the rendering engine is in maintenence mode to support MSN for MacOSX - but there's been little improvments). Personally I use Camino but most in my unit use Safari. Of the people outside my unit most use Firefox under Windows, there's not that many people keen on IE so Firefox has taken a hold here, there's still many people who still use IE here because of loyalty to their employer but we're not officially banned from using alternative browsers so many of us do.

I've actually met a few of the WinIE developers, don't blame them for the stagnant product, until Firefox hit the radar then most of the team were placed on alternative projects. Personally I think they've got their work cut out, IE needs a total rewrite, its last major rewrite was for version 4 - with some of the code dating back even further (check the about box if you run windows).

Re:They want the source out as far as possible

[DickBreath]

Most of Slashdot is so negative toward Microsoft

You don't suppose there could be a reason for that? Or is everyone who disliked Microsoft just a fanatical zealot?


A friend of mine from high school....at Microsoft who is smart, honest, and writes brilliant code.

Obviously, Microsoft has some very bright people. They can afford to with all that ill gotten booty. Nobody disputes this. In fact it is so well known as to be not worth mentioning -- especially on Slashdot -- that MS has lots of brilliant people.


it's interesting to read what's actually going on inside Microsoft.

I for one would love to hear more about why things are done certain ways, or what the development culture is like at Microsoft.

I agree it is always interesting. Nonetheless, this does not change their business practices -- the reason that so many people don't like them.


Windows XP is a pretty good piece of software... things are improving.

....and I have noticed over time how much better the products have become.

Many of their products today are outstanding. Again, buckets of money to pour into development. (i.e. what economists call "monopoly rents".)


I run linux, but I run Windows at work

Me too.


overall the usability and stability of Microsoft products has noticably improved.

Give them enough times and they'll get it right. Just like with Word 4. But let me ask this. How many real businesses could afford multiple product release screwups or even disasters until they figure out how to get it right. Many ordinary developers might not even survive until a release 2.0. But look at how many tries it took Microsoft on some of their products.

This gets back to business practices and monopoly money.

In short, maybe there is an actual reason that so many on Slashdot dislike Microsoft. And do you even suppose that this phenomena is unique to Slashdot? (Hint: it isn't.)

But maybe I'm wrong. Maybe it is just completely irrational that people dislike Microsoft. They're just jealous. (But then why would the same people not be jealous about, say GE which is huge? Maybe because of business practices?)

Maybe anyone who dislikes Microsoft is just a raving lunatic. Nobody could possibly have a rational basis to dislike poor Microsoft?

We should all just bend over and take it.

There is a reason why the next release is called Longhorn. Because with all the DRM, you're really, really going to get screwed this time!

Am I too paranoid?

[bahamat]

I sort of get the feeling that Microsoft is releasing source code to some so that they could eventually attempt to do what SCO is doing. UNIX may have no trade secrets left, but Windows certainly does.

Re:Shared source will not work for MS

[PingXao]

I agree to a point. IBM does not fully embrace open source, however. They sell plenty of products that do not include access to the source code. If you run one of their mainframe OSes like VM or MVS or zOS (or whatever they're calling them these days) you pay dearly for the privilege of having access to the source code. About 10 years ago - granted, a long time - the company I was working for was paying upwards of $50,000 per year just to get access to the source code. This was above and beyond the normal charges just to license and run the things.

More recently - well, same time frame actually - OS/2 had a killer desktop: the Workplace Shell. It was totally object-oriented. AFAICT Windows 2000, NT, XP, Longhorn, etc. use completely object-oriented desktop models. People have been pleading with IBM for years to Open Source the WPS. 10 years later it would still be an improvement over the Windows offerings. IBM refuses.

Bye-bye, documentation...

[dpbsmith]

One of the few good things about closed source is that if a vendor wants third-party developers to write to its closed-source OS, it is beneficial to the vendor to create and make available well-written, accurate, complete, up-to-date documentation.

Admittedly Microsoft's documentation for developers has been going downhill lately, along with almost everyone else's. The physical volumes became CD's which became help-system files which became scraps of sample code. In order to develop to the Windows SCSI API, it is necessary to use guesswork, intuition, trial-and-error, and the assistance of the Windows community's "tribal knowledge." The PC community has long been used to using magazine articles and "Undocumented WIndows" books as sources of information.

But it is now about to get worse. I potentially foresee a situation where favored developers have access to source code, and documentation will decline to the point where it is difficult or impossible for non-favored developers to work in any development environment but VB.

In the Apple world, documentation was absolutely superb from about 1983 to about 2000 and underwent a precipitous decline with the advent of Darwin-based OS X. (A noticeable portion of the official documentation seems to have been generated automatically from header files!) I don't think this is a coincidence.

Re:This is somewhat good

[Anonymous Coward]

-- the solution is not many, many eyes--

it is, if we organize those many eyes properly. We could do it like SETI@home does it literally. we can make a program that takes half a page or a full page of code, of whatever program we happen to be working on at that time, and mails it out to many people. We tell them to find any errors in the code we mailed them. This way they are only focused on ONLY that page mailed to them, not the 250.000 lines of code before it and not the 250.000 that comes after it. EVERY page will be looked at by MANY people eventually.

Re:Shared source will not work for MS

[dasmegabyte]

that microsoft merely wants to go through the motions of open source, while not being open at all.

Fucking duh. Microsoft doesn't want to be "OPEN" source any more than we Microsoft developers want to write the OS or programs we pay for. But they do want to benefit from getting eyeballs into their code, allowing for peace of mind an enhanced security. And we want to benefit from being able to finally figure out what's going on in Function X that's always been a bit twitchy.

If Microsoft can satisfy the security and customizability needs of their customers with Shared Source, then Shared Source WILL "work" for Microsoft. Not all software has to be Open Source, you know. There's this thing called "choice" which is quite popular with purchasers, and some of them may "choose" to go with a controlled source solution.

Re:MVPs? Odd choice.

[pointbeing]

To be an MVP, from what I could see on Usenet, you don't have to do much else than lick Microsoft's boots in every single on-line comment you ever make (try to find an MVP who thinks product activation is a bad idea, for instance, or who disputes the Microsoft interpretation of an EULA), and demonstrate that you have competent skill (and even that I'm not so sure about, from when I still read the microsoft.public.* groups).

Why wouldn't this be proferred to those with MCSE credentials or MSDN subscriptions -- y'know, those guys who pay to know all this stuff?

I'm a Microsoft MVP - I'm also an MCSE and have an MSDN Universal subscription ;-)

I've questioned WPA on Microsoft's campus in front of 600 other MVPs. My specific question was since WPA pretty much guarantees paid licenses and since the price of piracy was *already* built into Windows I asked whether MS was going to lower the price of Windows since sales on previously pirated copies of Windows were pure profit - development, marketing and distribution costs were factored into the price of Windows before WinXP ever came out.

I'm not going to dispute their interpretation of the EULA because they wrote the software and can license it any way they like.

And you're right - the quality of MVP answers varies widely, but if the MVP has any sense at all he'll STFU about stuff where his skills aren't strong - I know I've been corrected by both MVPs and newsgroup users a couple times in the past two or three years.

Oh, geez

[rd_syringe]

Yes, Microsoft has created an initiative whose sole purpose is to hopefully inject proprietary code into an OSS project somewhere down the line in order to sue them.

And people wonder why Slashdot is considered a poor source of fringe journalism.

redalert !! redalert!!

[Anonymous Coward]

If you look at their code you open yourself to litigation.

they will say anything you code in the future you got the idea from looking at their code.

so beware and be very afraid - unless you have a lawyer as a friend.

If you want to see how M$ code can be, just look..

[Assmasher]

...at MFC, the source shipped with Visual Studio for years.

Ugly, ugly, ridiculously poor code documentation (if any), odd workarounds left-over from the days of WIN32s, 16-bit thunking days, et cetera...

Be GLAD that you don't have access to the Windows source, god forbid anyone should code that way (of course, there's plenty of Open Source Software that is just as ugly or worse ;).) Sometimes end user features bury the desire for elegant (or even functional sometimes) architecture...

Re:IBM understands quite a bit

[realSpiderman]

Au contraire, just look at Eclipse. This was IBM VisualAge for Jave. The OpenSourced it, and now it is taking HUGE steps.

There's also JFS, OpenAFS and a lot of other small things.
Sure, they will never open source DB2 or other multi million dollar middleware, but would you hope for an OSS Oracle???