Microsoft Discloses 14,000 Pages of Coding Secrets

OrochimaruVoldemort writes "In an unexpected move, Microsoft has disclosed 14,000 pages of coding secrets. According to The Register: 'This is Microsoft's latest effort to satisfy anti-trust concerns of the European Union, which is possibly a tougher adversary for the company than Google.' The article mentioned that this will be done in three phases. 'Between now and June it will garner feedback from the developer community. Then, at the end of June, Microsoft will publish the final versions of technical documentation — along with definitive patent licensing terms.' Lets just hope those terms are pro open source."

stupid summary

[moderatorrater]

Lets just hope those terms are pro open source

Come on, guys. There's no chance in hell that the licensing terms will be pro open source and we all know it. Can we please stop propagating false hope?

Damn right!

[Ungrounded Lightning]

Unexpected, as in they told us very loudly that they were going to do it?

Yep!

They've told us a LOT of nice stuff they're "going to do" that they turned around and either didn't do or poisoned.

Embrace, extend, extinguish.

I'll believe it when/if it's finally done. (And even then I'll wonder what "gotchas" are included.)

Re:bring on the virii

[Starrk]

Perhaps. Or perhaps it brings on suggestions from security experts that will prevent virii.

Why is parent flamebait?

[WindBourne]

MS has NEVER done anything yet that is pro open source. They have gone to great lengths to make sure that something has the appearance of such, but that it would not help. The only question should be, how far ahead is MS thinking? They have always been a pretty good chess player.

Admitting They're Lying is Reassuring?

[Doc Ruby]

So Microsoft finally releases a huge tome of secrets Microsoft uses to compete with other vendors on its closed system. After years of denying that, after years of keeping them secret from even the thousands of paying customers buying what they thought was equal access to the MS platform.

And somehow that admission that MS has been lying about something so central to protecting its anticompetitive abuses of its monopoly is supposed to reassure antitrust investigators?

Re:WINE

[TheRaven64]

Not really. With the exception of a few bits of Microsoft-written software, most Windows software is written against published APIs with the occasional work-around for bugs in the APIs. WINE 'just' needs to implement the already-public APIs (including replicating bugs) and code will work. The WINE team only need access to secret APIs if code has been written using them.

Re:Why is parent flamebait?

[The End Of Days]

I have to wonder, is the flamebait mod for damning Microsoft with faint praise, or not damning them enough?

Ummmm, no

[Sycraft-fu]

People said this same thing when the Windows 2000 source code leaked. Nothing happened. Multiple problems with that theory but one of the biggest is simply that it is wrong. Lots of people have the Windows source code. MS has a license where universities can get a copy for research. One university I know that does is ASU in Tempe, Arizona. So this idea that only MS has ever seen the code is false, thus the argument is invalid, never mind the other problems with it even if it weren't.

Re:WINE

[Tatsh]

I believe Wine, ReactOS, and MingW are using MSDN and "clean room reverse engineering" to develop (meaning a group writes documentation, another group implements). And they are well making sure that no code in the trees are taken from the leak of the Windows 2000 code a few years ago, and no code is written via direct reverse engineering Windows. This information MIGHT be helpful, but Microsoft is unpredictable when it comes to enforcing its patents and loves them. If I were on any of these teams, I would advise to stay away from this documentation until it is cleared with FSF that the licence is compatible with GPL (which I highly doubt it will be).

Well of course not

[Sycraft-fu]

Since "Pro open source" seems to mean "Can't cost anything, and can't put any restrictions on it other than requiring the code to be open." That is pretty much going to kill almost anything from being pro open source.

I imagine it'll be similar to MPEG-4 and such as it'll be an open standard with RAND licensing. What that means is anyone can get a copy of the standard and licensing to use it, and the price of that license will be reasonable and standard. However, that does mean you have to pay if you want to use it. I can't see them just wanting to give it away for free.

So if you are willing to adjust your definition of open source to accommodate things that are open standards, where it is open to all, but you do have to pay a license, then I imagine you'll be happy. However if you take the stance that it cannot cost any money, well then you are probably SOL.

Re:Unexpected?

[Compholio]

Well until now, we assumed it was just an idle treat.

I assume you meant an "idle threat", but what we got IS a treat. Now whenever someone claims that open source is not viable for business applications we can claim that even Microsoft supports open source.

Re:Oh come on now ...

[Missing_dc]

Please,

Improbibility is not required....

Think business. What better source to find your bugs than the many thousands of angry coders who are not M$ fanbois. Let your hatred consume you Luke, find the flaws in the code..... or rather "Your hatred, a tool, it is. Fix that which is broken, and glory you will find" /yoda voice

And you suckers ^h^h^h^h guys will do it for FREE!!

Re:Admitting They're Lying is Reassuring?

[Doc Ruby]

As far as I can tell, those protocols weren't documented for consuption by anyone outside Microsoft. Yet programmers inside Microsoft were able to use them to write software.

That does indeed make them as useful as "secret APIs" to programmers writing for Office/Exchange 2007.

Re:WINE

[Anonymous Coward]

> The WINE team only need access to secret APIs if code has been written using them.

And LOTS of such code has been written. Examples include Internet Explorer, Windows Media Player, etc. Why do you think Microsoft hides the source-code of these applications? It's not because of ANY of the many popular reasons. It's ONLY because these applications interact with the OS in many undocumented ways, i.e., in ways that can't be used by non-Microsoft applications (I avoid using the word "API" because it is the most misused word in MS related discussions). The result? No matter how competent Firefox developers are, IE works faster than Firefox or consumes lesser resources. Beginning to smell Microsoft's anti-competitive behavior? Imagine the huge penalties that Microsoft would have to pay if the sources are disclosed. In fact, ANY application that's linked with the libraries that ship with Microsoft's proprietary development tools is going to have SOME of these undocumented interactions too. That's the reason why DevC++ or GCC/Win can't create applications that are faster or feature-rich than the ones created using Microsoft's commercial tools. The Result? "Developers", as some these MS fanboys call themselves, choose Visual Studio and thus creating yet another source of revenue for MS.

In summary,

The functionalities of various windows applications are as follows:
Microsoft's In-House Applications (is superset of) Applications developed using Microsoft's commercial tools (is superset of) Other windows applications.

The speed (or performance) varies as follows:
Microsoft's In-House Applications > Applications developed using Microsoft's commercial tools > Other windows applications.

The difficulties for wine in running the application:
Microsoft's In-House Applications > Applications developed using Microsoft's commercial tools > Other windows applications.

Please please stop wasting resources fighting for Open Source. It's either far ahead of time or very stupid. Instead, fight for Open Standards, Open Specifications, Open Formats, Open Protocols, Open Interfaces. These are the immediate needs.

Re:WINE

[evilviper]

This information MIGHT be helpful, but Microsoft is unpredictable when it comes to enforcing its patents and loves them.

Whether you copy implementation details from a document or not has no bearing on patent rights.

Re:Why is parent flamebait?

[nametaken]

I'd say releasing 1 project under open source license is a "pro open source step".

But that aside, there are at least hundreds (thousands?) of examples of open source code available from MS. Many MS platform developers know this.

Now, that's not to say MS is what anyone would call an open source supporter, but it often benefits them to release tons of source code under very liberal licenses. You provide me with truly free framework for a particular kind of application, I'm more likely to accept your platform for development. That means anyone who wants to consume it has to use a closed product that makes them gobs of money. It doesn't make the original project any less open, though.

why not, if it's under patents?

[CarpetShark]

we never expected MS to disclose 14k pages of anything but contracts.

Without clarity over patents involved, those pages amount to a contract anyway. By using that stuff, you're signing the contract. The only thing is, you aren't getting to read the contract yet, until the patent issues are disclosed.

Re:Why is parent flamebait?

[Spliffster]

mods, these days, are probably on crack. i see tons of funny and even deeply sarcastic posts modded interesting or even insightful.

I wonder how and why these people get mod points.

Cheers,
-S

Re:bring on the virii

[The_reformant]

Im sure there are plenty of OSS apps which havent had thousands of security experts poring over them (presumably for fun). Outside the relatively few OSS apps that are used in the enterprise I would imagine code quality is worse since commits typically aren't buddy checked (buffer overflows ahoy!) and regression testing is a relatively new addition to the OSS scene.

Re:The problem is

[BruceCage]

Only requirement is you have to pay licensing. I see no problem.

First of all, definitions, definitions, definitions. It all depends on what definitions you use for "open standard" and "open source" (and "free software").

For open source one should be using the definition from the Open Source Initiative (OSI) [opensource.org] since it's a term used to indicate software that has been released under a software license compatible with the definition from the OSI. Note the very first criteria from the definition, "The license shall not restrict any party from selling or giving away the software as a component of an aggregate software distribution containing programs from several different sources. The license shall not require a royalty or other fee for such sale.".

The same goes for "free sofware" which uses the definition from the Free Software Foundation [gnu.org]. On that page it is explicitly stated that, "Thus, you should be free to redistribute copies, either with or without modifications, either gratis or charging a fee for distribution, to anyone anywhere. Being free to do these things means (among other things) that you do not have to ask or pay for permission."

For "open standard" one could look to the definition by the European Commission (IDABC programme) [eu.int], which most importantly includes: "The standard has been published and the standard specification document is available either freely or at a nominal charge. It must be permissible to all to copy, distribute and use it for no fee or at a nominal fee.".

It should be clear now why the bit about "licensing fees" (or royalties or whatever) is exactly the problem and would prohibit such software from being referred to as either open source or free software. Once I receive software or a specifications document I should be able to distribute it without asking or paying anyone for permission.

Note the difference between paying a one-time fee for receiving and paying fees on distribution. See also the article "Selling Free Software" [gnu.org].

Re:Why is parent flamebait?

[sohp]

Here's how I view that contribution. First, it's not code, it's CSS. Data.

More importantly, the contribution could be viewed as Microsoft trying to influence the W3C standards to its benefit. Of course IE already passes the 700 cases they release. Now if Microsoft can get W3C to adopt them, IE instantly is complaint, no extra work. Even if the test cases are sloppy, or worse, in their interpretation of the standard. It's not really that much different that Microsoft "contributing" OOXML to the document standards process. Just a backdoor way to get their implementation as the standard. It doesn't add anything of value to open source development.

Re:Why is parent flamebait?

[Anonymous Coward]

First, it's not code, it's CSS.

It's not programming, but it is code. Regardless, its nature is not important, the fact that it's licensed under open source terms and is a benefit to open source projects is.

Of course IE already passes the 700 cases they release.

Actually, released versions of Internet Explorer don't pass them all, not even close. Those testcases were produced as part of the work to make Internet Explorer 8 more conformant with the CSS specifications.

Now if Microsoft can get W3C to adopt them, IE instantly is complaint

You mean "compliant", and no, the W3C adopting them would not make Internet Explorer any more compliant than before.

no extra work.

You're ignoring the work that Microsoft did to pass those testcases in the first place.

It's not really that much different that Microsoft "contributing" OOXML to the document standards process. Just a backdoor way to get their implementation as the standard.

Either you are trolling, or you don't know what you are talking about. Testcases don't change what's in the specifications and can't make any implementation the standard.

It doesn't add anything of value to open source development.

That'll be why the announcement [w3.org] on the W3C CSS test suite mailing list garnered a positive reaction from developers for all the major browsers then? (Quotes: "Awesome guys!", "Great to see them published!", "your contribution of this test suite is very encouraging.", "Thanks for the CSS 2.1 tests."). Testcases are valuable tools for QA.

And, most pertinently, there was an ongoing thread on the list about relicensing the existing testcases under the BSD license before Microsoft's announcement because the usual W3C license was a bit too restrictive for some people, open source projects in particular.

So when open source developers, who explicitly require testcases under an open-source license received news of Microsoft's contribution, they were pleased. And you really want to argue that Microsoft have never added anything of value to open source? Really? Then why do Mozilla and Webkit developers disagree with you?