Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Will a Microsoft Breakup Help WINE?

By rusty in News
Mon Jun 12, 2000 at 05:56:53 AM EST
Tags: Interviews (all tags)
Interviews

Just so that nobody is disappointed after reading that title, I should make it clear right now that the answer to that question, according to Douglas Ridgeway, WineHQ maintainer and general WINE PR hacker, is: probably not.

I recently spoke with Ridgeway via email about the imminent MS breakup, and whether he thought the ruling would have any effect on the WINE project's goal of re-implementing the Windows API for Linux and Unix-like platforms, to allow native Windows software to run unchanged on *nix. His opinion was that the easiest means of decoding MS APIs would continue to be cleanroom reverse-engineering, and that assuming a Microsoft OS company was eventually forced to share API specs openly with the world, "if more info becomes available, we'll use it, but I don't think it has a big impact on the people writing code."


The following is a combined, and very slightly edited, transcript, in Q&A format, of my emails with Ridgeway. My questions appear in italics, his answers are normal text.

Judge Jackson has ruled that MS is a monopoly, and proposed a two-part breakup where Microsoft will be split into an OS portion and an Applications portion. The terms of the ruling include the stipulation that the OS company cannot share any technical information with the App company that it does not make simultaneously available to anyone else under the same terms.

How does the WINE project feel about this ruling, in general? Do you think it was deserved, and will acheive it's goal of promoting competition in the software market?

I'm not sure that I can speak "for the project", which is, after all, just a group of people with some shared ideas about code. From what I understand about the case, it seems clear that Microsoft has made a practice of engaging in behavior which is illegal under the Sherman act.

I don't know whether or not the ruling, if/when implemented, will promote competition. The argument that it will simply create two monopolies holds some weight with me. However, it seems like it may help prevent Microsoft from illegally extending its monopoly into other areas.

More specifically to WINE, do you believe it will help your project achieve it's ultimate plans of fully re-implementing the Windows APIs for Unix?

If and when the breakup does occur, will WINE license or otherwise attempt to obtain the same API documentation that the Windows Application company will need?

If more info becomes available, we'll use it, but I don't think it has a big impact on the people writing code. Think about it: suppose there's a call you don't understand. Right now, your options are to read the available docs and do some experimentation and reverse engineering. After the ruling is implemented, you might, at best, have the additional option of filing a legal challenge in order to get access to some mythical "internal documentation" or the source. I suspect that even if this becomes possible, the former options will still be easiest for most of the developers most of the time.

Let me give you an example. I've been offered the source for Windows 95 multiple times, both in original and reverse engineered form. I've turned it down. I don't see any use for the original source in any coding that I've done, and other coders I've talked to feel similarly.

If you will seek the API documentation, will this change the direction or focus of WINE in any way? Is there anything that the WINE developers really want to do, but thought they'd never be able to implement because of the previously non-existent chances of getting detailed and accurate API docs from Microsoft?

No. Secret APIs in Windows were mostly a Windows 3.x thing, and went out after the 1995 consent decree. Microsoft's strategy since then to protect the Windows franchise has been quantity, not secrecy: add new APIs, and blackmail developers into using them. (ActiveMedia, eg.)

What I thought might most affect the WINE project was the part of the ruling that the OS nanosoft can't make API docs available to the apps nanosoft without making them simultaneously available to anyone else, under the same terms. I understand that the biggest challenge for WINE has been the "moving target" problem-- do you think the "open spec" condition will slow the pace of development of new APIs, since it will no longer be so easy for MS to rapidly develop APIs and implement applications on them in-house, before making any information available to anyone else?

I think that it will slow the "moving target". This is in fact the key point -- it's precisely what Microsoft execs are referring to when they speak about the "freedom to innovate". So it's recognized that this is where the true anticompetitive aspects of the Windows franchise lie, and this is the most important thing for Microsoft. Note that it's not really a question of when / how documentation is made public. Users won't buy a new version of Windows because it supports a new API, they care about running apps. So you can only force an upgrade if you convince application developers to require the new API. Unless the API provides truly new and unique functionality (something Turing showed impossible early last century) application developers have no incentive to use the new API, because it would reduce their market by requiring users to upgrade. So there's a problem introducing the new API. Microsoft solves this by controlling both the OS and major applications: they can use their new API in their own apps, using their control of both to maintain their monopoly ("freedom to innovate"). Publicizing the documentation is irrelevant to this dynamic, and is necessary to get third parties to use the API anyway. (Other, less straightforward mechanisms are required to blackmail third party developers into adopting new APIs.)

Also, if the OS nanosoft has to openly share information with the App nanosoft, will this assist WINE in it's reverse engineering efforts? My thinking is, OS nanosoft develops new API. They must make either the docs or actual working binaries available openly, before the App nanosoft can develop to the new API, so WINE developers could start reverse angineering and re-implementing the new API at the same time as the Applications nanosoft begins developing for it, thus potentially cutting the lag time between a new MS API and the WINE implementation of it. Does this situation seem likely to you at all?

Again, I don't think it's a matter of documentation, it's a matter of whether Microsoft can force adoption. Unused APIs don't matter. Microsoft knows that a fixed target, no matter how complex, badly designed and documented, can be cloned. They must maintain the pace of change in order to maintain their position, no matter what the social costs are.

Do you think this decision will change or accelerate your timeline for having a more complete and stable product? Will we be seeing MS Office on Linux in the near future?

No, and no, unless the Office Nanosoft decides to do a port.

Are there any other comments you'd like to make about WINE, the breakup, or, really, anything at all? :-)

It'll be interesting to see what comes out of the antitrust effort. But I really see it as an entirely orthogonal effort to Wine.

I, and the Kuro5hin readers, very much appreciate your time answering these questions, and we wish you and the whole WINE team the best of luck!

Thank you.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Related Links
o Kuro5hin
o Douglas Ridgeway
o WineHQ
o Also by rusty


Display: Sort:
Will a Microsoft Breakup Help WINE? | 12 comments (12 topical, editorial, 0 hidden)
One of the useful things which may ... (none / 0) (#4)
by Bradley on Mon Jun 12, 2000 at 02:02:39 AM EST

Bradley voted 1 on this story.

One of the useful things which may come out of the antitrust stuff is patents. There were a couple of posts (and followups) on Slashdot were along the lines that MS using patents to prevent other implementations may be invalid now. (cf VirtualDub, etc) I don't know if this is true or not, but it would be interesting. What other patents do MS have that they've been sending cease-and-desist around for?

As for WINE though, I've found that windows things are documented reasonably well (at least the bit I'm working on). Even if the courts end up giving access to windows source (and in places outside the US), I suspect that looking at the source code for Windows would tend to disqualify you from creating a windows clone, in the same way the the free java developers don't want anyone who has looked at the Sun code.

Posting at +1 only due to the fact ... (none / 0) (#1)
by Fish on Mon Jun 12, 2000 at 02:06:15 AM EST

Fish voted 1 on this story.

Posting at +1 only due to the fact that Rusty has put a lot of effort into this article. The topic doesn't really interest me.

How could a breakup help Wine? Win... (none / 0) (#6)
by PresJPolk on Mon Jun 12, 2000 at 02:46:58 AM EST

PresJPolk voted 1 on this story.

How could a breakup help Wine? Wine's goal is to write a 100% compatible implementation of the Windows API. The hypothetical Windows Baby MS would no more want to help Wine, than the current Microsoft does.

Disclosed APIs are only the first step. Wine works to be compatible with the implementation, not the specification. That includes duplicating all the bugs in Windows.



Re: How could a breakup help Wine? Win... (none / 0) (#7)
by AArthur on Mon Jun 12, 2000 at 06:55:10 AM EST

<p>But I wouldn't be surpised if the Microsoft Applications company wouldn't want to help support WINE, as it make it easier for them to increase there sales (especially if the Linux x86 platform takes off). Obviously Microsoft Windows company wouldn't want to do this -- but they are forbidden now to have hidden APIs.

Andrew B. Arthur | aarthur@imaclinux.net | http://hvcc.edu/~aa310264
[ Parent ]

Re: How could a breakup help Wine? Win... (none / 0) (#12)
by PresJPolk on Mon Jun 12, 2000 at 09:19:45 AM EST

Yeah, but if the split happens, then the MS Apps company would be in no better position to help Wine, than any other Windows software company.

Though, given how much spare cash Microsoft always seems to have lying around, a sudden infusion from the MS Apps company to Wine probably wouldn't hurt. :-)

[ Parent ]
Why is it we keep hearing about the... (none / 0) (#3)
by soulhuntre on Mon Jun 12, 2000 at 03:20:03 AM EST

soulhuntre voted 1 on this story.

Why is it we keep hearing about the "secret" API issue when event he WINE developers say it is bogus?

Re: Why is it we keep hearing about the... (none / 0) (#8)
by Imperator on Mon Jun 12, 2000 at 06:56:29 AM EST

If Microsoft were to publish their APIs at the same time they distributed them internally, with all the same supplementary information (e.g. "this function doesn't actually work") they tell their own developers, it would be much easier for applications to keep up with upgrades of Windows. (Not that this would avoid the constant upgrade issue, but it would decrease the relative advantage of MS apps.)

There are still some secret APIs. MS apps use them to tie directly in to the kernel and gain significant performance advantages, or to do things they don't want any competing product to do (esp. under NT). But they aren't as much of a bargaining chip for them as they were when they exchanged APIs for favorable as a way to control the market. You can write Win32 apps without any of the secret APIs.

[ Parent ]

Re: Why is it we keep hearing about the... (4.00 / 1) (#9)
by Anonymous Hero on Mon Jun 12, 2000 at 06:58:04 AM EST

It's mainly an historical thing that's grown to become an urbane legend.

Back in the days of MS-DOS, people were finding undocumented BIOS and system calls all over the place.

That continued with Win 3.0 - 3.1. It was very obvious to anyone who actually had to write code for those versiona at the API level that M$ just hadn't documented an *aweful* lot of it.

There was also the whole fiasco over DR-DOS at that time. Basically, it was discovered that Win3.1 would generate some really bogus messages if it was installed on a machine running DR-DOS instead of MS-DOS.

So basically, a lot of people just assume that Win95 and higher contains undocumented API's.

Whether or not it does, I don't know but if it doesn't, it's simply because M$ has become cautious about being caught out on that one before and didn't want the anti-trust problems that would have gone with it.

Hence, the adoption of the 'moving target' strategy.

You might be strangling my chicken, but you don't want to know what I'm doing to your hampster.



[ Parent ]

Let me give you an example. I've be... (3.00 / 1) (#5)
by Marcin on Mon Jun 12, 2000 at 03:53:34 AM EST

Marcin voted 1 on this story.

Let me give you an example. I've been offered the source for Windows 95 multiple times, both in original and reverse engineered form. I've turned it down. I don't see any use for the original source in any coding that I've done, and other coders I've talked to feel similarly.

I find that pretty interesting. I wonder if it was offered by disgruntled Microsoft employees or what? As to not needing it, I can see why not.. they don't need to see why MS did something the way they did it, they just need to know what it does so they can redo it 'properly'.

Very interesting interview.
M.

Re: Let me give you an example. I've be... (none / 0) (#10)
by Imperator on Mon Jun 12, 2000 at 07:01:45 AM EST

I doubt they wouldn't mind having the source when they're stuck; it would certainly speed things up in many ways. But they're careful to avoid it because they need to make sure that WINE is developed without any knowledge Microsoft might claim was improperly used or obtained. When Microsoft publishes an API, it's often under terms which prevent it from being reimplemented. If WINE were contaminated by code from one of these documents under the Monopoly Private License, the legality of the whole project would be thrown into question.

[ Parent ]
Re: Let me give you an example. I've be... (none / 0) (#11)
by Marcin on Mon Jun 12, 2000 at 08:55:23 AM EST

On a vaguely unrelated topic, it'd be interesting to find out what Microsoft have deep in their research labs.. like maybe working versions of IE5 for Linux or things like that.

Like someone I remember saying on IRC somewhere that there is a VMS for x86 somewhere deep within the bowels of whoever makes VMS..
M.
[ Parent ]

Wasn't the answer in the second lin... (none / 0) (#2)
by davidu on Mon Jun 12, 2000 at 04:05:14 AM EST

davidu voted -1 on this story.

Wasn't the answer in the second line? I like your questions, but...in the end, does it matter? It seems WINE has been fine so far. They will continue to be fine.

Will a Microsoft Breakup Help WINE? | 12 comments (12 topical, 0 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!