Fog Creek Software
Discussion Board




Saving the API: Should Microsoft back WINE?

It seems to me that Microsoft could keep the Windows API ubiquitous by backing the WINE [1] project.  I'm sure that the project would leap forward if

If WINE was successful then the Win32 API could probably be ported to other architecture without too much trouble.

Windows software would then run on other architectures, but they would always have that 'not emulation' layer slowing things down.  A genuine windows platform will always perform better.

I think this would give the same results as Crigley's old suggestion about Microsoft building the next windows on Linux [2], but would take a much less radical move.

[1] http://www.winehq.com/
[2] http://www.pbs.org/cringely/pulpit/pulpit20030116.html

Ged Byrne
Friday, June 18, 2004



Part of the premise of the WINE Proejct is that it's a clean room implementation where they do not have to worry about *any* copyright/patent issues as none of their people have seen the internals of Windows (other than the BSOD).


If MS gets into it, they could put actual Windows developers on it, which, although it would speed things along, would risk the project ni the longterm as things could be "accidentally" copied into the source.

Look at what SCO is trying to do to Windows.  MS would be able to do the same.

KC
Friday, June 18, 2004

>>Look at what SCO is trying to do to Windows

You meant Linux, right?

Anyway, I think Microsoft would have to pushed very far before giving up on their lock in strategy.  Wine, they accept because its not really dangerous yet. Mono they accept as it currently probably brings more developers to .Net than away from it.

But when the path to MS becomes the path away from MS (as I'm sure will happen some day) Mono, Wine and others will find them self at the wrong end of some lawsuit.

Eric Debois
Friday, June 18, 2004

I have high hopes for WINE, but it's taking a long time to get ready. I think they have problems tracking a moving target.

bill@lingolanguage.com
Friday, June 18, 2004

Main misconception about the Windows API:
"I think they have problems tracking a moving target."

The target isn't moving.  I believe they are targetting Windows 2000 compatibiliy -- which is currently 5 years old.  New software still runs on Windows 2000 and will for some time -- perhaps even after Longhorn is released.

Almost Anonymous
Friday, June 18, 2004


Yeah, I meant Linux.  Brain and fingers weren't moving in unison.


They don't have much of a moving target at all (with a few exceptions).

Since MS is stuck making sure that Office 95 can run on XP along with Office XP, that's what the Wine crew does.

One of the nasty moving targets is DirectX compatibility.  It's a mess because many devices (video cards especially) have DirectX drivers/compatibility layers.

KC
Friday, June 18, 2004

Microsoft already has a reference implementation of the Wndows API... The Windows API itself.  Why would they support/use WINE?

And before you say portability or anything else, remember that Microsoft has supported chipsets other than x86 from time to time and from everything I've seen from the outside, I believe they have always kept their OS code mostly portable so that their company's fortunes don't completely depend upon the continued success of Intel.

Mr Fancypants
Friday, June 18, 2004

I'm not talking about Microsoft coding for or using Wine, but rather supporting it by making internal documentation, etc available.

If win32 applications could be made to run on other platforms, it would reduce the demand for cross platform development.

The performance of Wine applications, however, would probably still be sluggish compared to those run under Windows, which would be good for Microsoft.

Ged Byrne
Sunday, June 20, 2004

*  Recent Topics

*  Fog Creek Home