An open letter to Bill Gates and Steve Ballmer Letter
By newsdesk@theinquirer.net
The Letterman: Saturday 25 October 2003, 15:16
Messrs William Gates Jnr & Steven Ballmer Microsoft Corporation One Microsoft Way Redmond Washington, USA
Dear Sirs
I see you have been active again in making interesting and to some degree highly amusing statements about Free/Libre Open Source Software and the many and varied people who make up its community. I in particular would like to thank Mr. Steven Ballmer for your entertaining exposee of Linux's deepest, darkest secret - that it can seriously worry the senior executive of a convicted predatory monopoly, without that ever having been the intention of its principal software designer and initial developer. I would also like to thank you for humming and hawing around the question of the release of source code to people who can use it, in the light of the new MVP source code entitlement program. Well, are they deserving members of the Windows development team or not? In relation to your comments, Steve Ballmer, on Linux's "road map", I will refrain from expounding on Linus Torvalds' comment on the cover of one of Bill Gates' books, showing him standing in the middle of an empty road. It's not nice to make jokes like that, is it, Your Billness? Road kill is no joke, even if some enterprising chef has written a book about it. No, I have something else on my mind, something much more worthy. I would like to challenge you to a software coding bake-out, a bet to see which methodology works, and which doesn't. You have made some progress with your NT source tree, anyone can see that - Windows 2k3 is a more serious product than Windows XP, and definitely a more realistic - and much more massive - product than Windows 95. Congratulations. You have also declared that Windows 95, Windows NT 3.x and NT 4.x are discontinued, end-of-line, unsupported products. And Windows 98 is shortly going to be in the same category, having already been discontinued. And Microsoft is attempting to roll the Win9x features into the NT line. XP is the nearest you have come to success. In the process, Windows users have enjoyed an interesting remote use of RPC and other features that might otherwise bug you. And in the process you have put back Longhorn's release date.
My challenge is this - release the entire range of discontinued, end-of-line and unsupported Operating Systems mentioned above (Win9x, NT 3.x and Win4.x) and their related utilities and Productivity Applications, as Open Source under the BSD/MIT license, since you have stated at sundry times and in diverse manners that that license is one you can live with. You are of course expected to sanitise the source trees - we don't want trouble with absurd IP cases. Release the sanitised source trees, minus any bits and pieces of third-party encumbered code Microsoft may have in the Win9x and NT 3.x and 4.x source trees, to the ftp servers at the MIT, ibiblio, the U of Calif. at Berkeley, and the U of Cambridge, UK, with prominent notices stating that they are released under the terms of the BSD/MIT licenses placed in slashdot.org, newsforge.com, computerworld.com, news.com.com, www.theinquirer.net and www.theregister.co.uk and other industry news outlets.
My bet is that in the time it takes Microsoft to come up with a half-way decent Windows product, the Open Source development process starting from an earlier, identical initial source tree without constraints will produce one better. The length of time is going to be the same. On one side you have the multi-billion dollar transnational corporation, on the other you have an amorphous world-wide community. One has a head start, but the code bases for this challenge are the same. The only catch - Microsoft is not allowed to use the source code produced by the open source effort until after it has rolled out Longhorn - thus preserving the independence of the challengers, who will not have access to the Longhorn source tree. After the challenge has finished and the bets have been tallied up, then it is a totally different story, because the BSD/MIT license doesn't prohibit incorporation within a closed-source code base, only the denial of attribution. But should Microsoft use the independent effort's code during such a challenge, it would be an admission that the Free/Libre Open Source community is right, and must be met with an appropriate forfeit - the sanitising and opening of the Longhorn source tree. I propose in the interim that the challenge in the interim be named something other than Windows or Office - precisely what will have to be decided upon later.
So, there you have it. Are either of you betting men, able to face a challenge?
Yours Sincerely
Wesley Parish (wes.parish@paradise.net.nz)