Microsoft and Federal Government Reach Tentative Settlement

Wednesday October 31st, 2001

MSNBC is reporting that Microsoft has reached a tentative settlement with the Federal Government. The deal calls for a 5 year consent decree that forces Microsoft to release Windows without a variety of currently bundled programs like Internet Explorer, MSN Messenger, and Windows Media Player.

As you may know, the US Government has had consent decrees before, and Microsoft has completely ignored them, we feel. As in the past, we expect the version of Windows without the bundled packages to be almost identical in price to the current version of Windows. The last major question remaining is whether the State Attorneys General agree with the deal. We hope they do not, and push for much harsher punishment.

#45 Why Johnny Can't Parse

by choess <>

Friday November 2nd, 2001 5:57 PM

You are replying to this message

One of the questions that arises out of Microsoft's ridiculous claims of "standards compliance" and so forth is "How can they get away with it?" An interesting question; after all, why should the W3C be expected to sit quietly while MS makes grossly misleading claims about supporting its "standards"? The problem here is the composition of the W3C. While it's done a fairly decent job of acting like a neutral standards body for the past few years, it is an industry consortium, and as such, operates in the interests of its members. It's all well and good to write standards, of course, but who wants to be held responsible for *implementing* them. After all, it could take a lot of programming time, money, and effort to completely implement them; it would be embarassing if companies were called to account for that, and it would hurt people's feelings. Better to use nice fuzzy terms like "strong support" and pick and choose the easy bits. Of course, since there's no reference implementation or comprehensive test for standards, the Teeming Millions will eat up any pap you give them about your support vs the competition's.

Of course, it doesn't help that the W3C has taken a quick-n-dirty approach toward the documents it produces; "Recommendations" have the advantage of being able to go from draft to recommendation in "Internet Time", but they lose some of the rigor associated with IETF standards. Furthermore, recs like CSS2 were churned out without regard for implementation, leaving a document riddled with ambiguities and errors. (There's quite a few bugs in Bugzilla with some variation of "ask wg for clarification" in the status line...) Fortunately, the W3C Process Document now requires a WG to demonstrate an implementation of each feature in a Candidate Recommendation before it can advance to Proposed Recommendation, but it will take a long time to undo the damage. The W3C has also, encouragingly, begun a Quality Assurance activity, which is making vague noises about certification, but it will take them a long time just to assemble the tools for the job, let alone actually starting to evaluate software.

As bleak as this may sound, I think there are signs of hope. It looks like RAND is being forced into retrograde motion, such that new recommendations may be required to be royalty-free, and I think the dedication of projects/companies like Opera and Mozilla may give the W3C the courage to stand up to MS and say, "We're going to do what's best, whether in writing recommendations or enforcing them, because people respect and rely on them, and going home to write your own standard isn't going to fly anymore." (Look at Samba, for example, which seems to have been quite successful at reverse-engineering a "commoditized" protocol.) The Web community has begun to see what a powerful thing universally-accepted standards are with XML; MS won't find it so easy to break things now, I think.