RE: Bad News on IE6 XML Support

[Eric Bohlman <ebohlman@earthlink.net>]

9/10/01 1:06:30 PM, Michael Rys <mrys@microsoft.com> wrote:

>Eric, the excerpt below is what our XML teams are planning on doing
>(look at MSXML and the System.XML support). IE team members (I think)
>are not even reading this list. 
>
>However, data outlives code, so once you allow people to store data in a
>certain form, you simply cannot disallow them. Warning is an option, and
>providing transition periods (that can range several years/decades).
>This does not move in internet time, I am afraid...
>
>Best regards
>Michael
>
>PS: As to the embrace and extend arguments. I give you two replies.
>Choose the one that Occam's razor would suggest:
>
>1. MS intentionally screwed the XML parser up as an evil plot to embrace
>and extend and rule the world (insert evil laughter).
>2. MS engineers did not read the XML recommendation as careful as they
>should have and made mistakes that are now hard to fix and take time...

Of course it's 2: my point is that even though this was unintentional, it creates a strong vendor 
lock-in incentive in favor of Microsoft, i.e. it has the same *effect* as a deliberate embrace-
and-extend decision.  Therefore, I think Microsoft has *some* responsibility to the XML community 
at large to help alleviate the negative effects of this mistake.  Perhaps it could develop some 
conversion algorithms and submit them to the USPTO *as examples of prior art* and, of course, make 
software based on them available on an open-source (not necessarily GPL, in fact preferably *not* 
GPL) basis.  Examples might include adaptor classes specifically tailored to specific 
products sold by Microsoft competitors.  Certainly updating product documentation to make it 
extremely prominent that the current behavior with illegal characters is a bug, not a feature, 
would be called for.  If you're going to do any certifications involving XML, then several 
questions regarding legal characters should be a part of the test.