Re: [seul-edu] Food for thought regarding M$ and SIF

[Bruno Vernier <vernier@vc.bc.ca>]

On Mon, Jan 17, 2000 at 10:49:20PM -0600, Eric Sandeen wrote:
> I'm sure it's worth keeping an eye on, but Microsoft can write all the
> DTDs they want, for all I care.  If they start breaking XML, then that's
> another thing, but I haven't seen any evidence of that.  In fact, XML is
> much harder to break than HTML, due to its more extensible nature.

Microsoft does not need nor want to break XML: 

Please keep in mind the principle of yin-yang  and  what Marshall McLuhan
said about overdosing on new artifacts; how we end up somewhere near where
we aimed to get away from when we first introduced the said artifact.

The reason I got excited (and by extension why I think so many digerati got
excited) about XML is that it aims at getting away from proprietary data
formats allowing free flow of information (content and commands)
to and from any conceivable application, anywhere, anytime, easily.

I won't speak for the commercial educational software industry, but i'll
guess they are getting excited about XML because their prospective clientele 
is reluctant to purchase yet another standalone product blind to the
existence of other programs already in the school network: sale prospects
look better if the sales pitch can include something about not having to
maintain yet another authentication database, and not having to re-enter
marks once they have been "earned".

The articles on slashdot suggests that MS is excited about XML because 
1. it is something they can graft onto windows without breaking much,
2. it will make them look pioneerish (one the 2 co-authors of the
   XML definition Mr. Bosak works for MS), I.E.5 is the first browser to
   do xml, many XML dtds are being sponsored by MS. incl both SIF and IMS
3. if XML really takes off, the MS machine will be in a good position to 
   take business advantages;  Its people are becoming quite XML-literate,
   and its machinery can handle/use the huge volume of xml coming from
   e-commerce transactions.  Knowledge may be power.  Transactions 
   management is where even more power might be. 
   
Which brings me to my point:  I want XML in order centralize all educational
transactions (marks earned, levels attained, units completed, attendance 
records etc...).  Apparently, I am not alone:  commercial software companies
feel compelled to open up their portion of educational transactions (via
XML) in order to please school decision makers.  Microsoft wants XML in order
to be in the middle of the transaction action so to speak.

Therefore, no one seems to have an interest in breaking XML.  Yet, I
predict that if XML does get "broken", it will be broken by the open
source community,  accidently.  By creating hundreds of permutation on the
XML theme  (middleware that communicates content and commands between
applications), the intensely creative open source community is accidently 
re-creating something not unsimilar to the proprietary and closed data 
formats it seeks to replace:

instead of a plethora of closed proprietary content and command sets,
we are about to have a plethora of open, possible content and command sets.
 - thousands of XML dtds, many of which overlap incl. SIF, IMS, EduML
 - COM/DCOM, soap, webdav, Java and company
 - Corba ,  many variations on that theme
 - RPC, xml-rpc, wddf, and other variations on that theme (including my own
    attempts within EduML)
 - Python, Perl, TclTk, Emacs, VB, and sundry scriting languages which
    are being enhanced to handle remote handling of content/commands
 - Zope, Frontier, Casbah, Baboon, and cgi-based permutations
 - SQL and its new extensions into this arena (and the noSql alternative)

Though I enjoy the freedom of potentially studying and modifying these new
sets, in practice, they can paralyse me:

which of the open, possible sets should I focus my time-limited attention on?
and is there a half-life of any given set?  Yes, it is much easier to
change from one to the other of these sets; but really: nothing is ever
completely easy and simple; any conversion necessarily takes precious seconds.

Part of the trouble is that communicating content/commands between
applications is not unlike creating languages (noun/verbs) between human 
beings;  the variations are potentially infinite and each permutation has 
its charms. Translations are always possible and sometimes unecessary if the
dialects are close enough.
 
Conclusion:  Microsoft does not need nor want to break XML.  They may be
calculating that Open source creativity may have a marketing weakness;  
it may be easy for MS to recycle the old argument: look at the unix world; 
it is a real tower of babel; stick with us, for we may not be the best,
but we can impose a de-facto standard in the area of middleware.

p.s. Yes, I know I should have included IBM as a key player in this drama.