[MOBY-dev] XML Schema vs Biomoby

[Martin Senger]

>> Well.... the blame game is not useful.  Our use of XML is a valid one,
>> and XML Schema is not capable of representing it.  End of story.
>>
>   You are wrong. End of story.
>
   During last several days, we have exchanged several messages with
Mark about the message above and above the role of XML Schema in
Biomoby. There are mainly two points:

   Point one is that I was told that "you are wrong" is not a proper
choice of wording in English. Well, I can always improve my English
but not being an English-native speaker I will never be sure. Saying
that, however, does not prevent me to regret using such phrase. Sorry,
Mark.

   Point two is probably much more interesting for the biomoby developers.
We all know, and there were never doubts about it, that Biomoby XML
messages are valid XML messages. But they cannot be expressed using XML
Schema. Two reasons for that. One is (the one I was talking about): the
Biomoby messages are changing from call to call (a service *must* expect
to get more specialized message anytime). The other one (the one Mark had
in mind) is that even one single message cannot be fully described by XML
Schema. Eddie gave me an example and - as far as I know - he is working
now on a full list of such cases.

   The important message for me is: isn't it now a time to talk again (I
am saying 'again' because I started already a lobbying for that in
Vancouver) about why Biomoby services protocol is not compliant with the
XML schema, and what could be done to make it compliant? I know about
recent email exchange between Richard and Mark, referring also to the work
Ben has done, talking where to put syntax and where to put semantics in
the Biomoby messages. All of this is very promising and in good
directions, seems to me.

   It seems to me that perhaps now is a good time to talk about
it. Are we ready for that? Perhaps Eddie's list of incompatibility and
Richard/Marks comments may help to start the discussion. Perhaps
changing fundamentaly biomoby messages does not need to be so painful
now when we have libraries for service providers for both main
languages (Perl, Java).

   With regards,
   Martin

-- 
Martin Senger
   email: martin.senger at gmail.com
   skype: martinsenger
consulting for:
   International Rice Research Institute
   Biometrics and Bioinformatics Unit
   DAPO BOX 7777, Metro Manila
   Philippines, phone: +63-2-580-5600 (ext.2324)