Re: [seul-edu] SIF and EduML revisited

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

On Thu, Jan 13, 2000 at 08:02:51AM -0600, Eric Sandeen wrote:

> I was thinking about this last night, and came up with this:
> It seems that SIF's XML system is designed more for _network_ data
> transmission (changes and updates are communicated over the network in
> XML snippets), independent of the actual data file format each
> application uses, which may actually be a closed spec.  The only
> requirement is that an "agent" is available to handle the XML queries
> from the network.

Yes, I agree.  SIF website suggests ZIF (ms win32 source available) as an
agent.  Someone on the list mentioned the CASBAH project (which is similar
to Frontier) as another possibility, and I think Zope could do the job as
well.  That is the beauty of XML; it does not matter what handles the XML.
What matters is that all programs speak the same XML (and if not, at least
some sort of XML so that it can be relatively easily converted)

> EduML seems to be more suited to a _data_ file format, so that
> educational applications based on it could have direct access to the
> same stored data.

EduML proposes to be both a data format and a network transfer format;
something like: programmers can save their content directly in EduML should
they want to, and/or use it to send to an agent/server who will take care of
integrating that data along with the various other educational programs' data.
So yes, both can definitively co-exist.

> Now, I'm no expert in either system, so perhaps I'm completely wrong,
> but I'm thinking that there's no reason SIF and EduML can't coexist,
> even in the same application...?

They can because they are both XML (a meta standard).  And we should
remember IMS as well.  www.imsproject.org has just released another version
of their proposed XML standard, and it looks like it is moving along at a
fast rate.  Remember that Jose Lacal (www.openclassroom.org) has chosen it
as the XML for education for that project.

After taking another long encyclopedic look at all these things today, I
think we are still only at the beginning of a process that feels like trying
to catch fish with our bare hands.  It is because everything is related to
everything else and sometime I feel like we (I) are trying to define the
universe and give three examples.

Personally, I know I cannot wrap my head around every permutation of
XML enhancements and middleware between data and internet applications  
that thousands of creative programmers are designing these days.

My best guess (for what it is worth) is to stick to python/zope/xml and
hope for the best that whatever comes out on top will also work with
python/zope/xml.

Bruno