Re: [school-discuss] Thoughts on XML data interchange

[vernier@vc.bc.ca (Bruno Vernier)]

On Sun, Jan 27, 2002 at 01:34:32PM +0000, Gunnar Stefansson wrote:

> Being a newbie to xml (and this entire discussion) I wanted to take a
> look at just what you're doing, but
> 
> https://robertscap.bc.ca/rec/REC/Eduml/school_identification
> 
> does not seem to work and I got a bit stuck with where to go with
> http://k12admin.cmsd.bc.ca/cgi-bin/data. 

both of these links are not intended ultimately for humans, but for other
programs that will process the raw XML inside and transform it to some
useful report or quiz or educational activity for real humans

so the correct way to take a look at them for now is to "View Source" them
in mozilla/netscape and they should show up as an xml tree in later I.E.
browsers.  In lynx, use backslash to show the source.

The lesson here is that every participating school or learning community
needs to speak the same XML schema ... and so as more people join in, we
should always be seeing the same type of XML. 


> The idea of linking all such on-line schools together sounds great.  I
> would very much like to be a part of that with my alpha-stage
> tutor-web.net, which provides a framework for lectures, tutorials,
> courses and associated tests.  So far I've only been using simple html
> (and a little javascript) generated by perl programs.

sounds like exercise 3.  Again, it does not matter what you use; simple text
editors, or perl/javascript or J2EE/java ... in the end the output from
participating schools will have to be a commonly-agreed form of XML.  That
is the sine-qua-non prerequisite to any chance of benefitting from
interoperability.  (I know you know all that Gunnar, I am just trying to
explain the process to anyone new to this game)

> Reasonably soon I will need to include a formal registration
> procedure, but security is not really much of an issue as far as I am
> concerned at this stage (at the moment I simply register a user by
> querying for an e-mail address when the first question is requested
> during a session).  Since I've only used a single course as a live
> testbed grading (though automated) is done through the csv database in
> perl (I can still see what I am doing:-).

so from where I stand, if you were to do exercise 1, you would write a cgi
in perl that  extracts from the cvs database (and for the rest of us, it
does not matter that we do or don't know what perl and csv mean or do) and
spits out the willing participants name through a URL in common XML for use
in exercise 2.

> Do you have a formal classification scheme of course content in place
> (i.e. is it covered by one of the acronyms you folks have been
> using:-)?  It'd be nice if I could for example say in the introdution
> to one of my tutorials that the content is x and to take this I would
> recommend taking some course with content y and a mouse click would go
> through superschool to Bruno's course with content y.

that is indeed the ultimate goal ... but we are indeed at the very
beginning so we can't show or do that yet.  We have yet to formally 
agree on a common XML.  Yes, all of Eduml v1, SIF, IMS, Cancore, Dublin Core
offer specific Schema for these learning objects.

> My courses would certainly gain a lot if they can be hooked up to
> other (free?) on-line courses.  The converse is probably not true
> today (partly to to language) but I would like to think it will become
> true as I enhance the tutor-web (i.e. more courses in English).

yes, everyone wins (the commercial companies in the IMS and SIF consortia are
very convinced of that from a purely commercial point of view) and we, in
the open source world, have even less trouble to be convinced :-)

> So: Is there a way to get into what you are doing through a reasonably
> smooth learning curve?  If you have pointers to a couple of examples
> of what you need to generate and how you do it, that'd be great.

which is why I have proposed these basic exercises.  I hope we will all
assist each other in completing the exercises so that no one is left behind
who does not want to be left behind.

> Sorry for being so basic...

On the contrary, we are desparate for confident people like you who are not
afraid of forcing this conversation to stay basic, in simple english with
clear examples ... and to continuously force all of us to stay or become
that way.

Bruno