[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [school-discuss] ZIS Server Dreams

On Wed, 2002-01-23 at 06:12, Tonnesen Steve wrote:
> On Tue, 22 Jan 2002, Les Richardson wrote:
> > 
> > I've been looking at the SIF specs and muttering...oh my, oh my, ...
> > 
> > It seems to me that there are several things that have to come before the
> > ZIS server
> > 
> > To me it would seem like more pressing solutions would be:
> > 
> > 1) Building SIF compliant applications that have a built in agent to talk
> > to ZIS servers. One would like to have a small complement of apps to carry
> > out the core administrative functions, which IMHO are:
> > a) attendance
> > b) student demographics
> > c) report cards (ie. student reporting)
> > 
> > 2) A toolkit (or software assemblage) including good documentation to SIF
> > enable existing applications.
> > 
> > 3) A good database schema for student demographics, grade book/report
> > cards, attendance, 
> > 
> > 
> > I would think that #3, and #2 would lead to number 1...more SIF capable
> > apps. Once we have a few of those, then ZIS server could occur. Otherwise,
> > I would say that it's premature...
> The SIF specifications has some schemas laid out already.  My cursory
> glance through them makes me think that there is plenty missing.  Bruno
> Vernier's EduML might be another good starting point. 
> My thinking is that developing a ZIS server would be an excellent way to
> come up with a solid #2.  It's going to be pretty hard for open source
> developers to add ZIS support to an application if they don't have access
> to a ZIS server in any case.  Also, I believe that we already have some
> applications that could benefit from the addition of SIF.

	I had not much time to read the spec but of what I've read I think the
server in itself isn't much work. The main parts are the communication
libraries used by the agents and the server. I think the development in
of the toolkits would be the main part of the development of the server.
So server and toolkit developments should take place in parallel
(permiting unit and functionnal testing along the way).

	But offcourse the server would be useless without a database and
applications that could use it. Here too, the toolkit are the main work
to do, because a lot of applications are advanced enough to plug easily
a new communication module. The difficulty would be to come up with
toolkits useable in the various languages that are used for those
applications (c, perl, python, php, etc...)

	My 2 cents(euro ;) participation could be :

	- The implementation of a part of a java version of the toolkit.
	- Pluging this toolkit in my already developped application in order to
have a provider agent. This would be easy because I already use XML to
generate HTML and PDF (through XSL:FO) and SIS would just be another
output format (input would need more work but nothing impossible). 
	- Implementation of Jython test suites in order to test the server and
the agents.

	I hope this will help,