Re: [suse-linux-uk-schools] MIS, databases
I agree with Christopher's comments on the choice of technologies for an open source project. I would have though that the ideal choice of technologies for the job would be to use MySQL or PostgresSQL (or other good open source database) as a back end, PHP for the programming and all user interaction through a browser (it seems silly writing native clients in this day and age - particularly for this type of job). I can't foresee any performance issues with this setup and such a solution could be hosted on almost any modern computer platform, with client access via absolutely anything with a web (or even WAP) browser. HTTPS could be used between browser and server to ensure that all net traffic is encrypted and that secure access is available if required from inside and outside the school. Probably the hardest component of the overall job would be the database design so Richard may well have already done the tricky bit. Cheers -- Phil Driscoll Dial Solutions +44 (0)113 294 5112 http://www.dialsolutions.com http://www.dtonline.org
On Sun, 29 Apr 2001, Phil Driscoll wrote:
I agree with Christopher's comments on the choice of technologies for an open source project. I would have though that the ideal choice of technologies for the job would be to use MySQL or PostgresSQL (or other good open source database) as a back end, PHP for the programming and all user interaction through a browser (it seems silly writing native clients in this day and age - particularly for this type of job). I can't foresee any performance issues with this setup and such a solution could be hosted on almost any modern computer platform, with client access via absolutely anything with a web (or even WAP) browser. HTTPS could be used between browser and server to ensure that all net traffic is encrypted and that secure access is available if required from inside and outside the school.
For what it's worth, Fen Systems is currently bidding to create such an open source school admin database, using PHP/PostgreSQL. If it goes ahead, then there should be a first working version by December, with a full working UK version by Summer 2002. BTW, PostgreSQL is probably a better choice than MySQL, because support for transactions may be required. SAP is a possibility, and has the advantage that (AFAICT) it supports replication properly (MySQL supports only master/slave replication and the PostgreSQL replication server seems not to be properly open-source). Michael Brown Fen Systems Ltd.
I've mailed Miles Berry to express my interest in this idea, but discussion seems to be taking off here too, so I'll chip in: Clearly the hardest parts of this are the data modelling and database design. We've already had one offer of a Free database schema, and perhaps working towards that would be a good way to start. (Is there a schema based on the DfEE 'Common Basic Data Set'? If not, maybe we should ask DfEE if it would support definition of a (Free) one?) I'll second the choice of PostgreSQL for the database engine, and mostly web-browser interfaces. However, I wouldn't exclude the possibility of writing special clients if functionality required it (and I'd write them in Tcl/Tk, for portability). Bob Gautier Ateb Limited
participants (3)
-
Michael Brown
-
Phil Driscoll
-
Robert J Gautier