Well, it would be OK for *me*, but ideally it should be cross-platform.
Why? Does it need to run on Windows? No. If it did, might as well buy the M$ stuff. Does it need to run on Macs? Any toy library (aren't they usually cash-strapped like all the other early edu institutions) got any Macs? I don't think so. You have only Linux to think of. The other supplier of M$ toy lib software couldn't care less about Linux. In fact, he will actively fight it (it's cutting his profits).
Remember, your position is one of fighting for survival. You need to come up with a solution which does any job at all, let alone a good job. The biggest argument in your favour would be cost, and you need to play it in cold blood, against established commercial supplier's and Microsoft's "gifts". (How many free copies can M$ cough up for toy libraries if they need to?) If you want to support Windows, your biggest drawcard is half gone. Cross-platform is at the bottom of the list.
You're right. And wrong. The problem is, the toy library that I know is running windows on both computers, and some software that they paid good money for. I have to be able to demonstrate that an OSS solution works without major changes, and is considerably better, before they'll contemplate chucking out the investment (money, but mostly *time*) in the old product. If I was getting there first, it'd be Linux all the way.
It would seem to be prudent to start small and make version 1 "work", even in the Microsoft-definition of "work", and to leave improvements for version 2. Once version 2 is done, you can think about porting to other OSs. In any case, neither php nor python restrict you to Unix. postgres could be a bad choice, mysql is available for Windows IIRC. I still wouldn't care at this point, once you have a foot in the door migration from postgres to mysql, if that is even necessary, isn't such a big step in comparison.
My 0.02 * $CURRENCY
appreciated.
Cheers, Carl.
