22 Apr
2010
22 Apr
'10
5:53 a.m.
Philip Pemberton wrote:
Probably something along those lines - it wouldn't need to be anything
like real time.
?- Tables of contents (where available) for books,
journals, etc.
? ?Very handy when you're looking for a specific article and don't
? ?know where it is.
The current data model has TOCs for a lot of documents and I'd
definitely want to keep that feature. So for books/journals, you'd get
that for free.
?- Open-source, with as few dependencies as possible.
PHP+PDO would
? ?be my choice (maybe PHP+CodeIgniter). Should be possible to install
? ?a version on a "home" server or desktop machine (Linux/OSX box?) to
? ?play with.
Open source and runable at home: certainly. But let's not get too hung
up on the technical implementation before we know what it needs to do.
I'd really like to see a core library/API which can be used by
different languages.
?- XML-based API. Make the data available to other
systems. Maybe limit
? ?this to registered users to prevent (or at least reduce the
? ?likelihood of) data-scraping and the associated bandwidth/CPU costs.
And possibly enabled/disabled at the discretion of the person hosting
a particular mirror.
* Replication
of the entire system to mirrors, or at least the ability
to pack the whole shebang up and move it to a different host with
relative ease.
Mysqldump and a 15-minute (or whatever) period each day when the DBs sync
against each other? I'd be happy to contribute hosting, development
time and such.
Great! Looks like we have a good few volunteers. I'm wondering how we
can best co-ordinate our efforts. If may be enough to simply keep
others informed of what we're thinking about and working on. Myself,
I'm going to brainstorm a bit more, but I wouldn't mind drafting a
preliminary design some time soon.
Cheers,
--
Steve Maddison
http://www.cosam.org/