[Freecats-Dev] Free CATS meeting (June 27), API and TM database options

[Kemper DOC (Nerim)]

Hi Yves and all,

1) Meeting
----------------
Here is my (brief) report about our meeting: it was a nice and fruitful
moment. Here come the real things.


2) API
---------
In the short term, we need Yves Champollion to have a look at the draft API
in the Development Roadmap document - I'm enclosing it again just in case.

More specifically, we need feedback from Yves Champollion (and other
experienced developers) concerning:
- any forgotten function or parameter
- any design / logic error.

For Yves, this is the time for asking us to take his specific needs
(Wordfast as a Free CATS TM server client) into account.

I'm CC'ing Tim, our developer, and Eric Cousin, enseignant-chercheur in
computing at the ENST-Bretagne (I hope I'm right), who will:
- follow the project rather closely
- provide feeback concerning the project's software architecture
- and... after assessing our first results, influence the ENSTB's decision
about whether to provide a major help for clients development (next year).

As some other mailing list members, Eric might react more quickly if his
email address is included in the recipients list.

Of course, Tim (presently on holiday for a few days) will also say what he
thinks about the API.


3) TM database options
---------------------------------
We also have to quickly decide how we will implement the TM. As stated in my
previous message, the main options presently considered are:

A) Free native XML database
I noted two seemingly mature projects:
        eXist DB                http://exist-db.org/
        Apache's Xindice        http://xml.apache.org/xindice/
(none of these project teams have been contacted yet - feedback & help
appreciated on these)
Some help from any of these teams would be a great asset. Any volunteer?

B) Free text-based database
        (Stanislav recently expressed concerns about Berkeley DB)
        there might be other free projects of interest...
        if only with search engine-related technology (?)
        Feedback from DB gurus appreciated

C) Our own system, as close to the filesystem as possible
        (for instance, all TUs from a TM in a single flat, Unicode text file)


Cheers,

Henri Chorand
Manager

Kemper DOC   Documentation Engineering
Documentation - Translation - Localization - Referencing