[Freecats-Dev] [Fwd: Re: About OmegaT and Free CATS]

[Henri Chorand]

Hi all,

I just contacted Keith Godfrey, developer of OmegaT, a Java-based CAT 
tool, as well as Marc Prior, a translator who worked on this project. 
Here is the answer I just sent to Keith's nice email.

Hopefully, we'll be able to collaborate, one way or another.


Henri

-------- Original Message --------
Subject: Re: About OmegaT and Free CATS
Date: Thu, 13 Feb 2003 22:53:48 +0100
From: Henri Chorand <address@hidden>
To: Keith Godfrey <address@hidden>
CC: marc prior <address@hidden>


Hi Keith,

> Sorry for the late reply - I'm in Central America right now and am
> only checking email periodically.

Thanks for your kind reply. I hope you have a nice time and a better
weather than in Brittany at present.

> As with yourself, I do believe there's a need for quality open source
> CAT tool.  Unfortunately, I haven't had the time to put into OmegaT to
> give it what it deserves as I simply have too many concurrent projects
> going.  I can't make any promises on how much I can contribute to your
> project, but I am very interested in hearing what you think and what
 > you have in mind.

Let's be frank. I decided to start this project two months ago. Our
(still small) team is mainly made up of translators (some coders
subscribed to our Dev-list), some members having (like me) a limited
coding experience. At this stage, we hope to be freely allocated 1,200
man hours of coding by a French engineering school, starting next fall,
and so we have a few months left to build up a detailed and realistic
specification. Any other (most welcome) help will come from the free
software community. The rule for free software is that, in order to
succeed, you need to produce something first in order to attract others,
and better code than a mere design specification.

So, all of us are particularly eager to exchange ideas in order to get
closer to our (very probably common) goal. And, whatever you can/want to
contribute, please subscribe to our mailing list so that all of us can
benefit from your feedback :-)

I believe the present team's strongest assets are:
- an in-depth knowledge of translation & localization industry
- trying to always keep the end-users' point of view in mind (some
projects are too much designed by developers or for large translation
companies)
- better knowledge of computing than average translators
- good understanding of what free software may help to change
(you probably noted that our project is under GPL)

> Some questions I have are what type of tool do you have in mind?
 > Something along the lines of Trados (i.e. integrated into a
> word processor) or ForeignDesk/OmegaT (a database design)?


So, ideally, we would prefer a standalone, interactive translation
editor like Trados Tag Editor or Déjà Vu.

If we are to select a word processor, there are only two reasonable choices:
- MS Word, but we don't want it:
        we are not founding members of Bill G's fan club
        VBA is a developer's nightmare (bugs, performance)
        not a portable solution
- Open Office
        it might be the best way to establish a strong cooperation with their
team, which we need for various reasons (mostly all the work they did
with import/export filters which we are to reuse and possibly adapt)

So, we are presently allocating most of our efforts to a TM server
module, to which any translation client, whether standalone or embedded
within a word processor.

We started with the idea of a standalone editor, but this means adapting
OO's famous conversion filters. This is why we're also considering a OO
plugin (no adaptation needed), at least for a start, and also if it
proves a more reasonable development effort.

> What platform(s) and file formats would you seek to support?

Platforms: we want to support OS X, and also MS Windows and in order to
allow professional translators to migrate smoothly to Linux if/when they
feel ready for it.
File formats: in order to be able to use this tool in our daily work, we
need as many of these as possible. The "mandatory" ones are, starting
with the most important ones:
- HTML & XML (including OO's Writer format)
- RTF and possibly MS .DOC (done by OO)
- ANSI text and lots of text-based resource file formats (.RC,
WinHelp/HTML Help resource files - these should be rather easy to do)

Apart from these, we would also like to add Latex and Linux resource
files (eg. .PO), & proprietary DTP files if/when more developers help us.

> Might OmegaT find a place in your plans (i.e. re-using some of it's
 > code and functionality in another product or migrating it to serve
 > somewhat new/different purposes)?

Definitely, yes.

I must add something about Java. We know that it's quite popular and
portable, but is has performance issues and it's difficult to learn.
This is why, ideally, we believe we might choose Python. But we don't
want to lock ourselves in a "theoretical" (rigid & stupid) position.

Python is good at integrating any other code. Some of us (Julien and I)
are presently learning it. So, we might be more hopeless if you want to
show us Java code ;-)

> Please do send a copy of your specification documents - I suppose
 > they would probably answer many of my questions

Done. I'm sending these as RTF, which we choose as our exchange format
because OO's HTML has a few problems and OO is still a beta on Macs.
As you'll see, they are drafts to be improved - it's a start.

At this stage, we're designing an architecture, addressing various
issues, assessing & selecting existing GPL code. At this stage, I also
hope you can help the indexing features by comparing what you did in
OmegaT with our proposal.

Apart from the switch from Tcl to Python, the latest info (not
implemented in our reference documents) are:
- instead of a database, we might as well use the server's filesystem.
- for the server component, we are considering using the Twisted
framework. This stuff looks a great tool in the hands of experienced
developers.

> :-)  Also, I have to admit up front that I do have a somewhat strong
> anti-Microsoft bias.

Well, I must admit we will all agree here.

> I CC'd Marc Prior, an English/German/Italian/??? translator and also
> administrator of the linux-for-translators forum (at yahoo) to establish
> contact.  I believe it's through him that I heard about a new CAT tool
> in planning/development for the Mac.  I don't recall the details on that
> project, but it might not hurt to establish contact with that team in
> case it's possible to pool efforts and resources.

Fine. I'm CC'ing Marc.
Both of you are most welcome in Free CATS project team if you feel like
it, whatever time & resources you have.

I'll also CC (without the attachments) to our Dev-list.

Some time ago, I was forwarded some emails by fellow translators about a
new proprietary CAT project for Macs. It looks like they were trying to
raise interest in order to determine whether to go along with their
project. You might be referring to another project, though.


Cheers,

Henri