[Freecats-Dev] The other side: "Commercial" TM tools (cont.)

[Henri Chorand]

Hi everybody,

As I'm not sure I can post to GNOME l18n & documentation lists before 
subscribing these, I hope Simos Xenitellis will kindly forward them.

> From what I can see in the discussions in both lists, the FreeCATS
> project has strong background in actually using 'commercial'
> translation tools, what are their advantages/disadvantages and
> what facilities an open-source tool should provide while in the
> gnome-docs/gnome-i18n lists there are experienced developers/
> translators with extensive experience on gettext that could
> implement it.
>
> (...) I am sure Henri (FreeCATS project leader) is able to provide
> a better summary than this. Sorry. :(

So, it's my turn now :-)

First of all, a big hello to the GNOME team. Let me briefly, as Free 
CATS' (Free Computer-Aided Translation Software) project coordinator, 
give you some background info about us and Free CATS.

This project recently started up from end-user requirements. Our project 
team started around:
- a small group of technical translators who have a quite good 
experience of commercial CAT software and who are getting tired of 
buying expensive, Window$-only commercial software which still lacks 
features
- several other people interested in developing a full-fledged free tool.

Some of us have coding experience, but not always related to free 
software's tools (like me for instance). As a whole, our knowledge of 
software development & coding is still higher than technical 
translators' average.

As mentioned by Simos, a number of free CAT software projects were 
launched since some time now, but none of them convinced us enough to 
decide to join it and follow. Of course, the more teams from different 
projects can cooperate, the better. This is not a harsh judgment in our 
minds - the scope of most of these projects was quite restricted from 
the start (e.g. Java-only files) or still at the design stage after more 
than one year's existence.

We want to build a portable, end-user friendly set of tools which, thanks
to its modular architecture, can gradually become a comprehensive tool,
as good as (and even better than) existing proprietary software. Here 
are the modules we want to build (starting from the first one):
- Translation Memory server with HTTP-access
- Remote administration client
- Interactive translation client
- Input / output file conversion tools for a lot of file formats
- Terminology database features
- Legacy translations alignment client

We also want to enable translators to work on as many different file 
formats as possible. In order to do this, the translation client will 
always work on our internal (tagged) bilingual document format. So, 
we'll build a pair of input/output converters between any native file 
format and our working format.

We presently work on specifications, not code (yet), but we have 
selected a number of tools (e.g. Python language & bits of C++ later 
where necessary) based on the features and performance we want.

As we (still) lack coders, we are especially eager to obtain some help 
in order to achieve our goals faster. This is not to say we want to find 
people to write it for us, but we're dealing with a variety of topics 
for which experienced free software developers should be able to guide 
us and help things go smoothly, with an interest in seeing it succeed.

We see this project as a win-win situation if translators & free 
software developers can cooperate on it:

- Probably more than one million professional translators in the world 
still lack the free software tools that will enable them to get rid of 
proprietary software if they want to. Helping them is therefore a Good 
Thing.

- The very existence of a successful free product, in itself, should do 
a lot to bring more help to free software & documentation translation 
projects, as translators will feel the benefit of free software for 
themselves. Even if free software developers are gradually putting up 
some tools that help them (for instance KBabel & similar), these are far 
from offering the same benefits as full-fledged CAT software. As soon as 
professional translators will be able to use their favorite CAT tool to 
participate in free software translation projects, much more will come 
to help than presently, and as they will have better tools, they'll work 
faster and produce higher quality work. I believe any free software 
advocate can agree with this.

- More than that, if today, a translator wants to use his/her CAT tool 
in order to work on a free software translation project, most of the 
time, it's not possible, because this CAT tool knows nothing about the 
lots of file formats common in free software: custom resource files (.po 
& similar), Open Office / Latex / etc.

- As soon as we have realized a first version of our server and that we 
have a client running (even if restricted to basic file formats, e.g. 
text-only and resource file), or even sooner, of course, I hope we can 
convince free software developers to come and help us, if only to help 
making these import/export filters between their own software's native 
file format and our universal working format.

We intend to publish our specifications document on our project's page 
on Savannah within a couple of weeks. Presently, all our project's 
activity is happening within our discussion list.


Well, I don't want to be too long, and I hope my little speech sounds
convincing enough. Please feel free to provide any feed-back you feel
appropriate.

Henri