[Freecats-Dev] Tr: Welcome to Free CATS list

[Henri Chorand]

Hi all,

These are a few remarks on Dan's feedback.


> > Recently (last summer), a few "real" developers have
> > expressed interest in this project and joined the team
> > (Michael & Tim), but no code has been released yet.
> > Tim is presently trying to see how the TM server could
> > be built with MySQL, and I hope Michael can do something
> > about its overall client-server architecture.
>  >
> > Once we have a first draft of our TM server, Yves
> > Champollion, the author of Wordfast, the proprietary CAT
> > tool, has promised he would interface his software with
> > our server, and at that stage, we would then begin to have
> > something. I also hope that other free CAT software project
> > teams go along the same lines, in order to provide a
> > common root.
>
> That would be great. Are you planning a web-services
> interface; and if so, do you have any sort of interface
> specification yet?

As a rule, we professional translators consider that a web services
interface is not a perfect goal, even if it would be handy, and nothing
would prevent anybody from developing one compatible with our TM server, of
course.
There are two reasons:
- A web interface can't be as productivity-oriented as a traditional GUI
interface, and we do hope that some Win/OS X/Linux portable tools are
becoming mature enough for something to come along out of it.
- We realize that some of the people who are considering the web services
approach start a bit too much from the large localization agencies' point of
view (top to bottom, taylorist approach to translation) and that it might
give the translator less control over the process than is desirable - for a
variety of reasons, first because real life is complex and that we intend
our tool to provide maximum flexibility on the management (read: ability to
hack) of the source materials.


> It's a bit surprising, given the need for large-scale opensource
> projects to translate documentation, that there aren't better free
> tools yet. I'm not sure how feasible this is, but I wonder if it would
> be possible, once your project really gets going, to get server space
> somewhere to set up a centralized TM repository for free software
> documentation? It seems to me that would really help the cause of
> getting more open source docs translated.

We quite agree with that. I see this as a win-win deal:
- The free software community helps giving professional translators the free
tool they have been dreaming of. This enables translators to free themselves
from the "trio infernal": M$ Windows / Office / Trados.
- Thus doing, this free software community gets two things:
1) They provide themselves with a set of tools that will greatly help the
work of translators helping to provide quality documentation for their
projects in as many languages as possible.
2) By giving away the results of this effort, they have an opportunity to
attract a much larger pool of translators who will actually help them
localizing their stuff.


We started this project with some known strengthes and weaknesses:
- Starting from professional translators requirements, we wanted to achieve
something that would target a much, much larger potential users base than
the .PO translation tools that are emerging (KBabel or POEdit).
- This implies a lot of work, and we lack developers. In fact, the idea is
that, once some of the free software "leaders" hear more about Free CATS,
they can publicize this project in order to attract volunteers and to give
it the momentum it deserves.


> > Also, I sincerily hope that a free and portable TM server would
> > help the existing free projects a lot, by providing a common
> > ground around which different clients could be developed.
>
> That's definitely the right architecture; but I think it might really
> be useful to have an interface specification first. That way, people
> could also write different server implementations. This is what I've
> done with my package; the server implementation is pretty trivial
> and won't scale well, but at least I have something I can test a
> client against.

This is great news.

You will see a number of messages in the archives dealing with the
possibility to develop a client as an Open Office plug-in, for several good
reasons:
- Portability would closely follow the portability of OO
- We would reuse all their document import-export filters, which represent
an awful lot of work that we translators also need.

That's all for now... ;-)


Cheers,

Henri