Luis, Mougey is his last name. It's customary in France to state the
family name first in some formal texts ;-)
You should call him Camille.
Camille, I agree with Luis here : I don't think you can really have
everything fit in the summer.
Plus, remember that, like for me last year, you'll probably have the
school's end-of-year project to deal with until mid-June (unless you can
convince the teachers that you don't need to do it...), so the first
three weeks will have to be lighter.
Aside from that, your proposal is interesting. I'll leave a few more
comments below.
4) Provide more formatting modules; currently only ODF, OOXML,
html and
text are supported. Mediawiki (using apertium-mediawiki) and
others are
wanted.
-> Add module for Mediawiki formatting
-> Add module for Rich Text Format formatting
-> Add module for Pdf formatting
-> Perhaps consider the possibility of using OCR for
Pictures ( by using
Google Apis for example, or OpenCV which is open source )
I would suggest you to take a closer look at apertium deformater and
reformater. PDF is a closed format and its handling would imply a hell
of complication. Even if you use a third party tool. Sergio Ortiz has
some experience on it using a propietary tool (which is out of
consideration) and there are quite a bunch of problems with that.
Indeed. Pdf will probably be an incredible mess; it isn't meant to be
decoded. You might want to drop this.
However, Mediawiki and RTF support should be easy to add, since Apertium
already has formatting scripts for those. So this part (adding
formatting modules) could be made very short in the planning.
5) Localisation, make it possible to translate the interface into
different languages.
-> The localisation is given by IP Address, or by the
browser, or set by
the user
-> The choice is save ( cookies, .. )
-> The interface texts are load from XML files, which
contain the text
for every button, checkbox, ...
If you choose to store language files in XML, you may also want to
design some kind of cache : parsing an XML file at each time the
interface is ran would be quite an important cost if the interface were
to be used by thousands of people :-)
Though it's not an immediate concern anyway.
-> Perhaps allow the user to propose new interface text (
by making XML
file ), or using his own
I don't think that's really useful. Normal people will just want the
localisation for their language to be there, and people who do want to
write a language file should share it with the Apertium community :-)
9) Design a server-side TMX database, so that the memory generated
after a
translation be stored and reused automatically for next
translations in
this language pair. (It might be wise to add some kind of
validation too,
to make sure that people don't mess with the whole system by
submitting
wrong translations...)
-> These tasks go together : The main idea is to permit to
the user to
give or alter current translations, export them to TMX format. At
the same
time, these modifications are saved in a server-side TMX database,
which
contain two kinds of dictionaries by language :
- Submitted and awaiting for approving translations
- Approved translations
-> Add in the user interface the possibility to use
Submitted and
awaiting for approving translations ( not recommended ) and/or
approved
translations ( recommended )
-> May we can see here a way to use the logging system
I think this should be left to the user. If someone is interested in
reusing his translations, would be able to do it using the export TMX
function and then manage his TMs using some TM merge solution. But I
think this should be done outside the web interface. In addition,
there is the confidentiality problem with content submitted to the
engine.This can be solved by asking the user for permission to use his
text for improving the system but should be done in a very simple way
that I think would discard storing whole TMXs. I think that efforts in
this sense should address how to capture user edits on the Apertium
output so they could be exploited to identify improvement points for
the engine.
Actually, here I quite like Camille's idea (after all, I was the one to
suggest the idea of a server-side tmx db ^^). As I see it, it'd be more
useful for regular users to have an automated process (with simply a
checkbox "Reuse old translations to improve translation" and another one
"Share translation results", or something like that), than downloading a
tmx file.
13) Make installation easier. Fully describe the installation
process on
the Wiki.
-> Add screen-shot/video tutorial to the Wiki to describe the
installation.
-> For Debian based system, may create a *.deb package,
*.rpm for Red
hat, ..
-> Perhaps create a simple interface for the installation
process ( more
attractive for Ms Michu )
"Ms Michu" isn't exactly understood by non-French people ^^
Jim, it's just a way of saying "a random person"; kinda like John Doe
means an unidentified person ;-)
Anyway, I wouldn't focus on deb packages. First, because Apertium itself
doesn't really provide debs (the ones on Ubuntu are usually not up to
date at all, etc.), and second because one may want to install the
interface anywhere on the system whereas debs contain the hard path to
files, if I'm not mistaken.
For this task, you should rather focus on an interface to create the
config file easily, or something like that. So that one would just have
to extract the files, visit the install script in the browser, fill the
information, and be done with it.
14) Improve integration with Wikipedia — it should be able to
fetch pages,
translate, allow to be revised and then published.
-> Integrate module for recognize Wikipedia format
-> Make or use existing Wikipedia API to fetch pages, and
post them (
allow to be revised and then published )
Ideally, FTyers would like to get it integrated to a tool like
http://translate.google.com/toolkit . Building such an interface would
probably be too much for this project considering the other tasks, though.
Good luck with this proposal !
Regards,
Arnaud
------------------------------------------------------------------------------
Create and publish websites with WebMatrix
Use the most popular FREE web apps or write code yourself;
WebMatrix provides all the features you need to develop and
publish your website. http://p.sf.net/sfu/ms-webmatrix-sf
_______________________________________________
Apertium-stuff mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/apertium-stuff
