Loading ...
Sorry, an error occurred while loading the content.

We've got a poll on the Yahoo page...

Expand Messages
  • jc_helary
    Ok, I see the group is a little sleepy :) So, here, I put a poll at: http://groups.yahoo.com/group/OmegaT/surveys?id=1644793 so that we can get a priority list
    Message 1 of 21 , Mar 9, 2005
    • 0 Attachment
      Ok, I see the group is a little sleepy :) So, here, I put a poll at:
      http://groups.yahoo.com/group/OmegaT/surveys?id=1644793

      so that we can get a priority list for Maxym :) and other would be developers :)

      Besides, I just got myself a fun deadline (like "right now !") so I'm getting offline now.

      Also I just updated the moderation status of a good bunch of you so don't feel afraid to
      post.

      OmegaT exists because you use it and because you contribute to its evolution...

      Cheers,

      Jean-Christophe
    • Sabine Cretella
      Hmmmm ... they are all interesting ... Ciao, Sabine
      Message 2 of 21 , Mar 9, 2005
      • 0 Attachment
        Hmmmm ... they are all interesting ...

        Ciao, Sabine

        jc_helary wrote:

        >
        > Ok, I see the group is a little sleepy :) So, here, I put a poll at:
        > http://groups.yahoo.com/group/OmegaT/surveys?id=1644793
        >
        > so that we can get a priority list for Maxym :) and other would be
        > developers :)
        >
        > Besides, I just got myself a fun deadline (like "right now !") so I'm
        > getting offline now.
        >
        > Also I just updated the moderation status of a good bunch of you so
        > don't feel afraid to
        > post.
        >
        > OmegaT exists because you use it and because you contribute to its
        > evolution...
        >
        > Cheers,
        >
        > Jean-Christophe
      • Marc Prior
        Hey JC (when you get back from your fun deadline), Where s the option for OmegaT is already the TM app I ve always dreamed of ? I expect the poll will provoke
        Message 3 of 21 , Mar 9, 2005
        • 0 Attachment
          Hey JC (when you get back from your fun deadline),

          Where's the option for "OmegaT is already the TM app I've always dreamed of"?

          I expect the poll will provoke quite a bit of discussion.

          Marc
        • suzume@mx82.tiki.ne.jp
          ... the guy says he sends me the file but he forgets to attach it ;) _never_ be in a hurry ! ... that s the whole point... btw, I checked the cvs on
          Message 4 of 21 , Mar 9, 2005
          • 0 Attachment
            > Hey JC (when you get back from your fun deadline),

            the guy says he sends me the file but he forgets to attach it ;)

            _never_ be in a hurry !

            > Where's the option for "OmegaT is already the TM app I've always
            > dreamed of"?

            :) I knew you'd say that :)

            > I expect the poll will provoke quite a bit of discussion.

            that's the whole point...

            btw, I checked the cvs on sourceforge and I noticed that nothing had
            been put there in the last 4 weeks.

            besides i don't hear of Maxym much, any news from the dvpers side ?

            JC
          • S. Tomaskovic
            No multiple selections? I could have checked at least three items from that list. Nevermind. But since this poll is supposed to provoke some discussion, I just
            Message 5 of 21 , Mar 9, 2005
            • 0 Attachment
              No multiple selections? I could have checked at least
              three items from that list. Nevermind.

              But since this poll is supposed to provoke some
              discussion, I just realized that no one is interested
              in a better glossary function or scriptability or
              *direct MS Office support*.

              While I don't think that scriptability is a must, I am
              surprised that no one wants a better glossary function
              nor direct support for MS Office. Are you really happy
              with how the glossary works right now?

              Where are those people in favour of MS Office support?


              Sonja ;)






              ___________________________________________________________
              Gesendet von Yahoo! Mail - Jetzt mit 250MB Speicher kostenlos - Hier anmelden: http://mail.yahoo.de
            • Marc Prior
              Maxym already has the road map all neatly planned. My guess is that he ll tear his hair out when he sees the results of the poll. :-) My thoughts on the
              Message 6 of 21 , Mar 9, 2005
              • 0 Attachment
                Maxym already has the road map all neatly planned. My guess is that he'll tear
                his hair out when he sees the results of the poll. :-)

                My thoughts on the various features:

                * XLIFF support (TM file format)

                This sounds unglamorous bt should solve a few other problems, like the lack of
                one-to-many translations. But how much work is involved in switching OmegaT's
                internal TMX file to XLIFF, I don't know. I suspect that there are indexing
                problems.

                * better TMX support (TM file format)

                Upgrading to TMX 1.4b should really be no problem at all.

                * Trados TM support (TM files format)

                Does anyone know anything about the current Trados TM format? Isn't it
                encrypted?

                At any rate, Trados supports TMX and OmegaT supports TMX (it would help if
                Omegat supported 1.4b, of course), so compatibility between the two shouldn't
                be a problem. The real need is for support the "Trados uncleaned file
                format". Now that OOo supports the hidden text character property, It should,
                I think, be possible without too much programming effort to generate output
                in the form of a bilingual OOo file that only need be saved as MS Word within
                OOo in order to produce an uncleaned Trados file. I've been meaning to figure
                out what such a file would need to look like (not difficult), but haven't had
                time.

                * Direct MS Office file format support (project files)

                We all know that this isn't going to happen. It's a huge amount of effort. The
                best we can hope for is to integrate someone else's routine into OmegaT as a
                plug-in. For instance, a call from OmegaT that launches an OOo script which
                in turn saves the selected MS Word file as OOo Writer. The OOo script
                required is actually fairly simple and there are implementations of it
                already available on the web, so what we really need is someone willing to do
                a little coding in Star Basic, then our Java developer(s) to add code to
                launch it.

                * RTF support (projet files)

                More realistic than direct MS Office support but not really any improvement
                over converson via OOo, in my opinion.

                * generic XML support (projet files)

                My vote. :-) A function that enables end users to select which tags should be
                treated as segment breaks and which should be treated as inline markup would
                be really nice. In the meantime, though, the existing file filters are not
                that difficult to modify, and what I'd really like in the interim is a
                generic Unicode XML filter that a user can modify. But I know Maxym has this
                in mind.

                * internal segmentation function

                Nice but involves a fairly major rework of OmegaT.

                * internal spellcheck

                Should be very easy to do with Aspell, and it's already on Maxym's list.

                * better glossary function

                Would be nice, but if you make good use of the "Find" function, the glossary
                becomes less relevant. I also have a hunch that large glossaries may involve
                performance issues that we haven't yet noticed.

                * scriptability

                I discussed this with Maxym once (adding bindings to external scripts), but we
                didn't reach a conclusion on how it could be done.

                Marc
              • Samuel Murray
                ... Here I am. But a better GUI would also be nice. Typing in a Java application is just plain difficult. I have seen Java applications that react to
                Message 7 of 21 , Mar 9, 2005
                • 0 Attachment
                  S. Tomaskovic skryf om 11:15 AM op 09/03/2005:

                  > Where are those people in favour of MS Office support?

                  Here I am. But a better GUI would also be nice. Typing in a Java application is just plain difficult. I have seen Java applications that "react" to input with the same ease as any Windows program, and I've seen ones (like OmT) that react like a typical Java app... sluggish, non-standard keyboard shortcuts, etc. I'm not complaining, though... OmT certainly is useful. But it won't become my main t9n tool even with direct MS Office support (unless I make the switch to Linux).

                  Direct MS Office support depends on how far the OpenOffice guys are with their separate filter set, aint that right? So its a case of wait.

                  Samuel
                • Marc Prior
                  ... The filters are already very good and probably a major reason for OmegaT s acceptance so far. I m expecting them to improve but they ll never be perfect.
                  Message 8 of 21 , Mar 9, 2005
                  • 0 Attachment
                    Am Mittwoch, 9. März 2005 11:36 schrieb Samuel Murray:
                    >
                    > Direct MS Office support depends on how far the OpenOffice guys are with
                    > their separate filter set, aint that right? So its a case of wait.

                    The filters are already very good and probably a major reason for OmegaT's
                    acceptance so far. I'm expecting them to improve but they'll never be
                    perfect. An automated routine that required OOo to be installed but made its
                    use invisible to the user would put an end to a lot of the "OmegaT doesn't
                    work with MS Office" that I hear.

                    Marc
                  • Samuel Murray
                    ... See here, isn t a Trados uncleaned file already segmented by sentence? Then OmT would simply have to recognise the segments (based on the Trados hidden
                    Message 9 of 21 , Mar 9, 2005
                    • 0 Attachment
                      Marc Prior skryf om 11:36 AM op 09/03/2005:

                      > * Trados TM support (TM files format)

                      > The real need is for support the "Trados uncleaned file format". Now that
                      > OOo supports the hidden text character property, It should, I think, be
                      > possible without too much programming effort to generate output in the
                      > form of a bilingual OOo file that only need be saved as MS Word within
                      > OOo in order to produce an uncleaned Trados file.

                      See here, isn't a Trados uncleaned file already segmented by sentence? Then OmT would simply have to recognise the segments (based on the Trados hidden tags) and voila, you have sentence segmentation in OmT to boot.

                      Samuel
                    • Samuel Murray
                      ... On a slightly different note... one obstacle in the way of general acceptance of OmT is the fact that you have to install and configure Java before you can
                      Message 10 of 21 , Mar 9, 2005
                      • 0 Attachment
                        Marc Prior skryf om 11:36 AM op 09/03/2005:

                        > Maxym already has the road map all neatly planned. My guess is that
                        > he'll tear his hair out when he sees the results of the poll. :-)

                        On a slightly different note... one obstacle in the way of general acceptance of OmT is the fact that you have to install and configure Java before you can use OmT. Now I have seen a genealogy program called GenJ which is written in Java, which has an applet version that runs inside any browser that uses Java 1.4 or higher.

                        Would such a thing be possible with OmT? I have no idea how much work it would be or how possible it would be, though. The startup file could be an HTML page that spawn two windows (one with the main window applet and one with the match/glossary window applet), or alternatively both could simply run in a single window. Most [Windows] users with browsers have a Java enabled browser, and most newer versions have Java 1.4 or higher.

                        Any comments from those in the know?

                        Samuel
                      • Rodolfo M. Raya
                        On Wed, 2005-03-09 at 11:36 +0100, Marc Prior wrote: Hi, Some comments below. ... I don t know what you mean with indexing problems, but I can tell you that
                        Message 11 of 21 , Mar 9, 2005
                        • 0 Attachment
                          On Wed, 2005-03-09 at 11:36 +0100, Marc Prior wrote:

                          Hi,

                          Some comments below.


                          > * XLIFF support (TM file format)
                          >
                          > This sounds unglamorous bt should solve a few other problems, like the lack of
                          > one-to-many translations. But how much work is involved in switching OmegaT's
                          > internal TMX file to XLIFF, I don't know. I suspect that there are indexing
                          > problems.



                          I don't know what you mean with indexing problems, but I can tell you
                          that much, much work is involved in the change. Definitely not a short
                          term task.


                          > * better TMX support (TM file format)
                          >
                          > Upgrading to TMX 1.4b should really be no problem at all.


                          I still don't understand why the program is restricted to TMX 1.1 . Can
                          anybody explain this?


                          > * Trados TM support (TM files format)
                          >
                          > Does anyone know anything about the current Trados TM format? Isn't it
                          > encrypted?



                          Yes, it is encrypted, but Trados is now able to export as TMX 1.4 . If
                          OmegaT is improved to support TMX 1.4 it will be automatically able to
                          reuse Trados TM


                          > The real need is for support the "Trados uncleaned file
                          > format". Now that OOo supports the hidden text character property, It should,
                          > I think, be possible without too much programming effort to generate output
                          > in the form of a bilingual OOo file that only need be saved as MS Word within
                          > OOo in order to produce an uncleaned Trados file. I've been meaning to figure
                          > out what such a file would need to look like (not difficult), but haven't had
                          > time.



                          An alternative:

                          - Write some routine to extract the hidden text from the open office
                          file and create a TMX version.
                          - Translate the OpenOffice document normally, using the TMX created in
                          previous step
                          - Deliver an OpenOffice document or a .doc exported from OpenOffice.



                          > * Direct MS Office file format support (project files)
                          >
                          > We all know that this isn't going to happen.


                          Well, this can actually happen if there is a filter for XML files and
                          you deal with MS documents saved in XML files. Complicated, I know, but
                          a possible solution.


                          > * RTF support (projet files)
                          >
                          > More realistic than direct MS Office support but not really any improvement
                          > over converson via OOo, in my opinion.



                          Many people still insist on using MS Office. Many people simply ignore
                          that OpenOffice is quite good at converting MS Office documents. Many MS
                          Office documents can't be properly converted to OO format. These are
                          some interesting reasons for supporting RTF.

                          On the other hand, I can tell from experience that writing a filter for
                          RTF format is a nightmare.


                          > * generic XML support (projet files)
                          >
                          > My vote. :-)


                          I would vote for this too, but I only had one choice and decided to vote
                          for another option.

                          This is a must. Many tools are slowly adopting XML. If OmegaT supports
                          XML properly, it will be able to support many formats in the future.
                          Simple example: Abiword uses XML

                          BTW, a real XML filter should be used for XHTML. OmegaT currently can't
                          really be used for translating XHTML files without destroying them. :(


                          > * internal segmentation function
                          >
                          > Nice but involves a fairly major rework of OmegaT.


                          This is what I voted :)


                          Suggestion for the future: please implement support for SRX. See http://www.lisa.org/blogs/index.php?id=16


                          This is also something for medium or long term. Better not to do
                          something quick and dirty.


                          > * internal spellcheck
                          >
                          > Should be very easy to do with Aspell, and it's already on Maxym's list.
                          >
                          > * better glossary function
                          >
                          > Would be nice, but if you make good use of the "Find" function, the glossary
                          > becomes less relevant. I also have a hunch that large glossaries may involve
                          > performance issues that we haven't yet noticed.


                          May be very important if the user wants to verify that the terms in the
                          glossary and not something else were used in the translation.


                          > * scriptability
                          >
                          > I discussed this with Maxym once (adding bindings to external scripts), but we
                          > didn't reach a conclusion on how it could be done.


                          Is this really useful? Why can't scripts be kept outside OmegaT?

                          Regards,
                          Rodolfo
                          --
                          Rodolfo M. Raya <rmraya@...>
                          Maxprograms


                          [Non-text portions of this message have been removed]
                        • Marc Prior
                          Hi Rodolfo, ... much, much work is involved in the change. Definitely not a short term task.
                          Message 12 of 21 , Mar 9, 2005
                          • 0 Attachment
                            Hi Rodolfo,

                            At the risk of starting another long discussion:

                            > I don't know what you mean with indexing problems, but I can tell you that
                            much, much work is involved in the change. Definitely not a short term task.
                            <

                            I mean that the internal TMX file is a hash table. That would not be the case
                            if XLIFF were used as the internal memory file.

                            > I still don't understand why the program is restricted to TMX 1.1 . Can
                            anybody explain this? <

                            No one has offered Maxym enough money yet.

                            > Yes, it is encrypted, but Trados is now able to export as TMX 1.4 . If
                            OmegaT is improved to support TMX 1.4 it will be automatically able to reuse
                            Trados TM <

                            That's what I said. Trying to support proprietary formats when there's no need
                            for it is a waste of effort. When it's justified, OK.

                            > - Write some routine to extract the hidden text from the open office file
                            and create a TMX version.
                            > - Translate the OpenOffice document normally, using the TMX created in
                            previous step
                            > - Deliver an OpenOffice document or a .doc exported from OpenOffice.

                            "Uncleaned Trados files" actually necessitate two functions: those of
                            receiving and delivering uncleaned files. Your routine describes the first.
                            Delivery is actually more likely to be needed, however. Both should be fairly
                            easy to implement.

                            > > * Direct MS Office file format support (project files)
                            > >
                            > Well, this can actually happen if there is a filter for XML files and you
                            deal with MS documents saved in XML files. Complicated, I know, but a
                            possible solution.

                            We discussed this at length on the development list last year. Several
                            possibilities were considered, including:
                            - using the OOo filters (incorporating a transparent Star Basic routine, as I
                            described in my earlier message), i.e. making the current manual procedure
                            automatic
                            - using Abiword's command-line converter
                            - using Upcast and Downcast
                            - supporting Microsoft's own, new WordXML format

                            > Many people still insist on using MS Office. Many people simply ignore that
                            OpenOffice is quite good at converting MS Office documents. Many MS Office
                            documents can't be properly converted to OO format. These are some
                            interesting reasons for supporting RTF.

                            Is conversion via RTF really much better than conversion via OOo? In my
                            experience, both are very good, but neither is perfect.

                            > This is a must. Many tools are slowly adopting XML. If OmegaT supports XML
                            properly, it will be able to support many formats in the future. Simple
                            example: Abiword uses XML

                            Agreed. Though supporting Abiword is more a proof of concept than anything
                            else; I don't anticipate much demand for it as such.

                            > BTW, a real XML filter should be used for XHTML. OmegaT currently can't
                            really be used for translating XHTML files without destroying them. :(

                            I think the dev team is aware of this problem.

                            > > * internal segmentation function
                            > This is what I voted :)

                            Why is there so little interest in the external segmenting utilities?

                            Marc
                          • Samuel Murray
                            ... Does Abiword use the same filters as OOo? I didn t know that Abiword has a command-line filter. I better check this out. Abiword GUI crashes on my PC,
                            Message 13 of 21 , Mar 9, 2005
                            • 0 Attachment
                              Marc Prior skryf om 2:08 PM op 09/03/2005:

                              > We discussed this at length on the development list last year. Several
                              > possibilities were considered, including:

                              > - using Abiword's command-line converter

                              Does Abiword use the same filters as OOo? I didn't know that Abiword has a command-line filter. I better check this out. Abiword GUI crashes on my PC, but perhaps the command-line works fine.

                              Samuel
                            • Dmitri Gabinski
                              Why only new features, not improving the existing ones? I find some problems with OOo file formats: 1) While Writer and Impress documents are generally treated
                              Message 14 of 21 , Mar 9, 2005
                              • 0 Attachment
                                Why only new features, not improving the existing ones? I find some
                                problems with OOo file formats:

                                1) While Writer and Impress documents are generally treated well, Calc
                                files (of both 1.x and 2.0 format versions) are corrupted, I mean, the
                                resulting files that open as weird Writer documents. At the same time,
                                no error messages appear during translation or compilation.

                                2) Very often are problems with Writer and Impress documents obtained
                                by conversion from MS Word and PowerPoint respectively formats: while
                                they can be easily translated and even compiled without errors, the
                                resulting files often do not open.

                                So, before introducing new file formats, it's wise to polish the
                                existing filters.

                                Dmitri Gabinski
                              • S. Tomaskovic
                                ... Not at all! I am frequently using Ben s macro. However, I am somewhat reluctant to install Tcl/Tk on my PC because it involves different components, and if
                                Message 15 of 21 , Mar 9, 2005
                                • 0 Attachment
                                  > > > * internal segmentation function
                                  > > This is what I voted :)
                                  >
                                  > Why is there so little interest in the external
                                  > segmenting utilities?

                                  Not at all!
                                  I am frequently using Ben's macro. However, I am
                                  somewhat reluctant to install Tcl/Tk on my PC because
                                  it involves different components, and if it's only
                                  needed for the segmentation of a document, Ben's macro
                                  does the trick for me.

                                  I too voted for the internal segmentation function.
                                  The problem with any external utility is that you
                                  can't change the segmentation once you've loaded it
                                  with OmT. I am actually hoping that internal
                                  segmentation will allow that, maybe even split/join
                                  segments.

                                  Just my 2 cents.

                                  Sonja ;)






                                  ___________________________________________________________
                                  Gesendet von Yahoo! Mail - Jetzt mit 250MB Speicher kostenlos - Hier anmelden: http://mail.yahoo.de
                                • Dmitri Gabinski
                                  ... Far not ANY browser supports Java 1.4. Well, in fact, mainly the latest Opera versions (if downloaded as with Java packages) do, if we speak about
                                  Message 16 of 21 , Mar 9, 2005
                                  • 0 Attachment
                                    Samuel Murray:

                                    >that runs inside any browser that uses Java 1.4 or higher.

                                    Far not ANY browser supports Java 1.4. Well, in fact, mainly the
                                    latest Opera versions (if downloaded as "with Java" packages) do, if
                                    we speak about Windows. And for all the rest, you still have to
                                    download Java. But who needs a system witout Java ;-)

                                    Dmitri Gabinski
                                  • Marc Prior
                                    Samuel, Abiword s command-line functions are among the best-kept secrets in the community. I understand there are two ways of using Abiword for conversion from
                                    Message 17 of 21 , Mar 9, 2005
                                    • 0 Attachment
                                      Samuel,

                                      Abiword's command-line functions are among the best-kept secrets in the
                                      community.

                                      I understand there are two ways of using Abiword for conversion from the
                                      command line. There is a built-in function in Abiword. Its syntax is

                                      abiword -to=[ext] [filename]

                                      So to convert the file test.doc to Abiword, you cd to the directory containing
                                      it, then type

                                      abiword -to=abw test.doc

                                      There's also a plug-in that does the same thing, but supposedly better. I've
                                      never tried the plug-in. I tried the built-in functionality last year with
                                      mixed results. Can't remember what worked and what didn't. I tried it again
                                      briefly just now and it worked OK (abw <-> doc), but I didn't try to find its
                                      limitations. The "doc" output is in fact RTF, but that's true of every other
                                      filter I know with the exception of OOo's, and given Abiword's restricted
                                      function set is hardly material.

                                      I stopped experimenting last year as soon as I found out that it was possible
                                      to execute OOo functionality from the command line; there didn't then seem to
                                      be much point continuing to experiment with Abiword.

                                      You'll have to dig around the 'net for documentation of the Abiword
                                      command-line functions - it isn't very accessible.

                                      I don't think Abiword uses the same filters as OOo; in any case, it doesn't
                                      support anything like all the same functions as OOo, although Abiword has
                                      added a lot of functions in the last three years or so. However, there's a
                                      lot of cross-pollenation between the OOo, KWord and Abiword teams, so I don't
                                      know the current state of play.

                                      Re Abiword crashing, I've never found it particularly stable. No more stable
                                      than MS Word running on Crossover, and nowhere near as good as OOo.

                                      Marc
                                    • Samuel Murray
                                      ... I ve just given AbiWord s command-line conversions a spin (on Windows 95). It handles uncleaned files well, but only if the format is very, very plain.
                                      Message 18 of 21 , Mar 9, 2005
                                      • 0 Attachment
                                        Marc Prior skryf om 3:26 PM op 09/03/2005:

                                        > Abiword's command-line functions are among the best-kept secrets in the
                                        > community.

                                        I've just given AbiWord's command-line conversions a spin (on Windows 95). It handles uncleaned files well, but only if the format is very, very plain. It completely loses images and watermarks. Text in textboxes disappear, although the textboxes remain. The conversions didn't take a lot of time, though (as compared to opening AbiWord or OOo, opening the document in it, and converting it).

                                        Samuel
                                      • Marc Prior
                                        ... although the textboxes remain.
                                        Message 19 of 21 , Mar 9, 2005
                                        • 0 Attachment
                                          Am Mittwoch, 9. März 2005 16:01 schrieb Samuel Murray:
                                          >
                                          > It completely loses images and watermarks. Text in textboxes disappear,
                                          although the textboxes remain. <

                                          This is probably par for the course with conversion within Abiword, and not
                                          necessarily a reflection upon the command-line option.

                                          Marc
                                        • Rodolfo M. Raya
                                          ... Hi, ... Thanks for explaining this. I don t intend to open another discussion about XLIFF. ... The best reason :) OmegaT code is stored in Sourceforge;
                                          Message 20 of 21 , Mar 9, 2005
                                          • 0 Attachment
                                            On Wed, 2005-03-09 at 14:08 +0100, Marc Prior wrote:

                                            > Hi Rodolfo,

                                            Hi,


                                            > > I don't know what you mean with indexing problems, but I can tell you that
                                            > much, much work is involved in the change. Definitely not a short term task.
                                            > <
                                            >
                                            > I mean that the internal TMX file is a hash table. That would not be the case
                                            > if XLIFF were used as the internal memory file.



                                            Thanks for explaining this. I don't intend to open another discussion
                                            about XLIFF.


                                            > > I still don't understand why the program is restricted to TMX 1.1 . Can
                                            > anybody explain this? <
                                            >
                                            > No one has offered Maxym enough money yet.


                                            The best reason :)

                                            OmegaT code is stored in Sourceforge; Sourceforge allows requesting
                                            donations via PayPal. Why not request voluntary contributions to be
                                            distributed among the developers?


                                            > Is conversion via RTF really much better than conversion via OOo? In my
                                            > experience, both are very good, but neither is perfect.


                                            Sometimes RTF is better, specially when there are images used as
                                            watermark in the file. There are still many features not supported by
                                            OOo.

                                            The biggest disadvantage of RTF is that you need to know the encoding of
                                            the file.


                                            > > This is a must. Many tools are slowly adopting XML. If OmegaT supports XML
                                            > properly, it will be able to support many formats in the future. Simple
                                            > example: Abiword uses XML
                                            >
                                            > Agreed. Though supporting Abiword is more a proof of concept than anything
                                            > else; I don't anticipate much demand for it as such.


                                            Abiword is just an example. You can also think on DocBook, which is more
                                            important.



                                            > > > * internal segmentation function
                                            > > This is what I voted :)
                                            >
                                            > Why is there so little interest in the external segmenting utilities?


                                            I have not tried the external segmenting utilities. Never downloaded
                                            them.

                                            I'd like to see sentence segmentation incorporated in the application
                                            just for compatibility with other tools.

                                            In the second half of this year we will add support for SRX in
                                            Heartsome's tools and it would be great if OmegaT and our tools could
                                            share segmentation rules.

                                            I'm thinking on contributing the SRX module to the XLIFF-Tools project,
                                            http://xliff-tools.freedesktop.org/wiki/ , to be used with the filters
                                            we open sourced. Perhaps someone from OmegaT can join the project and
                                            help.

                                            Regards,
                                            Rodolfo
                                            --
                                            Rodolfo M. Raya <rmraya@...>
                                            Maxprograms


                                            [Non-text portions of this message have been removed]
                                          • Henry Pijffers
                                            ... What I heard is that the OpenOffice developers are reworking their filters into a separate module. Perhaps OmegaT could use that? ... I reckon having to
                                            Message 21 of 21 , Mar 10, 2005
                                            • 0 Attachment
                                              Marc Prior wrote:
                                              >
                                              > * Direct MS Office file format support (project files)
                                              >
                                              > We all know that this isn't going to happen. It's a huge amount of effort. The
                                              > best we can hope for is to integrate someone else's routine into OmegaT as a
                                              > plug-in.
                                              >
                                              What I heard is that the OpenOffice developers are reworking their
                                              filters into a separate module. Perhaps OmegaT could use that?

                                              > * better glossary function
                                              >
                                              > Would be nice, but if you make good use of the "Find" function, the glossary
                                              > becomes less relevant. I also have a hunch that large glossaries may involve
                                              > performance issues that we haven't yet noticed.
                                              >
                                              I reckon having to explicitly "Find" stuff is much slower than having an
                                              automatic glossary function. And perhaps caching can help performance a lot.

                                              > * scriptability
                                              >
                                              > I discussed this with Maxym once (adding bindings to external scripts), but we
                                              > didn't reach a conclusion on how it could be done.
                                              >
                                              I know of at least one Java-based project that can execute JavaScript
                                              scripts.

                                              Henry
                                            Your message has been successfully submitted and would be delivered to recipients shortly.