Today I've attended a Webinar about the http://www.opentm2.org/ open-source version of the good ol' IBM TM2 translator's workbench. Very interesting. OpenTM2 is surely one of the most mature and efficient translation GUIs on the market (sure, after 20 years of development by IBM…). It's probably the tool I would use if I would be a translator…
But it's basically that: A tool for the translator. No collaboration or even TM sharing. Even using a TM on a network drive doesn't fully work apparently (strange errors coming from the M$ server…), so you'd have to send files and TMs around via mail.
But there may be a solution: Let's connect openTM2 to a TinyTM server! Suddenly the combined system would be able to compete with Logoport and other tools.
I asked Gerhard Fetz (IBM openTM2 architect) about the difficulty of this change and he answered that it wouldn't be easy. It seems that the translation memory component isn't very well isolated from the rest of the system.
=> We should check the code.
=> I can't really imagine that they didn't encapsulate TM lookup functions into a reasonable API…
Cheers!
Frank
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
Today I've attended a Webinar about the http://www.opentm2.org/ open-source version of the good ol' IBM TM2 translator's workbench. Very interesting. OpenTM2 is surely one of the most mature and efficient translation GUIs on the market (sure, after 20 years of development by IBM…). It's probably the tool I would use if I would be a translator…
But it's basically that: A tool for the translator. No collaboration or even TM sharing. Even using a TM on a network drive doesn't fully work apparently (strange errors coming from the M$ server…), so you'd have to send files and TMs around via mail.
But there may be a solution: Let's connect openTM2 to a TinyTM server! Suddenly the combined system would be able to compete with Logoport and other tools.
I asked Gerhard Fetz (IBM openTM2 architect) about the difficulty of this change and he answered that it wouldn't be easy. It seems that the translation memory component isn't very well isolated from the rest of the system.
=> We should check the code.
=> I can't really imagine that they didn't encapsulate TM lookup functions into a reasonable API…
Cheers!
Frank
I had a brief look when the project was launched, I will take a further look at the code base when I have some time