summaryrefslogblamecommitdiffstats
path: root/2021/talks/omegat.md
blob: 10a1aca2ded29c961aae4f6349b9b4db4851bad9 (plain) (tree)





















































                                                                                                                                                                                                                                                                                                                                                                                                                                              
[[!meta title="Emacs manuals translation and OmegaT"]]
[[!meta copyright="Copyright © 2021 Jean-Christophe Helary"]]
[[!inline pages="internal(2021/info/omegat-nav)" raw="yes"]]

<!-- You can manually edit this file to update the abstract, add links, etc. --->


# Emacs manuals translation and OmegaT
Jean-Christophe Helary

Even if it is generally agreed that software localization is a good thing, Emacs is lacking in that respect for a number of technical reasons. Nonetheless, the free software using public could greatly benefit from Emacs manuals translations, even if the interface were to remain in English.

OmegaT is a multiplatform GPL3+ "computer aided translation" tool running on OpenJDK 11. That category of software is roughly equivalent for translators to what IDEs are for code writers. Casual translators can benefit from its features but professionals are the most likely to make the most use of OmegaT.

When Emacs, OmegaT and free software based forges meet, we have a free multi-user translation environment that can easily sustain the (close to) 2 million words load that comprise the manuals distributed with Emacs, along with powerful features like arbitrary string protection for easy typing and QA (quality assurance), automatic legacy translation handling, glossary management, history based or predictive autocompletion, etc.

The current trial project for French is hosted on 2 different forges:

1.  sr.ht hosts the source files
    <https://sr.ht/~brandelune/documentation_emacs/>
2.  chapril hosts the OmegaT team project architecture
    <https://sr.ht/~brandelune/documentation_emacs/>

The sources are regularly updated with a po4a based shell script.



# Outline

-   5-10 minutes: (brief description/outline)

In this format, I would show the basics of:

1.  using po4a to convert the texi files to the PO format
2.  using OmegaT as a team with 2 (or more) translators working at the same time
3.  using OmegaT features such as regex based string protection, legacy translation handling, autocompletion, QA, etc.

<!--
-   20 minutes: (brief description/outline)

With more time, I would discuss the same items, slightly longer, along with:

1.  a short description of how I came up with the regex used to identify non-translatable markers in the manuals, found here:
    <https://gist.github.com/brandelune/46faceb4c81bfee7c938282cc6dfe17c>
    and explained in French here:
    <https://goshikidai.blogspot.com/2021/08/#5695124486632169049>
2.  how all that team coordination is in fact a huge git job hidden from the users
3.  discuss what is called MTPE in the translation industry (Machine Translation Post Edit) and how that relates to/affects localization in general (cf. the recent LibreOffice documentation Japanese translation using MT)

-->

[[!inline pages="internal(2021/info/omegat-schedule)" raw="yes"]]

[[!inline pages="internal(2021/info/omegat-nav)" raw="yes"]]