For someone who speaks English fluently, interface translations are confusing even when done well. I therefore use English interface even when software is available in Bosnian (e.g. LibreOffice, Microsoft Office).
Thanks for the comment @dzenanz - yes, we’ve heard from English-speaking users who grew up speaking other languages that they need English for their work anyway so they prefer to use the English interface. We are also hoping that some people find other languages convenient for day-to-day work and for learning the program.
Since documentations, slicer community,… are in English I find it convinient to also have the interface in English. But I think is a good idea, specially for new users that may be not familiar with 3D Slicer/medical imaging language.
I am a native french speaker and the challenge I see using scientific/technical software in my native langage is the limited resources available when looking for help:
the forum posts, tutorials and sometimes manuals are typically in english
the names of the functions in toolboxes and source code comments are in english
the literature is in english too…
Making the link between all these great resources and the translation is sometimes more time consuming and confusing than just learning the software in english. So in the end, I use the software in english but I realize other people may have different preferences and be intimidated (I assume you have identified a target population interested by the translations. students? clinical staff? others? people using specific procedures?).
As a suggestion, you probably thought about it already, but anything that could facilitate the link between the translation and the original langage may be useful in case a user is looking for help. These could include allowing both translation and original names in small in the menus, or a dynamic switch between langages, tooltips with original names, easy access to a dictionary with the translations, etc. Unfortunately, I have not seen that in other software (many need to restart to change the langage).
Thanks for the answers, it is very useful to see what are some common experiences.
I had similar experience with Hungarian translation of Microsoft Windows and Office. It was really painful to use them. However, in about 10-15 years the translations have become better (and probably people also learned and got used to Microsoft’s Hungarian terminology) and so the translated interface is just as easy to use as the English original.
I agree that getting information/support in English is easier, as magnitudes more information available in English. However, with Copilot, ChatGPT, etc. the gap is getting narrower. For example, I can ask any question about Slicer in Hungarian from Copilot and I get answer in Hungarian. The answer is clearly collected from English sources but it is provided in Hungarian. In the response the name of menus and buttons are translated, too, but unfortunately those translations don’t match the words that we chose to use in the actual Hungarian translation of the Slicer GUI. But maybe if we translated the user manual, too, then chatbots would start to use the correct words.
These are very good points. Adding a mode where the original English text was included in the translated text would be easy, but unfortunately at many places there is not enough space for double amount of text. Adding English original text in tooltips would be difficult, too, because there is no support for this in Qt (Slicer’s GUI toolkit). Dynamic switch between languages would be possible to implement (Qt provides the infrastructure for this), it would just require significant extra effort in all widgets to implement retranslation. Access to a dictionary may be the easiest to implement. We already have a glossary, which could be improved and maybe it could be used to enhance non-English conversation with chatbots.
This is a common issue even for much bigger projects including windows and office. I am not sure if there will be a solution that we can emulate. However, there is one thing, we can address and that’s the user guide.
I suspect once people knows what buttons and menu items to push, it doesn’t matter that much what it says on that button. But for that we need to have training materials that are in their own native language. Are there any plans to translate the user documentaitons to other languages, or is the plan to have people rely on services like google translate?
Yes, that’s an area we have been working on. So far there’s a SlicergGlossary project and French versions of some documentation and tutorials, but we have been discussing how best to approach it.
It depends in part on whether we continue to be funded on this work, but it’s a bit of an open question at this point if the automated tools will keep getting better without any effort from us or whether we’ll get better results by doing custom training based on our Slicer translations.
This is why we are interested in hearing from the community, especially people somewhat new to both English and Slicer, about what methods are working well for them and what is still needed.
I did ask students about this. They are first year medical students (6 years undivided program in Hungary). They told me that using Slicer in English is a valid option, provided, you already are familiar with the technical terms (), and the software (/interface).
() both the medical terms, which is not the subject of this discussion and the medical image processing-related terms
They told me that the “mixture” of the two would be a way they likely would take: begin to use, and learn in their native language, and then transition to English.
Also, we are planning to translate the help/docs, and (at least some of) the training material, and provide the community with the experiences, and some insights. We involved a linguist, and she will join us in LP next week
(Since Microsoft was mentioned earlier: as far as I know she did take part in the Hungarian translation process of Windows and MS Office, so that already is some solid background)
I don’t disagree - it’s amazing what can be done automatically.
One of the goals of this LanguagePacks project has been to constantly evaluate AI language tools and identify what still needs human effort to make Slicer more available and productive for people who can benefit from it. So far, any of the purely automated UI translation efforts have had mixed results, but they keep getting better.
On the other hand it sounds like many in the community here have generally just learned English in order to get their work done, and maybe that’s what people will end up doing. But we also want to test if having the localized version actually helps people accomplish what they need to do.