For a non-English user, they would probably prefer asking questions by their first language. There might be a per-language general categories for them.
For a Language-specific question, like “How do I get some cjk-specific feature work?”, they might go to some cjk category to ask such a question.
This is a good idea! My main concern, however, is moderation. We cannot moderate questions that are not in English (or German, French). This means that for each other language category, we would need at least two trusted moderators from the community. If we can come up with these, we can give it a try!
Does Discourse support user language as a first class feature? Otherwise, if the forum ends up supporting more than one language, it could become hard to maintain the product of languages * categories manually.
This is a really constructive first post, great job! I also think a meta tag would be useful here.
I agree with the idea of languages + categories, and if I recall correctly, other communities like Blender also only have one single dedicated space for each language.
But moderators have to regularly monitor all the new posts, right? I’m personally not sure if I can do this, but as an on-demand moderator, I think I can.
It looks like Discourse has an extensive list of what can be tracked. IIUC, this should provide me with a setup where I get notified about every post for a specific category/language. Although I haven’t read everything carefully yet. But if it’s true, then I can give it a shot! :)
My first language are not English, but, I think separating forum by language are not the best idea. We have google, deepl for translation. So nowadays it’s not a big problem. But keeping forum in English will help find answers quicker because everybody knows that everything are in English
I would agree that using your native language is preferable, always.
However, I don’t agree that having a forum cater to every language is a good idea. It hurts discoverability (I.e. google search or forum search), fragments the forum, creates duplication of topics, and produces a larger amount of administrative overhead.
Indeed, having and not having per-language category has its pros and cons. In both cases, there will be some problems/downsides. But IMO allowing other languages in posts will result in net gain, because:
people that don’t know English won’t hesitate and can fully explain the question/issue in details that can be clearly understood by others that know the OP’s language;
this will make the community more inclusive and diverse, and as a result Typst can become more popular quickly (which can increase the number of contributors and sponsors, which will accelerate its development).
This is also important because the documentation doesn’t yet support other languages (other than creating a separate docs website for a specific language). And a lot of people won’t even consider trying something new if it doesn’t support their native language, specifically the web/CLI interface and documentation (and also the Discord server, for the most part). That is, unless they absolutely have to learn this (depends on person’s goals and intentions).
In programming, there is an idea: “don’t refactor too early”. Perhaps it is good to apply here.
There may be some need for language specific categories, but it’s likely better to wait until the need truly arises after seeing patterns. For example, if administrators start seeing a large and consistent number of posts from active users in a certain language, and perhaps about a particular topic, then creating a category (or similar) could be a good step at that point.
But in the current situation you won’t ‘start seeing’ posts in languages other than English, as this is an English-language forum, so people just won’t post here in languages other than English…
I think a lot of usability of this forum could be lost if we allow discussions in other languages than English. I think the whole point is the all content is written in a language most users understand - and that is English. This is the benefit of both the people asking and the people answering. I’d really hate seeing multiple active subforums in different languages. They would all contain a lot of content not accessible to the majority.
Will throw an idea I had yesterday: if it’s possible, maybe create multiple instances of the forum. So there will be either forum.typst.app or en.forum.typst.app or forum.typst.app/en and also, for example, ru.forum.typst.app or forum.typst.app/ru.
This will remove the issue of needing a separate category/tag for a language. Basically isolate different languages completely. When you then search something in your native language in a search engine, it will show you the results from the (Discourse) instance in your language. In each instance, you obviously can duplicate questions from other instances, but better yet when answering in the appropriate language also link a discussion from another instance, if it exists.
So, imagine someone asked a question in the English/default instance. Someone else that don’t speak English (or couldn’t find that post) but does speak Russian asked the same or very similar question in a Russian instance. Then people who answer to that Russian post should cite/link the first post from the English instance for better discoverability (and cross-instance high cohesion, something I remembered from the GRASP design).
In any case, I’m voting for having a multilingual forum in some way shape or form. We previously already had at least one instance where a Russian-speaking person tried to ask a question (more like discussion/offer) in English, but it was really hard to understand what was being said. So instead we switched to Russian and then quickly started to understand one another (after asking some more questions). This, however, from the very start raised an issue that not everyone knows Russian and not everyone could translate Discord messages into English. But as you can imagine, having a conversation (trying to understand the OP) and at the same time trying to manually translate everything into English is very close to mission impossible. I believe in a per-language instance, this wouldn’t be an issue.
Having multiple instances definitely requires more servers and probably more moderation overhead. But if moderators (for that language) are found (and spinning up additional instances is possible) then I don’t think there are any issues with this approach.