This site’s categorization by language doesn’t seem to be useful

Here’s another thing to throw into the mix:

This plugin isn’t available on our current hosting plan, but there’s a chance it might be in the future once it gets more stable – and it’s possible we’ll upgrade our plan in the future.

3 Likes

However, imho, the point is not the number of moderators for each language category. Even if we have enough moderators (the process requires at least two people willing and capable to follow the new category). The point is the number of people willing to use their native language on this forum.
We can have dozen of moderators for a specific language, but if people still prefer to use English (maybe for better chances to get an answer), we can’t do too much in order to increase traffic in such category. On the other hand, as said, if someone ask a question in a language category, and nobody speaking such language is able to help, again, honestly I have no time to use any kind of auto-translation system in order to help people speaking a language that I don’t understand (it’s already a lot if I understand English :sweat_smile:).
So. If people want to open a new language category, we should not discourage them. If people are not using a category, it doesn’t depend on us.

2 Likes

Bump.
Could we have some data about language category utilization?

For instance. https://discussion.fedoraproject.org/t/tastiera-non-funziona-per-alcuni-secondi-dopo-la-sospensione-del-computer/69591
Here someone translated to English the original post. Commendable initiative. But :thinking: why translating to English a post in the specific language category? Again. It’s clear that a post in English has more chances to get an answer. So, again, maybe language categories don’t work here. Then, do we have some data?

However. It is also clear that language categorization prevent us to create other categories. And even if I don’t see hypercategorized forums as useful, I can admit that we have a little issue.

3 Likes

Well, you can see the counts on the category page here:

cfdb945ff5ce27f69129692e5a8f1e9c6869c865.png

I guess I can ask discourse if they can provide per-category view statistics. Is there anything else you’d be interested in?

I guess another metric might be: number of views for the top post in each language. Those are, right now:

  • Spanish: 1.8k
  • Italian: 892
  • Traditional Chinese: 656
  • Simplified Chinese: 2.1k
  • Persian: 1.6k
  • English: 80.6k

I still think Language specific categories is not efficient use of the resources, and it limits the exposure of topics.

Because:
Computer Generated Translation is fairly usable. So for members having specific domain knowledge still can offer help to foreign language posts.
(To avoid potential fall outs, sometime myself will clear state I am using Translate Tools in the replies, etc )

If we put Language as a tag, ASK should still be able to provide predefined views, so that only specific language posts are shown, just like viewing a Category.

That’s what the Multilingual Plugin does — but that’s not currently available to us. (If we did that, it’d also be nice to use the Translation Plugin, including Language detection.

Of course, one can look at things by tag now — like #firefox topics (or just firefox, which takes you to the same place) — but that doesn’t AFAIK let you then filter by other tags, which is limiting.

1 Like

I guess it’s time for a straw-poll. This is non-binding (that is, the result will inform our collective decision, not dictate it). But I’d like to get a read on what people are thinking overall. I’m making this a three-week poll because it’s a big deal. Note that these responses are not anonymous.

I am okay with any of these (multiple choice)
  • Keep language categories as we have them
  • Drop language categories, introduce others (explain below)
  • Combine all non-English categories into “Ask in Any Language”
  • Drop language categories, make site English-only
  • Drop language categories, make one “Ask Fedora” category and accept occasional non-English posts.
  • Other (explain below)

0 voters

I think the best option is:
  • Keep language categories as we have them
  • Drop language categories, introduce others (explain below)
  • Combine non-english categories into one “Ask in Any Language” category
  • Drop language categories, make site English-only
  • Drop language categories, make one “Ask Fedora” category and accept occasional non-English
  • I do not have a strong preference.
  • Other (explain below)

0 voters

My usual note: with any of these that allow other languages, we’ll still need to have mods for non-English languages (until automatic translation etc. are possible to make it possible for everyone to interact with all languages)

Combine non-english categories into one “Ask in Any Language” category will be dificult to moderate.

Can we contact Discourse to make a exeption with us? and install the plug-in GitHub - paviliondev/discourse-multilingual: A Discourse Plugin that makes it easier to administer a Multilingual Forum.

Regards.,

Also I should note: I have an opinion, but this is one of those things where my opinion counts as “regular user”. So please take my opinion with that weight.

I want this site (and Fedora in general) to be welcoming to all, and that includes people for whom English is a challenge — or just plain not their first language. But I worry that having prominent categories which are almost ghost-towns actually is actually worse.

That’s why — although it’s not my preference — I’d be okay with “just drop them”. We might actually be better serving people by referring them to external local communities.

My thinking is that one combined category would

  • make it clear that we do have a place beyond English
  • spread things out less so at least there’s activity
  • possibly invite folks who want to use a language we don’t currently have represented.

… and if in the future one of those languages does become active to the point where it’s thriving (perhaps: at least half as active as Ask in English?) we could “graduate” it to its own level.

On moderation, I’m at least somewhat persuaded by @sampsonf’s note that we can use translation sites to at least see if something is spam, off-topic, or out-of-line in tone. It definitely would be nice to have moderators fluent in the most common languages, though.

@hhlp, I don’t think we can get an exception. Particularly, I think that plugin is pretty heavyweight and would likely be ongoing work to make sure stays working rather than just a one-time “hey, install this please.” (I think this same logic would apply if we were self-hosting, too — it’s the kind of thing I’d be wary of committing us to.)

But we are actively looking at upgrading the subscription — it’s just at the level where it’s a big request, and needs to go through a lot of compliance checking and so on, and I need to really sell it to higher-level budget deciders. This seems reasonable to me, but it is also slow and a lot of work.

2 Likes

I am in favor of being inclusive. So although I think we should replace the language categorization, I would prefer that we make it “Ask in Any Language”, rather than just “Ask in English”. I agree with @mattdm that having the current categorization may actually be hindering inclusiveness, though I understand how it came to be this way through the wish to make it inclusive. I agree the strongest with his point #3:

  • possibly invite folks who want to use a language we don’t currently have represented

Because right now there is no place for people with a language preference other than the 6 languages that are currently categorized.

I understand that there is a concern about moderation of foreign languages, but I am not convinced that this is a real issue. If the concern is that there is a possibility that a spammer could possibly post inappropriate content in a foreign language, I would say that this could already happen today. Is there any mechanism that prevents a user from posting a Russian spam message in the “Ask in English” category? Has that been an issue?

In addition, I think there is a pretty broad user base already present on the forum that could quickly identify “bad” posts in many languages. I, myself, can read multiple languages but I don’t have the time nor inclination to become a moderator for an entire language category. However, it wouldn’t be much trouble at all for me to flag a post as “spam” if I come across it in a language I can read. I imagine there are many people on this forum who can say the same.

2 Likes

Yes, switch to “Ask in Any Language” do not make moderations harder then the current arrangement.

Having said that, I think it is still good for ASK to have Moderators covering as many languages as possible - such that when there are language specific issues, we know who we can speak to when needed.

Mmh, if you can post in any language in the same category, the concern is all about aesthetic. It could look ugly to see a list of titles in different languages all in the same place.
It could be like (well, not really) a mailing list where everyone write in their language.
Now we have 5 “allowed” languages. If your language is not here, you are “forced” to use English. But after, when you are allowed to use any language? But this is only a concern. Let’s see how it goes.

In any case, I think that it would be nice to inform people about local community channels, when they exist, would they be forums or chats (Telegram/Matrix/IRC), like the ones listed here https://fedoracommunity.org (yes, this site should be eventually updated to the current status), but this is another story.

I agree with this 100%.

In addition to everything pointed out by @guiltydoggy, there are many other Linux forums that successfully help people in other languages without dedicated native language speakers or moderators signed up to help.

I personally have helped many people solve problems in languages I can’t read.

While it is true that many people won’t break out a translation app to participate, we don’t need everyone to be willing to do it. It only takes a handful of people who can communicate natively or are willing to use a translator.

the primary difference is this:

  • without moderators, people may or may not engage with languages they don’t speak—there’s no guarantee because there’s no one who has volunteered to take on the responsibility of looking after these languages
  • with moderators, there are at least 2 (if we say we want 2 mods per language) people who have agreed to take this responsibility. So whether others do or do not engage, at least the moderators will. It just ensures that there’s someone accountable.

If someone can somehow guarantee that someone or the other will engage with languages they don’t speak for sure, so that queries do not go unanswered, there’s no need for the mods. I just don’t see how this can be done.

I only suggest this because FOSS communities tend to have a high turnover rate. People join and leave all the time, people get busy all the time. The only way of ensuring that something is going to be looked after is find a few individuals who agree to take on the responsibility. It’s how package maintenance works, it’s how QA works, it’s how infra works, it’s how the magazine work—there needs to be a core team of people who will prioritise whatever responsibilities they’ve agreed to take on. It doesn’t usually work when we go down the “someone will do it” route. We’ve seen that again and again in the community over the years.

…Is it really the same, though? Let me ask: Are there English-language mods for Ask? Are there people who have taken on the responsibility of ensuring that someone engages with every post made in English?

If there are, then I’m curious how they keep up with the load.

If there aren’t, because they’re not needed / because engagement isn’t a problem for posts made in English, then at the very least that makes this entirely unlike packaging, or infra, or the magazine, where there are people on the hook for every commitment, and nothing happens without those people.

Ask is just not the same thing, because it’s user-driven, not moderator-driven. The site will still experience activity with or without the participation of any of the responsible parties. Which is completely untrue of packaging, infra, and the magazine.

The entire justification for the language sections seems to be predicated on the assumption that posts made in other languages will by default be ignored unless there’s someone who’s responsible for “handling” those posts. And honestly I agree with that assumption, while non-English posts are ghettoized into language subcategories.

But are we so sure that non-English posts would be ignored by the entire user community, if they were posted into the same category/categories as the English-language posts? People seem to engage with those posts plenty. To assume that they’re only doing so because the posts are in English, and that not a single one of them would do the same for non-English posts… feels like a bit of an unproven conjecture, at a minimum.

There are, yes—we don’t necessarily engage with each post because with English most get some sort of response, but when a post is flagged etc., we actively look at what needs to be done. Again the difference is that most mods (even ones that are meant to be mods for other languages) do speak English and so the load is very well spread out.

Well it’s all conjecture. The assumption that people will engage with every single non-English post is also an unproven conjecture.

We won’t know until we run an experiment. We know that there will be two possibilities:

  • people do engage with every non-English post
  • people do not engage with every non-English post.

If the former happens, all’s well and good. But what if the latter happens? Are we willing to allow for the possibility where some non-English speaking users will not be responded to, making them feel (even more) left out? So, if that does happen, what is the plan?

The site shows the number of views & responses on each thread. The existing mods can watch that and if a non-english post sits for more than a day with no response then they can reach out and respond (even if a translator is used) if someone has not already done so. Does not seem to me to be rocket science to possibly even set up some kind of alert if desired for posts that seem to sit stagnant with no response at all.

This is true of any post in any language. No post is guaranteed to get a response in any language, including English.

That being said, if we look at forums of other Linux distros, we can see that while there is certainly less attention paid to non-English posts, most do get responses. Further, the same is true here, even with moderators in place.

Again, having moderators doesn’t guarantee that either.

This doesn’t seem logical to me. The fact that their is high turnover is the reason that relying on a small group of dedicated people doesn’t make sense. Instead, make those posts more visible to everyone and rely on the community to help.

Maintenance, package management and the magazine all seem very different from an internet forum…

I guess my counterargument to this is that it works well for many other Linux forum communities, what makes this one so different from all of those?

It isn’t fully conjecture. It is based on seeing what has worked well for others.

Since we are already “making most non-english speakers” feel left out since we don’t have a place for them to post, this seems like a non-issue. Even if it turns out to be an issue, we can do what every other organization in the world does when a plan doesn’t work. We can make a new plan.

For what it is worth. I would be glad to commit to spending time answering and/or moderating non-English language posts.

To be fair, this has already failed. There are many unanswered posts in all language categories (English included). Have those moderators been held accountable for not answering those posts?

Persian - 8/72 (11%) have 0 replies
Simplified Chinese - 3/30 (10%) have 0 replies
Traditional Chinese - 6/13 (46%) have 0 replies
Italian - 6/26 (23%) have 0 replies
Spanish -23/115 (20%) have 0 replies

Even if we decide to keep the current language set up, then there needs to be a discussion on how to achieve these goals because its clear the current way is not.

2 Likes