[Debate] Translation of Gladys into Chinese, do we have the necessary resources?

Development
1

Hello everyone!

Recently, someone on Github made a PR to add Chinese translation to Gladys.

I’m very torn about this addition, but as this is a subject we will probably face with other languages, I wanted to start the debate to discuss it with the community.

First, you should know that adding a language to a product is « not as simple as that », it’s not just about translating a translation file once and that’s it.

1/ Continuously adding translations / possible development blockage.

We develop features almost daily on Gladys, and modify translations almost daily as well. If we decide to manage a new language, we therefore need continuous work to add the translations of these new features.

This work takes time, time that will not be spent on other features. We will therefore lose development velocity as we add more languages, as we would have to ping the « referent » translator of each language for each PR so they can work on the translations. It’s simple but it will take time. There will therefore be fewer features developed per year.

The risk is to get stuck: what to do if we have a feature that needs to be released and the translation is not ready? So far, everyone spoke English and French in Gladys developers. Here, it’s a language that none of us master.

We can put anything (Google translation), but is that really the goal (to do it badly)?

Also, we need to spend time testing the screens in each language to make sure the translations fit on the screen, and make any UI corrections if they overflow.

2/ The documentation must also be translated

Similarly, every new feature developed in Gladys is explained in the doc. This is a major point of honor in v4: we spend time on the documentation, it’s as important as the product.

Writing this documentation takes time, for information for each set of features that I release, I spend about half a day writing documentation (EN + FR currently).

If we add languages, we will need to spend additional time writing the documentation in the added languages, and ping the referents of each language so they spend the same time in their language.

Same question, what to do if the language referent disappears? Do we abandon the translations? Do we do all the work ourselves with Google Translate?

My personal opinion

In my opinion, the project does not have the necessary resources to assume the workload induced by new languages. I feel that we are already behind on many subjects, there are many things I would like to do in Gladys but that I cannot do because of lack of time.

We are still at the very beginning of this v4, and I think we have every interest in investing the few resources we have in the development of new features and content on the site in FR/EN.

The mantra of Gladys v4 is « we do few things but what we do we do well ». Adding Chinese, and possibly saying yes to all languages afterwards (why not add German next? Spanish? Portuguese?), would be, for me, going in all directions without real vision and we would lose overall quality of the product.

It’s sad but I don’t think we can say « yes » to all PRs, we have to know how to say no when we’re not able to assume the responsibility of development afterwards.

To top it all off, in the case of Chinese, we should know:

  • That apart from this user who made the PR, there are not really any other Gladys users in China
  • Gladys does not work well in Chinese, the specifics of their alphabet break some parts in the product.

Tell me what you think. I know my opinion seems quite clear-cut but I think it’s a subject that deserved to be discussed anyway, I don’t want to come across as the big bad guy, I’m just trying to make a choice that will be good for the project in general.

@contributors I’m open to your opinion!

The PR for Chinese: Support Chinese language by bulolo · Pull Request #1178 · GladysAssistant/Gladys · GitHub
Github issue for Chinese on the site: how to add a cn language supported · Issue #75 · GladysAssistant/v4-website · GitHub

2

Given the arguments you bring, indeed it seems complicated to add any other language. If we add new ones, we should expect partial or inconsistent translations, in addition to a probable visual change, or one that is not adapted. On the other hand, for the documentation, we could allow other languages, but not everything will always be translated either.

We might consider a service to add translations proposed by the community, with an upload in Gladys, but I think we’re not there yet.

3

Yes! And given the little enthusiasm for this message, it reassures me that we are not at all at the stage of translating into other languages :slight_smile:

4

I responded with a simple and respectful message about the contributor’s work. We’ll see what they say :slight_smile:

Screenshot 2021-05-06 at 09.12.24
Screenshot 2021-05-06 at 09.12.241938×1438 350 KB

5

Hello,

I think it should have been accepted.
Gladys would have shone further and encouraged other translations (perhaps).

For my part, in all the international developments I participate in, I only capture translations in French and English. By default, all sentences, words are displayed in English if they are not found in the requested language.
Thus, other participants speaking the language contribute by translating.

6

The EN + FR market is 1.5 billion people who speak these languages.

We are still very, very, very far from exhausting this market (you are 200-300 on the v4), and we don’t have the means to launch into a new one.

As I said in my explanation, it’s not just about translating Gladys, it’s a whole set:

  • Gladys currently doesn’t even work in Chinese
  • we don’t have documentation in Chinese
  • we don’t manage devices used by the Chinese

Don’t forget that saying yes to something is saying no to other things. Days are very short.

For me, Gladys v4 is still far from being a complete product even just for the FR market (look at the list of pending PRs, GitHub issues, feature requests), launching into another market is the best way to do everything badly ^^

7

What is @bulolo’s feedback on your response?
Wouldn’t he have the ability and especially the knowledge to handle it?
If he proposed this PR, it means he knows what he’s talking about, right?

Would he be able to translate the documentation as well?

8

I understand the arguments @pierre-gilles but I agree with @luke

I think the PR should have been merged.
The English fallback is a good solution.

What will potential future contributors think?
What I would think as a developer is that it’s not worth investing time and contributing. It closes a bit a project that is open.

It’s really not an easy decision.

9

Open-source doesn’t mean we have to say yes to everything!

As I mentioned in my response, I encouraged them to maintain their fork. That’s what open source is all about: being able to fork at any moment if you want to add something.

It really comes down to resources. If we had unlimited time (and money), of course I would have said yes.

But it’s the opposite right now. We’re already behind on many topics, and there are dozens of issues/PRs that I find more interesting that are completely abandoned due to lack of time. Just look at the list of feature requests, issues, and PRs.

If it’s just the act of « rejection » that shocks you, I could have kept the PR open indefinitely without ever touching it, like many PRs out there.

After that, I encourage everyone who thinks I’m wrong to give their time to help us move forward and allow this PR to be merged. In that case, absolutely :slight_smile:

But as it stands, there are other topics that I consider to have higher priorities.

10

It doesn’t matter if you’re 1 or 100 developers, if you only have French speakers, there will never be other translations.
Few people are bilingual; I have this situation at home. If there were no French, I wouldn’t be a user (my children and my wife don’t understand English).

We’re missing out on users and possibly on a committed contributor.

That’s how I see it; I’m just expressing my opinion. :wink:

11

We are also depriving ourselves of users if we prioritize Chinese over other features.

New actions in scenes? Improve Zwave? New boxes on the Dashboard?

In fact, everything is prioritized, it’s not a « yes » or « no », it’s « what is the most important? »

In the French Gladys newsletter, there are 3200 readers and many have not yet switched to Gladys 4 because some things are missing, and that’s understandable as we are still at the beginning of version 4.

For me, I find this more of a priority :slight_smile:

After all, it’s not easy, and there are no right answers.

I would also like to have the ambition to launch in as many countries, but we also need to be realistic :slight_smile: