Migrating from Pootle to Crowdin

There is no an official feedback from Legal Department of Crowdin

Hi Team,
Here are the comments from our legal:Our legal recommends having 2 definitions of our terms of service clarified.

  • Content - any data and information, other than Client Data , available through Crowdin Services or contained within the structure of the System, articles, documents, brochures, presentations, pictures, images, audiovisual works, other informational materials owned by Crowdin.

  • Client Data - Files and any other digital data and information, including translations which are subjected to the Crowdin Services or otherwise inserted to the System by the Client.

This should make more clear who is the owner of the data, uploaded to Crowdin.

Other than that, according to the GPL license terms and conditions, the translated version of the Program has to be what they call “modified version”.

Modified versions are allowed and according to GPL will only require copyright permissions from the author (TYPO3). As we see it, the translators are those who modify the software by translating it, so TYPO3 has to give the copyright permission to translators in order to be able to translate the software.

Because translations belong to you, you can use them at your own discretion in modified versions under the same GPL license, we don’t claim any rights to your data. We act as a tool that processes your data, nothing more.

The state of our current translation service is not relevant in the selection of a new tool. I think that most people agree that we should look at another tool (or recent version of Pootle). We should look at the features we need, and the pros and cons.
I this regard it’s a disadvantage the we don’t control the availability of the translation service of Crowdin.

The degree of completeness of the translation in any language is a bit misleading. There is still a lot of data present from obsolete versions (both core and extensions). Translators always have to consider priorities and CSH texts (a lot of them and mostly huge) always had the lowest priority. With the likely removal of CSH this won’t improve. It’s also not very popular to translate labels of obsolete versions or obsolete extensions.

It’s bad that the registration doesn’t work.
A login system that works with other platforms could increase the user base, perhaps. If you want to contribute to a project it’s usually not too much trouble to register there also. In fact, it’s weirder to require accounts on other platforms. But that’s no issue if SSO with t3o will be working.

I mentioned it because it was listed under “Far future”. It’s rather essential imo.

I was referring to using translations. If an installation uses extension “Foo” the translations can be fetched and updated independently from any releases of that extension. TYPO3 itself has frequent releases, but extensions may very well only receive updates once a year. By separating the language files from the code packages translations can be updated very often.

True, but it would still be convenient. I’m not looking for a replacement for Pootle that has exactly the same features (and problems). There are plenty of thing that nag me in Pootle. It would be awesome if we find a tool that is also easier to use in those areas.

It was presented as a cool feature together with screenshots.

It’s probably very handy for people who find an untranslated text or a very bad translation. Currently there are few translators so in order to find all language labels you just have to work in an interface that lists the available labels.

Again, it’s not a case of ‘Pootle is perfect but currently broken’ and ‘let’s find something that is exactly like Pootle’. I’ve looked at the crowdin site, the list of features and the rest of the description. If we switch to another tool we’d better find something doesn’t have the quirks and problems that Pootle has.

See the list I made earlier about the selections translators and reviewers need to be able to make. Most are (somewhat) possible in Pootle, a few are not possible in Pootle currently but these are the nags that should be solved in an alternative.

If I take all the pros and cons into account then my conclusion is that the missing features in the interface, the uncertainty with an external service, potential legal issues (which have to be looked into), the functionalities which are/were put on the “far future” list still outweigh the advantages of crowdin. This can change in the future if enough of the cons are solved somehow.

I adopted the initial post: far future is for me Q1/Q2 2020

It’s not that hard: if one wants to contribute in some way to TYPO3 a single account with a trustworthy organisation should be enough. Preferably this would be the typo3.org account.
Currently this is already made impossible «you see a mistake in the documentation? you need a GitHub account.» «Great idea, can you add it to this google docs document?» «The results of this meeting are in this google document.»

Once there was OpenID which allowed you pick a provider to store your identity. There were a couple of independent OpenID providers which made it really nice service for easy logins. This changed a lot after the independent providers stopped, OpenID2 came (which was oAuth2) and only the giants which had other core services but also provide oAuth2 remained.
It would be great if you could login with t3o using another identity provider but we shouldn’t force people to create accounts with all kinds of service if they want to contribute.

Ah, the all-the-people-we-could-attract-from-github argument :slight_smile: There is a hook active that converts PRs on GItHub to Gerrit patches. I don’t have recent numbers but the last time I got data about it it showed that was used rarely.

To me, not being forced to have an account on a commercial platform is actually a -1 for crowdin.

Hey Jigal,

Thanks for all your feedback. The point is: for me the current situation is really bad, having an outdated software which is hard to use and has currently quite some issues.

All I want to achieve is having a better tool which makes it easier to work with and which provides devs and translators more support like machine learning, translation memory and so on.

The tool is not only great to use for me. In the last days multiple extensions have been translated by the community which was not done previously for many reasons. It is fun, having a translation memory of all extensions together and just clicking some buttons to have a good translation.

I finished today the export api to get the translation in the format we need it and can now provide something which is not possible Currently with pootle: translations for version 10.
I can already provide branching for extensions which was not possible before.
The crowdin team has an really great support team with feedback within minutes which helps a lot, again something we don’t have yet. The product itself is being developed further and features I needed have been built within days.

I get the point that you don’t like it, so I won’t argue about that.

Who then decides who is trustworthy. Taking your statement literally, you say that TYPO3 Assoc or GmbH is trustworthy while Microsoft (Github) is not. Please elaborate.

If you are a developer on planet earth who does contributions to open source projects you have a github account, simple as that. No matter the owner, Github is the place to be as an open source developer. It doesn’t mean that open source software development exclusively happens on Github, but the majority of software is on Github and Github is their home to the extend that issues, documentation and discussions happen there. Yes, there are other platforms and the more important they become, the more you should also have an account there. Contribution is not only doing patches, it’s also giving feedback, reporting issues and so on.

Well, I don’t have a time machine, neither do you. So unless we can change that, let’s put that aside. We are living in a different present right now.

That’s exactly my point regarding the typo3.org account. You have 5 or more options to use crowdin while you only have one (1!), that allows you to contribute at translations.typo3.org. And while my github account gives me access to the whole world of github and plenty of other services (blackfire.io etc.), my typo3.org account just gives me access to gerrit and the shop.

Indeed, like a github account for instance.

How convenient… Wait, aren’t you the guy who brought up the convenience argument in the first place to support your opinion against crowdin? So, you care about an easy way to contribute and handle a service. Of course you do, because all humans do. There’s your reason why very few people use github for pull requests to create patches on review.typo3.org.

I don’t have any numbers either but I have some experience with people who create patches via github. Quite a lot aren’t even aware that this triggered a patch on review.typo3.org. Therefore, those people don’t create an account, don’t interact on review.typo3.org with us and don’t push new patch sets. btw. as mentioned earlier: Being on Github doesn’t only mean that you can create PR’s with a hook. It means, that the whole development takes place on that platform. That’s the interesting thing. Getting feedback from people on site, not on another platform you maybe don’t know or don’t have access to in the first place.

Creating a PR on github (for TYPO3) feels like throwing a bottle mail in the ocean.

I am sure you didn’t mean to put that “not” in the sentence but this way I can only agree.

Just a small hint to

The documentation team keeps translations for documentation open, but because of too few resources they can not support to keep such translations maintained.

Still an option on central part for the documentation I’d say (get started, T3 explained, etc). We could need translated docs for customer, editor and decision makers

This topic was automatically closed after 14 days. New replies are no longer allowed.