Please post a list of troubles users face on using translations, then solutions could be considered good or not ar even proposed new ones.
As it is now it looks you already have a goal and a timeframe to implement, just not the tech implemented. So make that it more clear and open for all.
As for the usual issues you considered:
“can’t find translation”
SalesAgility is to blame: the actual link is almost hidden on the main page. You are going to a store listing. Good. Hope is not the only place to post those links.
I had also posted translations zip in Sourceforge so people could download outside Croudin (or transifex that was the first platform used). I even have a site with guides and faq for translators). Its good to see things moving on your side with the store. Do not forget the wiki!
“have to go through an awkward process to contribute”
What?! Create an account, login, translate. No software install required, no code knowledge, no work loss.
Where is the problem?
“Trouble to install - language do not parse and blank screen”
Confirmed. Some languages breaks and do not parse on the first install. I consider this solved for all languages by controlling some problematic strings. But code should be cleaned to avoid that kind of problems
“Trouble to install - unzip problems”
Always a system issue, on permissions and code. Need to be simplified on the code side!
“Trouble to install - Local language edits are not implemented”
Need to rebuild and clean cache! Almost no software needs to rebuild just to use languages. This is a major issue and should be solved on code side
Translation size?
I’m sure someone has already complained about the translation size: 640493 words, 139 files on 266 folders, for a total files size of 1,44MB unzipped!
“not a complete translation”
Accepted. There is a place to collaborate, we need more collaborators… but where can you find an official link to collaborate here?
You want to go for an automatic translation system? Go for it, its a decision a company needs to take and are allowed to take.
Just not with me. I will not have my name as translator for my language if that language has also automatic translation. This is a red line for me and an end of conversation for me!
To consider your options, Crowdin allows different export methods
- Skip untranslated strings
If checked, untranslated strings will be absent from the exported file.
- Export only approved translations
Includes only texts that are both translated and approved.
This will require additional efforts from your translators to approve all suggestions.
So you can have an exported files missing the english lines so you can merge that with the autotranslated ones.
Then you can create your 100% autotranslated (that will include the approved lines) … and proper translators can have their own language packs not “all translated” but “well translated”!
do so in a way that takes little or no time to upkeep
provide support on how to install the packages
provide directions on how to contribute to these translations
This will solve the immediate and biggest needs that prospective SuiteCRM users have. An important note: we have seen first hand over the years that a very small percentage actually ever bring up any sort of problem. What happens is that when presented with a barrier like running into an installation problem they just bounce to the next CRM. So when we see a significant number of live chats about languages alone since we launched the store it means that it is something that is causing a significant to bounce.
Some of the other problems you fairly listed we will be able to help with including the repair & rebuild and file permissions issues some run into.
Once that is going that may end up being where other places, such as the wiki, end up linking to.
This is great info. Thanks for listing these possible problem points and for your help with all of this!
Jason, what is your relationship to SalesAgility at this point? (I ask with as positive expectation: despite my differences with you on the present topic, I regard any collaboration from you with this project as an excellent added value!)
as we discuss the integration of language packs into the Store, it occurred to me that language packs shouldn’t require a visit to any store, they should simply be selected from the SuiteCRM admin screens. We should just select the language(s) we want and Suite would download and install the packs by itself. This basically requires the Module Loader to be aware of the URL where it needs to fetch the packs, which the work we’re discussing here anyway.
Invaluable resources that we will help to bring to the forefront for people to find. Hopefully we can help bring in more contributors in the long term.
I don’t see any real major differences on the translation side right now. Just some small disagreements that we can all look beyond. This discussion has been really valuable.
I agree with you on the installation process and what we are talking about here is a big first step to that. We are working towards an in-app store which will alleviate many of the typical add-on install issues including having to deal with the module loader, file permissions, etc that typical users shouldn’t have to be experts in. Once we have the language discovery processed figured out here it will just be an extension of the in-app store to plug that into the install process.
Thank you each for a respectful and passionate debate.
As is often the case, there is merit on all sides.
From the project’s perspective we want to focus on the core application and to produce the best CRM we possibly can. We don’t want to focus on the ecosystem or to police it.
To that end we have partnered with Jason @ sugaroutfitters to provide the ecosystem store for SuiteCRM.
In the short to medium term we want to enable users to install plugins direct from the application and for those plugins to alert when updates are available. This will almost certainly be done through a single point and almost certainly through Jason.
So, it makes a great deal of sense if everyone harnesses their considerable energies towards that goal.
The vision is for a single point where high quality extensions such as language packs are available and manageable.
However, this is open source. We do not own the ecosystem. We’re very interested in a vibrant and high quality library of easily installable extensions but we cannot mandate how that will be delivered beyond facilitating a mechanism for the delivery of it.
I do not under-estimate or under-value the commitment that Horus68 has for the language packs and am in awe of what has been produced. Similarly, Jason’s contribution to open source, first with SugarCRM and now with SuiteCRM is considerable and of the highest quality.
You guys are all on the same side. I am not going to suggest a course of action. I sense that there will be a convergence of ideas and if I can help facilitate that then please just get in touch. You know where I am. In the meantime, please keep talking.
Thanks for chiming in. Apart from the new question raised by horus68 here, for me it all comes down to a few issues:
This new Store project and Jason’s efforts are very needed, very relevant and welcome; we agree on the usefulness of 95% of what is being planned to be done.
However, as a translator and translation user, nobody is going to convince me that adding automatically translated strings is a step forward for the project. I think it is counter-productive. I think we need to be very careful in how we communicate this to users, so that we are not simply fooling them. However, it seems this will go forward. I don’t object to that if it is a parallel effort, but not something to replace the better option (imho) that we currently enjoy.
I think Jason could prove the worth of his view before advancing full scale. Perhaps do one or two language packs and see how users appreciate them.
For me, the essential part would be to guarantee that the current Language Packs and philosophy is not hurt by this. This includes keeping the most knowledgeable and devoted people on-board (like horus68). I want to be sure that human-only translations will continue to be available, clearly labelled as such, and free from any noise. they could simply live on in Crowdin, even if they aren’t in the official Store. Can we get some assurance on this? It’s the least we can give to horus68, after all he’s done for SuiteCRM.
The most important thing for the SuiteCRM project, regarding translations, is to advance and complete all the human-translated packs. If I understand correctly, even Jason agrees with this; he just thinks there should be a temporary workaround before we can get it done. To advance this ultimate goal that everybody desires, horus68 has listed a few very practical things that could be done to unblock current delays. These should be given priority.
A big thank you to everybody in this valuable and respectful discussion.
I appreciate the passion, but I’m dropping the automatic discussion for now. I tried to make that clear in the previous posts by clarifying where my next steps are to help leverage and amplify horus68’s great work. Parallel work would be a wasted effort.
I just re-read the entire thread to make sure I understood everybody’s inputs until now. I see unanswered issues, but I suppose I can go into “wait mode” and, perhaps, as progress is made in the next few weeks/months it will be easier to see how the pieces come together.
Depending on how (technically) the different parts of translation are stored and retrieved, maybe there won’t be any need for duplicate work, and everybody’s concerns will be addressed.
Whatever system is chosen, it will have to keep track of which strings are translated automatically and which aren’t, so the basis will be there to serve the needs of both perspectives presented in this conversation.
Just to be clear, we are not going to touch the translation process that horus68 already has in place. It’s absolutely needed. We are only focusing on augmenting the process at this point.
Hundreds of reasons why. I haven’t forgot about this. We are working on processs, staffing, and a big new addition to the store platform which is critical to allowing us to focus on this and other items pending on our list.