Adding Localization/Translation

[Koto13]

Hi,

Why not add other languages ​​to Alfred in this case French.

 

If you are busy, why do not we appeal to members to help you in this task.

Thanks.

[mcskrzypczak]

Yeah. There were no translations in v1, but maybe Andrew could think about letting localize Alfred v2? I am sure many people would appreciate it.

[Tyler Eich]

This would be difficult, considering the smallness of the development team. Maintaining separate localizations would be time-consuming.

[jibbe]

Yes, I agree.

Maybe we could do a "participative" or "collaborative" translation through the forum ?

[mcskrzypczak]

Maybe we could do a "participative" or "collaborative" translation through the forum ?

 

IMO that is the only option to make it easy for us as well as for Andrew and Vero.

[Andrew]

Hi Chaps,

 

Once you start to release in multiple languages, it opens the floodgates to needing to be translated into all languages. Internationalising an app takes a HUGE amount of commitment from the translators, and as a new release would have to wait on the slowest translator, this would SIGNIFICANTLY slow down Alfred's very fast and iterative release cycle, not to mention add an additional burden on me and Vero managing and verifying translations.

 

To put things simply... If you love how fast paced Alfred's development is and like my iterative approach to releasing early and often, then multi-language is simply out of the question.

 

Hope that clears things up!

 

Cheers,

Andrew

[Tyler Eich]

Hi Chaps,

 

Once you start to release in multiple languages, it opens the floodgates to needing to be translated into all languages. Internationalising an app takes a HUGE amount of commitment from the translators, and as a new release would have to wait on the slowest translator, this would SIGNIFICANTLY slow down Alfred's very fast and iterative release cycle, not to mention add an additional burden on me and Vero managing and verifying translations.

 

To put things simply... If you love how fast paced Alfred's development is and like my iterative approach to releasing early and often, then multi-language is simply out of the question.

 

Hope that clears things up!

 

Cheers,

Andrew

Makes sense to me. Anything that slows down development is bad as far as I'm concerned

Keep up the incredible work and rapid updates!

[Vero]

Following on from what Andrew said above, your best bet is to create custom searches for your local language sites and changing keywords to be localised. The application itself, the support and documentation will remain in English only for the foreseeable future.

 

I'll be keeping an eye on this thread, and will lock the discussion if necessary.

[jibbe]

Thanks, explanations are limpid !

It's OK for me

[Koto13]

Hi Chaps,

 

Once you start to release in multiple languages, it opens the floodgates to needing to be translated into all languages. Internationalising an app takes a HUGE amount of commitment from the translators, and as a new release would have to wait on the slowest translator, this would SIGNIFICANTLY slow down Alfred's very fast and iterative release cycle, not to mention add an additional burden on me and Vero managing and verifying translations.

 

To put things simply... If you love how fast paced Alfred's development is and like my iterative approach to releasing early and often, then multi-language is simply out of the question.

 

Hope that clears things up!

 

Cheers,

Andrew

 

 

It is obvious that two is not simple and I doubt that this is not the priority. It's okay.

 

[cortig]

I don't know if it is doable, but could the localization be provided as some sort of a plugin containing the corresponing .lproj?

Then the plug-in would have to be updated by whoever provided it, without slowing down the releases. You also wouldn't take the blame for any lack of quality in the localization work…

Just a thought :-)

 

 

          Corentin

 

[Koto13]

@cortig Good idea

[face]

I understand your point of view.

But, please, take a look at 

http://crowdin.net

1 Password Team for example uses that service to translate their great app (4 version).

It's an incredible and simple tool that helps a lot developers and not interfere or slow down developer work.

As said take a look please at that and rethink about it

I'm here to translate your invaluable app in italian ...

[Vero]

I understand your point of view.

But, please, take a look at 

http://crowdin.net

1 Password Team for example uses that service to translate their great app (4 version).

It's an incredible and simple tool that helps a lot developers and not interfere or slow down developer work.

As said take a look please at that and rethink about it

I'm here to translate your invaluable app in italian ...

 

I appreciate your point of view and your input in two threads; English wasn't my first language either as I grew up raised in French, and I understand that not everyone is necessarily completely at ease with English.

 

Regarding the 1Password comparison, the (very lovely) 1Password team consists of 40 people, while ours is a team of two (with the help of jdfwarrior in the US and our community, of course) so we simply do not have the bandwidth to manage translations.

 

There are numerous translation services out there; I've used some of them, as well as managed a team of in-house translators, in a previous job I had and am aware of the time it takes. Once the translators have done their part of the work, every single word needs to be reviewed and assessed to make sure that the translator's perception of the word in English is correct in the local language (Three translators may translate the same thing in three different ways). In itself, this would add extra weeks to every release of Alfred.

 

Translating the app would also open the door to requests for support in non-English languages, which we can't provide, as Google Translate usually garbles up the simplest sentences. I pride myself in providing the best customer service possible, and this would be near-impossible with non-English support questions.

 

As such, we still are not planning to translate Alfred, as our priority continues to be developing and releasing new features quickly and efficiently, rather than add a two months delay every time we want to do a release.

 

As mentioned before:

1. Create custom searches for your loved non-English sites

2. Change keywords in system prefs, etc to local language keywords

3. If your English is better than your friends', be their guide and tutor

[Alan He]

1. Alfred is great app

2. crowdin is great website to do i18n

3. I think alfred should consider it , I am very happy to do Chinese translation.

[cortig]

11 hours ago, Alan He said:

1. Alfred is great app

2. crowdin is great website to do i18n

3. I think alfred should consider it , I am very happy to do Chinese translation.

 

Crowdin, OneSky, Transifex, there is no shortage of collaborative localizations platforms these days.

I've localized a number of apps before and I’d be more than happy to help out fora French localization of Alfred as well.

The question of the localization of Alfred has been raised years and years ago, it’d be nice to see the app finally accessible to non-English speakers too 🙂

 

Corentin

[deanishe]

1 hour ago, cortig said:

Crowdin, OneSky, Transifex, there is no shortage of collaborative localizations platforms these days.

 

They don’t actually solve any of the problems Vero mentioned above. They might make managing a team of translators a bit easier and faster, but all of Vero’s points still stand.

 

1 hour ago, cortig said:

I've localized a number of apps before and I’d be more than happy to help out fora French localization of Alfred as well.

 

You've translated apps or you've actually managed the localisation of apps?

[cortig]

47 minutes ago, deanishe said:

You've translated apps or you've actually managed the localisation of apps?

 

Both. I’ve also worked on localizations locally (from A to Z on my own Mac, which was a lot easier to do when iLocalize was still around and working) and through online platforms.

I understand very well both the technical and logistical challenges, but there are ways to somewhat ease the pain.

Véro raises some very valid points, but I’d like to add that many non-English speaking users don’t have the same expectations when it comes to localizations: They can’t use the app at all when it’s not in their language, so getting the app with a localized UI is a huge thing for them. They don’t necessarily demand (expect) support to be in their own language.

Many apps worked around the issue by stating (on their site or even in the UI of the app) that support was only provided in English.

As for managing the localizations in different languages, it can indeed be a challenge. It works best if someone on your team understands the languages, or if you can rely on a “lead” volunteer localizer to approve the strings. 1Password was mentioned in previous messages and that’s how they manage it (and I happen to be one of the “approvers” for the French localizations of 1Password.

 

As far as I am concerned, the biggest challenge for the localizations (especially for an app as popular as Alfred) is not to deal with localizers or their localizations, but to make sure the app itself is structured to be localizable and deal with difference in lengths between languages (Véro should be able to attest to it… French is notoriously longer then English for instance).

 

Again, I’m not trying to pressure our friends at Alfred in any way or to argue about how hard or easy it might be. I’m just saying: I understand your points, and if you ever want to reconsider and localize the app, I’d be more than happy to provide all the assistance I can 🙂

 

Corentin

 

[deanishe]

2 hours ago, cortig said:

Both.

 

Nice. Some very cool apps in there. But the best part is Josy's 2004-vintage, dialup-friendly website

 

2 hours ago, cortig said:

but to make sure the app itself is structured to be localizable and deal with difference in lengths between languages

 

Yeah, internationalising software is a huge PITA to do properly, especially if you didn’t build it that way from the start. The difference in the lengths of text is one of the simpler—albeit ever present—issues to deal with.

 

2 hours ago, cortig said:

there are ways to somewhat ease the pain

 

The problem is, it's only "somewhat". Regardless of how much you or I or anyone else chips in to do the actual translation, it's still a bunch more work for Vero coordinating the translation effort (taking away time she currently spends on support), and it would fundamentally change Andrew's development process.

 

As a one-man dev team, he can have a new build of Alfred available for download within minutes of deciding it's good to go. Inserting a dozen or more translators and proofreaders into the development–release cycle gives you a turnaround time of days to weeks instead.

[cortig]

17 hours ago, deanishe said:

Nice. Some very cool apps in there. But the best part is Josy's 2004-vintage, dialup-friendly website

 

Josy is old school for sure, but she’s an AMAZING proofreader. Nothing gets past her.

[cortig]

17 hours ago, deanishe said:

As a one-man dev team, he can have a new build of Alfred available for download within minutes of deciding it's good to go. Inserting a dozen or more translators and proofreaders into the development–release cycle gives you a turnaround time of days to weeks instead.

 

A number of developers in these situations make some updates available immediately and add point-point release updates with the localizations once they become available.

In most cases though, there isn’t all that much that’s needed for the localization updates and the localizers can easily keep up. Alfred has a string and devoted user base. I’d be surprised if you couldn’t find a good team of volunteers for all languages. In addition, having multiple localizers implies that you’re likely to almost always have someone available to cover up for a quick update when needed. That’s really not the aspect that would worry me the most.

Alfred is faaaaaarrrrr from the only one-man dev team app I’ve worked with. The first app I ever localized for instance was FinderPop from Turly O'Connor and back then, I had to redo most of the localization by hand through RedEdit for every update and we never had much of an issue (tools improved with MacOS X though 🙂 ).

 

Workload-wise, I’m more worried about the structure of the app. Not everything is in .lproj right now for instance. I also haven’t looked in the .nib, but I that’s also a frequent point of concern (using dynamic fields instead of fixed size etc.). The app might need some up front restructuring to ease the localization process in the long term, but once that’s taken care of, it could be fairly painless.

 

Corentin

[deanishe]

56 minutes ago, cortig said:

Alfred has a string and devoted user base. I’d be surprised if you couldn’t find a good team of volunteers for all languages.

 

And I’d be surprised if they stuck around. Look at all the abandoned libraries and workflows. I see no reason to think the community’s translator pool would be any different to its developer pool: let’s face it: translating is far less interesting than coding.

 

You're not really addressing Vero's statement that the Alfred team simply don't have the bandwidth. The most significant change since then that I'm aware of is that jdfwarrior is no longer involved, so the team is actually smaller than when she said that.

[Vero]

I've been following this conversation and just wanted to confirm that my last statement earlier in this thread still stands. Translating the app is only the very tip of the iceberg, as it then opens the door to requiring all documentation and support to be multilingual.

 

I appreciate the enthusiasm of some users who would like to help, but this is a much bigger beast than asking a handful of volunteers to translate a few words every so often. 

 

If we took the steps needed to translate, proofread, document and support Alfred in all languages, you could add an extra zero at the end of the price of the Powerpack license. It simply wouldn't be feasible without paying numerous salaries, taking days to respond to support queries rather than minutes, and taking months between every minor release. This simply isn't the path we are taking with Alfred.

 

Cheers,
Vero

[cortig]

14 minutes ago, Vero said:

Translating the app is only the very tip of the iceberg, as it then opens the door to requiring all documentation and support to be multilingual.

 

Out of the big ones like Adobe, Microsoft or Apple, support is rarely available in multiple languages (and when they do, it’s sometimes horrible — I’m looking at you Microsoft!!). It happens, but it is far from the rule (I don’t even think OmniGroup offers support in anything other than English). Many apps also don’t provide localized documentation. 

 

As I mentioned a while back, I wish the app could provide some sort of plug-in/extension for the localization though. That would shift the entire issue on the community of users: Zero liability or responsibility for you — blame the volunteers for not updating the plug-in when it happens 🙂

 

Anyway, again, I very much appreciate all the work you do on Alfred and I’m just here to say that if you ever want to reconsider, ping me and I’ll do everything I can to help.

 

Corentin

[deanishe]

7 minutes ago, cortig said:

and when they do, it’s sometimes horrible — I’m looking at you Microsoft!!

 

Microsoft's English-language support is also horrible, tbh.

 

7 minutes ago, cortig said:

blame the volunteers for not updating the plug-in when it happens 🙂

 

Doesn't work, tbh. The Alfred team still gets the support requests, and sometimes the blame, too, I'm sure.

Edited July 20, 2020 by deanishe