Re: State of l20n v3

[Стоян Димитров]

I am asking mainly because the docs are recommending it [1]. Still
I miss some of the functionalities from the learn by example page [2],
namely the multi-line interpolation and multiple variants.

P.S.
The l20n.org site is using some v1 beta [3].

___
[1] https://github.com/l20n/l20n.js/blob/master/docs/html.md
[2] http://l20n.org/learn/working-with-text-multiline-interpolation/
[3] https://github.com/l20n/l20n.org/tree/master/js

С

На 31.08.2015 г. в 11:49, Стоян Димитров написа:
> Greetings,
>
> What is the state of v3 of the project? Is it ready for production
> use? I plan to use it in a web page (html bindings).
> _______________________________________________
> tools-l10n mailing list
> tools...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/tools-l10n

[zbran...@mozilla.com]

Hi!

Sorry for late response. It's actually a really hard question for us right now, so I feel like all of us like to think we can avoid it ;)

On Monday, September 28, 2015 at 2:51:37 AM UTC-7, Стоян Димитров wrote:
> I am asking mainly because the docs are recommending it [1]. Still
> I miss some of the functionalities from the learn by example page [2],
> namely the multi-line interpolation and multiple variants.

Sure. We're aware that our current documentation and various tutorials are not in sync. I apologies for that.

> P.S.
> The l20n.org site is using some v1 beta [3].

Yup.

> На 31.08.2015 г. в 11:49, Стоян Димитров написа:
> > Greetings,
> >
> > What is the state of v3 of the project? Is it ready for production
> > use? I plan to use it in a web page (html bindings).

The current state is "in transition". Short explanation:

We have three versions of L20n library:

1.0 - the reference implementation of the original idea.
2.0 - the subpar port of the old FirefoxOS architecture to merge some l20n concepts of API
3.0 - the current implementation that builds on top of 1.0 and 2.0.

The 1.0 is the most complete implementation of L20n API and format so far, so if you need the full power of L20n, and you need it now, it's the best way to go for you.
Unfortunately, at the time when we wrote it we didn't really have a reference project to test it against, so we couldn't validate our API, performance etc.
In result, while complete in technical sense, I'd say that 1.0 is too complicated.

2.0 is not worth it, don't even try.

3.0 is bringing everything from 1.0 but using the lessons learned while working on our main driving platforms like Firefox OS.

If you are starting a project, I'd recommend 3.0. While it doesn't support everything 1.0 does, I believe it supports a lot and it has the best API.
We will be adding more features to it and eventually it should support pretty much everything that 1.0 did, but with better API and performance.

I know it's not ideal. Unfortunately, that's the nature of the project. All we can do now is push for getting 3.0 better and more complete which is what we are doing.

I can't even give you deadlines for when remaining features of L20n format will be in 3.0, although I expect a major increase in the coverage by the end of this year.

Greetings,
zb.