1001817 - [email] remove unused localization strings from Firefox OS 2.0

Status: VERIFIED FIXED

(Firefox OS Graveyard :: Gaia::E-Mail, enhancement)

Comment 1

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

Attachment: pull request, v1

Comment 2

[Andrew Sutherland [:asuth] (he/him)]

Comment on attachment 8413164 [details] [review]
pull request, v1

Thanks very much for doing this, :aryx.  I've validated this all looks safe to remove, but am torn over whether we should keep the 3 minutes string until we'll absolutely sure we'd never turn that option on.  :jrburke, if you could weigh in on that one...

My methodology is effectively:
  git log -S <string to search for> -p apps/email
then usually searching for the l10n string in double-quotes to locate the static uses without getting distracted by the locales stuff.  (If I didn't find a static usage, I'd switch to single quotes and/or change to just the js dir and searching without the quotes.)

For general interest, I was wondering what your methodology is in case it can help us avoid causing this in the future, etc.

Comment 3

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

(In reply to Andrew Sutherland (:asuth) from comment #2)
> For general interest, I was wondering what your methodology is in case it
> can help us avoid causing this in the future, etc.

It's only plain simple testing and finding the labels in the app which I do for verifying that the German translation fits into the available area and the context. So I have a copy of the translation in which I prefix the ids of verified strings (so I can run compare-dirs from compare-locales to check how many strings out of the total have been verified) and can see the ones which I never managed to find (I lack the time to re-verify the strings in new releases). So I mxr-ed or searched on Github for the usage of the strings, similar to your methodology (plain, double-quotes, single-quotes). If that fails, I search for parts of the string if it seems possible that it will be the result of a concatenation in the code.

Comment 4

[James Burke [:jrburke]]

Comment on attachment 8413164 [details] [review]
pull request, v1

Looks good to me. I am fine with removing the 3 minutes value. 

I would rather be clean and have everything unused removed, and 3 vs 5 minutes for me is not enough of a win to provide it (I say as just a developer not a product/ux person). We should encourage people to have a life outside of email! :)

So I say remove it. We can add it back later if need be, and while we may not be able to reuse the same key (would need to change for all the related keys), I think that is fine. If there are fans of a 3 minute interval, then by removing it, by fake science laws that makes it more likely that it will be added later, so it is a win for them too!

Comment 5

[Andrew Sutherland [:asuth] (he/him)]

landed on gaia/master since consensus this was good as-is:
https://github.com/mozilla-b2g/gaia/pull/18707
https://github.com/mozilla-b2g/gaia/commit/32b59d53c9221dc2d2777aa4b5d0450dc9745a3c

Thanks again for the fix, :aryx!

Comment 6

[Andrew Sutherland [:asuth] (he/him)]

advancing to verified; this is not something QA can verify apart from noticing no new regressions, and this already had two pairs of eyes on it.