On Wed, Jul 8, 2015 at 7:03 AM, Stefan Arentz <[email protected]> wrote: > So I did not take a11y into account. Well, I sort of did, but as you are > saying more patches are coming in. I kind of assumed that 'we will take > whatever we can get'. But if that does not work with a11y then maybe we need > to land those patches later so that we can land them fully localized? We can > always do small incremental updates after our v1 with a11y improvements.
I imagine a "we will take what we can get" approach will be fine with a11y (i.e., dusek), since I think where we're at right now is in pretty good shape. As long as he's aware of the freeze and doesn't get cut off without warning, we should be OK. Sounds like the next batch of PRs should address any remaining a11y strings that are needed [1]. > Emily ran into an issue where labels are sometimes unpredictable. Like when > they are dynamic. I wonder if we should start using accessibility > identifiers more consistently instead. AFAIK those are not end-user facing > and can be used to pinpoint specific elements without relying on strings or > localized strings. What do you think? I'm not sure what you mean by labels being unpredictable. They're dynamic, as you said, since they change based on the text content and app state, but the label should always be deterministic based on a given sequence of actions. If not, that seems like a bug. A similar discussion of accessibility IDs vs labels came up in KIF discussions. One of the KIF owners discourages using identifiers since KIF is supposed to be used from the user's perspective, so user-facing strings (labels) are preferable to magic developer strings (identifiers) [2]. I tend to agree with him, though there are some cases mentioned in that thread where identifiers could be useful. That said, if we're writing new tests and we need label changes, but we're string frozen, accessibility identifiers seem like a good fallback so that we aren't blocked. If we do that, we'll probably want to make a note to ourselves to switch from IDs -> labels after we ship in those situations. [1] https://github.com/mozilla/firefox-ios/issues/683#issuecomment-119566048 [2] https://github.com/kif-framework/KIF/issues/243#issuecomment-23363722 _______________________________________________ mobile-firefox-dev mailing list [email protected] https://mail.mozilla.org/listinfo/mobile-firefox-dev
