Closed clayallsopp closed 10 years ago
Thanks @clayallsopp can you link the HipByte ticket also?
Also, once this is fixed, do you guys want to do a release?
:+1:
1.3.1
or 1.4.0
?
Hmmm, looking at the diff, I think the correct semver thing to do is 1.4? ("add functionality in a backwards-compatible manner" vs. "make backwards-compatible bug fixes")
Is there any changed API, or just new API? Cursory glance seems to indicate just new API.
Yup, all new APIs, nothing that breaks backwards compat. (sanity check == specs that have red in diffs)
Only iffy thing I can see is an option was removed from BW::Reactor::PeriodicTimer
07b45e3e854716c7bc48b435468a6bb209306118 279776c12f0e636a03ff91e9fcb575b8ec772594 6810eda26a743435d02fe1504b126199b29807c9
Yeah. No more common runloop modes bollocks in Timer.
from #272:
it seems the fix to RM-3 has broken this test: https://github.com/rubymotion/BubbleWrap/blob/master/spec/motion/core/ns_notification_center_spec.rb#L18
the
notified
object outside of the callback is different than thenotified
inside - you can verify this by logging thenotified.object_id
: https://gist.github.com/clayallsopp/5995437If you make notified an
@ivar
, it passes, but I'd say this is an unexpected RubyMotion regression. I've filed a ticket with HipByte.HipByte issue: http://hipbyte.myjetbrains.com/youtrack/issue/RM-215