Conversation
- Run in random order so no dependencies are uncovered - Reset sessions after each test - When the order includes patching BEFORE running firefox tests, a couple of blank keys are added which made the subsequent tests fail. This is now fixed by using a lax(er), rspec matcher
…mpliance, so we need to re-tighten this restriction
…melcasing won't work otherwise
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good
So I've learnt two new things @mrdaniellewis order random is a must in every project. |
@@ -43,7 +48,7 @@ def capabilities | |||
# This is required to make local drivers work exclusively with Safari TP | |||
# This is required in V13 of Safari as the driver there is notoriously flaky | |||
# In V12 it doesn't hinder it | |||
# We don't support Safari V11! | |||
# Safari V11 is unsupported. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I wanted to use Safari 11 in the Casebook smoke tests, as we still have some users. But yes, it was completely broken. Just ended up using 13.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Funnily enough safari have basically said v12 is also "unsupported" so technically both are unsupported. But then you get into a minefield.
The safaridriver v12 I consider to be more performant and accurate than the v13 one. Primary reason being the v12 driver was one that contained both JWP and W3C stuff, and I theorise (It's closed source), that there are fallbacks e.t.c. in there that make things "work better"
safaridriver v13 just fails and falls over so frequently it was a nuisance when I first joined, so I just reverted to using safaridriver v12 again.
I'm midway through trying to get safaridriver v14 to work reliably, but outside of using it sporadically, it's falling over a lot (Although I believe that may be more browserstacks end).
We do the random order in the unit tests, but not the feature tests 😬, probably should change that |
Updated some comment docs and release v2