Conversation
DeeDeeG
force-pushed
the
0.37.0-redo-for-PRing-to-master
branch
from
57e0ba0 to
d6983d7
Compare
confused-Techie
left a comment
confused-Techie
left a comment
There was a problem hiding this comment.
Thanks for the PR!
Code changes look great, and like exactly what we need to get everything working in the places we care about. But boy is CI unhappy here, failing on nearly everything. Seems this PR may also need some additional CI love.
|
Yeah, someone got CI running but not passing a while ago here at this fork... I do not believe we have been able to get a green / passing CI run at this repo for as far back as I can remember at this fork. None of the ones shown here are passing: https://github.com/pulsar-edit/github/actions |
|
Well @DeeDeeG you've got a good point. But based on the last couple of comments on #2 I mention tests are "behaving as expected" and mauricio called out a flaky test. So makes it hard to tell if we expect CI to be failing so aggressively. |
|
Dang, you're right... Perhaps this is all since Electron 30? Might be possible to put together a CI run representing the state of this repo as of Pulsar 1.130.x (and testing against Pulsar 1.130.x rather than latest Rolling Pulsar) and see how broken or not broken that is. |
ppm is still using npm@6 under the hood. That ancient-ish version of npm is still only aware of lockfile v1. Modern versions of npm strongly prefer to operate on lockfile v2 or v3. Luckily, lockfile v2 is a highly interoperable hybrid of both formats. Yes, it is brutally redundant. It makes for a huge, huge, huge lockfile if you have a lot of deps. That said, having a more-or-less fully interoperable lockfile between all of our tooling is well worth it, IMO. So I try to keep this repo on lockfile v2 if I can.
DeeDeeG
force-pushed
the
0.37.0-redo-for-PRing-to-master
branch
from
d6983d7 to
53dc645
Compare
|
Rebased over As mentioned above, some tests are still gonna be broken. @savetheclocktower has a point here:
So yeah, those are testing known broken/(subsequently intentionally disabled and marked-for-removal) functionality (see: #45 for context.) The rest of the CI failures seem to need deeper work. |
Fixes "features.isViewApiEnabled is not a function" error savetheclocktower had already bumped this on the v0.37.0-pretranspiled branch, so I'm giving co-credit. Co-authored-by: Andrew Dupont <github@andrewdupont.net>
DeeDeeG
force-pushed
the
0.37.0-redo-for-PRing-to-master
branch
from
32b31e3 to
ea41960
Compare
3 participants
Redoing the
v0.37.0-pretranspiledinto a state ready for merging tomaster...-pretranspiledtag, for context / documentation of what pretranspilation is / is about.)And an update to animportdeclaration that I think might be needed (?) since the move from originalwhats-my-lineto@pulsar-edit/whats-my-linev0.37.0-pretranspiledbranch, I had merely misplaced it in my rebasing the first go around... I've now force pushed to reflect that bit having been done correctly by the original author@savetheclocktower, who did not in fact omit this bit originally!