Conversation
📊 Tachometer Benchmark ResultsSummaryduration
usedJSHeapSize
Resultscreate10kduration
usedJSHeapSize
filter-listduration
usedJSHeapSize
hydrate1kduration
usedJSHeapSize
many-updatesduration
usedJSHeapSize
replace1kduration
usedJSHeapSize
run-warmup-0
run-warmup-1
run-warmup-2
run-warmup-3
run-warmup-4
run-final
text-updateduration
usedJSHeapSize
tododuration
usedJSHeapSize
update10th1kduration
usedJSHeapSize
|
80efc74 to
ea827e9
Compare
|
Size Change: -59 B (-0.08%) Total Size: 78.4 kB
ℹ️ View Unchanged
|
da6d806 to
2bfe8b0
Compare
948a4e4 to
091a3ab
Compare
6f8b8ef to
0a4c379
Compare
There was a problem hiding this comment.
Lot of work went into this, great job!
However, I'm not in love with merging this as-is, entirely due to the state of Vitest, not any of the work here.
- It seems slower across the board
- It's an extra 100MB+ of
node_modules& an 800MB in~/.cache/ms-playright(tried using local Chrome install but I think that'd require different config/setup, it doesn't seem to function properly out-of-the-box) - It's a hell of a lot more dependencies
- We've had to switch to some Vitest-specific APIs. While the "lock-in" doesn't look to be too rough, we are switching from runner agnostic (or meant to be runner agnostic) tools to something that's specific & will need to be changed if we need to migrate again in the future
- We've encountered a fair number of Vitest bugs in the process of this PR
I suppose most importantly, I don't see a clear benefit to switching at the moment? It's something newer and that has some value, and we do shave a bit of config, but browser testing seems pretty well established, as are our tests, so I don't see any obvious possibilities this opens up. Karma, while deprecated, is super stable & doesn't seem to be holding us back at all.
I know Vitest's maintainers have been hard at work improving things, and with a bit of luck some of this will be addressed soon-ish, but at this particular moment it seems like a rather large regression instead of a clear improvement.
That being said, I don't want to hold this back at all; if people want this then let's go forward with it.
I agree, merging this feels like a regression, all though I am convinced that we should eventually move to a new tool because of karma being deprecated and it blocking upgrading our dependencies. I am going to explore the landscape a bit more, maybe there's a middle ground here. Ideally
This is more because
This was part of my intention, I wanted to uncover bugs so we can help the vitest people with testing browser mode as I am really looking for a community project to step into the shoes of karma.
Well, sadly, the setting doesn't work 😅 I agree with you, let's hold off on merging this, I will surface this with the vitest people so there's information sharing. |
|
+1 on waiting on some |
Oh fair enough, wasn't quite sure if it was sinon or vitest that was the problem there.
Huh. I thought the flag would act the exact same there. Admittedly I can't even run |
97ebbf3 to
5dc412d
Compare
5dc412d to
2bf8e23
Compare
41d9a27 to
e6a4080
Compare
5fed9f3 to
a2dcfc6
Compare
Requirements before merge
(done)testsoptionsproperty names in tests with regards to manglepnpm test:vitest:minwork, this somehow makes two tests failtscto work again... Broke with bump of types/node, this is part ofpnpm lintnpm run test:mochaplaywrightnpm run lint && npm run test:unitFollow-ups
sinonsinon-chai