Open himself65 opened 9 months ago
Thanks for opening up an issue.
Yeah, it seems like a bug. Can you dig into it?
exit status 0
is also weird.
exit status
0
is also weird.
Working on this
Working on this
From what I understand https://github.com/dai-shi/waku/issues/536#issuecomment-1966721276, dist/public/index.html
isn't created if /
is not isStatic
. So, I don't see how this happens. Hope you find a bug somewhere.
inside unstable_collectClientModules
From #573, what I'd say is it's not SSR (html generation) and it's a false positive (not leaking). So, I don't consider it's a bug per se.
However, I agree it's not a desirable behavior and I wish we could come up with a different solution and eliminate unstable_
.
btw, do you have a real issue with the current non-ideal behavior?
btw, do you have a real issue with the current non-ideal behavior?
Yeah, I will use getEnv
in the runtime, there's no env var during my build time
I see. It's unfortunate, but the current workaround would be add a check if such an env var is available or not (and do some fallback that would never be required in the runtime.)
Current status: https://github.com/dai-shi/waku/pull/573#discussion_r1514091960
Can we statically analyze the code?
I was just thinking about it. Analyzing statically is not possible with current architecture. defineRouter & createPages are designed to isolated from filesystem.
I hope there's a enviroment like WAKU_STAGE === 'build'
or globalThis.cleanup(fn)
to make sure I can disconenct the database or ignore some SSR
Try unstable_getPlatformObject
for now.
We might be missing something for emit...
.
See: https://github.com/himself65/waku/commit/5e8218f33fb1a7b9ec1ead5d9ce665f8199b75ac
Also exit code shouldn't be
0