@craigmaloney @noelle I've not seen that. I just looked quickly at the code. Since speculation isn't desired, I won't. But more context would be useful.
@lauren @craigmaloney That is, unfortunately, all the context I have, except that I'm using glitch-soc and not mainline Mastodon. I've updated ruby and node to the versions that 4.1.3 wants, and I've confirmed they're working properly. I've done all the updates steps - in fact, I've done several of them several times. This seems to genuinely be a webpack problem, but I can't figure out where the problem is.
@noelle @craigmaloney OK, and since I'm using mainline, the differences are likely significant (or at least possibly so) -- so I probably can't be of much help on this specific case.
@lauren @craigmaloney In other circumstances I would say that you're right, but the primary developer of mainline, ClearlyClaire, is also the main developer of glitch-soc. >_>
when people have encountered this sort of thing in the past, from trawling the github issues for glitchsoc
@noelle @lauren @craigmaloney@octodon.social oftentimes it's because webpack ran out of memory partway through packing, meaning it wrote the file saying that it packed the thing without actually packing the thing
that's almost definitely where i've run into it too (trying to run masto on too small of a VM)
in that instance, RAILS_ENV=production bundle exec rails assets:clobber unwrites that file (well, it clobbers all the assets) so that webpack forgets its previous attempt and does the thing again
but you can fix that issue and still not get anywhere if webpack just doesn't have enough memory. NODE_OPTIONS="--max-old-space-size=350" or some other configuration may also be necessary if node or webpack are just not getting the resources they need