[bookbrainz-site] snyk-bot opened pull request #479 (master…snyk-upgrade-c82bd8a457a52d9f0af0f950f9d89865): [Snyk] Upgrade prop-types from 15.6.2 to 15.7.2 https://github.com/bookbrainz/bookbrainz-site/p...
[bookbrainz-site] snyk-bot opened pull request #480 (master…snyk-upgrade-ab71637327b70edad54688c8bf48293b): [Snyk] Upgrade connect-redis from 3.3.3 to 3.4.2 https://github.com/bookbrainz/bookbrainz-site/p...
[bookbrainz-site] snyk-bot opened pull request #481 (master…snyk-upgrade-bee51cd3c6853a5923dea1777d5e2eb2): [Snyk] Upgrade express from 4.16.3 to 4.17.1 https://github.com/bookbrainz/bookbrainz-site/p...
[bookbrainz-site] snyk-bot opened pull request #482 (master…snyk-upgrade-4943cf7d8cf239b8b61c91ceb074cd8d): [Snyk] Upgrade superagent from 5.2.2 to 5.3.1 https://github.com/bookbrainz/bookbrainz-site/p...
yvanzo: when would the search server return a 404?
/search/error/no-results.tt is supposed to come up if the server 404s, but we usually hit "No results found. Try refining your search query." in PaginatedSearchResults instead since it returns but with 0 results
So I'm not sure we ever hit that, or at least not for anything connected to no results?
sumedh joined the channel
killme has quit
killme joined the channel
alastairp
morning
ruaok: I might have deleted the messybrainz mapping temp tables from bono. I was using the musicbrainz database there to run the script and didn't realised that it _wrote_ to the musicbrainz_db database. I'm running create-pairs again now and it's reinserting
yvanzo: hi, where do you prefer musicbrainz-docker tickets?
ruaok
Did you run the script and it nuked the mapping?
Perhaps that needs to become a separate step now that the code is closer to production.
alastairp
yeah, I was just playing around and ran the script and saw the first log message "Drop/create pairs table"
does that mapping include FKs to the musicbrainz database?
if not, I wonder if a separate database would be safer? instead of a schema on the same database
or at the very least a separate configuration string in config.py, so that you can point it to either the same database or a different one
CallerNo6 has quit
jmp_music_
Morning!
ruaok
alastairp: it writes the tracks to a different schema, but the key point of the mapping is that it needs to join to many MB tables.
having it outside the DB would be a freaking nightmare. so, a schema is the best sort of compromise....
but I agree, making the drop table part separate is now the right time.
alastairp
OK, sure. If it has to join to those tables then it should absolutely be part of that db - I wasn't sure if that was the case or not
ruaok
and really given the nature of bono, there is no problem with you messing around dropping random tables. it can all be recreated.
right?
alastairp
agreed, especially given how easy it was for me to rerun create-pairs and fix it
in fact, it's finished. I could have got away with not telling you
ruaok
then I wouldn't have gotten an important nudge to make it safer. :)
MajorLurker joined the channel
what is your intention with the web_home volume on the develop.sh improvements?
alastairp
it's so that bash and ipython can write their history to /root
and when you run it again, your history is still there
ruaok
ah, ok.
its not obvious why that is there.
maybe add a comment?
alastairp
sure. to the volume definition, or to the mapping in the web service?
tbf, I added that change in a separate commit which explained this :)
ruaok
future reader of the file will not be reading the commit history...
MajorLurker has quit
CallerNo6 joined the channel
CallerNo6 has quit
CallerNo6 joined the channel
alastairp
in the general case, yes. I've found that I look at annotate/history a lot these days to work out reasoning behind a change, but I realise that not everyone does it
a backend issue it seems, gateways are properly configured, cert was generated
it hits 10.2.2.31:13048
ruaok
agreed. gateways are working as expected.
I found a shortcoming in a the dataset hoster -- it can't find its templates.
v6lur joined the channel
yvanzo
hi alastairp: I prefer it as Jira tickets
ptitjano has left the channel
alastairp
yvanzo: thanks. on what project? mbvm?
wow. I managed to consistently crash musicbrainz-docker postgres!
in the end it was because shared_buffers was higher than maximum ram the docker VM on mac had
sumedh has quit
iliekcomputers
when you wake up at 8:45 to realize you have a meeting at 9 😥
alastairp
15 minutes? luxury
Gazooo has quit
ruaok
gah. anyone have experience shipping package-data or using MANIFEST.in in setup.py
I can't get distutils to include my template files to save the life of me.
alastairp
urgh. sometimes, but it always involves poking something, trying it, trying again, throwing everything against a wall and going for a bike ride, etc
ruaok
shit, I have no access to a bike, but that is what I've been doing.
Gazooo joined the channel
strangely works when I use the package in a dev server. but not in docker.
alastairp
ruaok: in messybrainz-labs, is `python3 formats.py | psql` required>
?
I'm just updating the readme to use run.py
ruaok
actually, the messybrainz stuff is in flux right now.
alastairp
right right now? OK, it can wait
ruaok
thanks for fixing things but I really need to clean up more before tossing you into the mix.
not right right now, but soonish now.
alastairp
OK. Sertan wanted a copy of the mapping, so I was passing the repo his way
ruaok
oh.
alastairp
but maybe a dump of the mapping tables might be enough for now
ruaok
hang on.
that might be best.
alastairp
ok. I can do it, unless you have one that you prepared earlier
ruaok
because you need at least 64GB of ram to create the mappings. :)
hmm. ok.
actually, go ahead.
I had made a branch to move the mapping to LB, but I've realized that that branch had problems and I since made other changes that are current in the messybrainz-labs repo.
but its on the short list to move this code to the LB repo.