change listenbrainzspark to listenbrainz_ansh as well everywhere in that file
ansh
It did not work
lucifer
it should be listenbrainz_ansh not listenbrainzspark_ansh
ah actually nvm, you need to change the network to musicbrainz-docker_default
because your other containers are running ther
ansh
worked
I have now imported the data into spark. I don't have an idea on how spark works exactly, but how will the recommendations come up? Like will it process previous listen history? Or will it work on new listens
lucifer
imported the data - means you imported listen dumps right?
bitmap: I started looking at https://tickets.metabrainz.org/browse/MBS-6741 (IIRC one of the plans we had by adding the area_containment table was to be able to implement something like this) but I'm not sure what'd be a good way of doing it :) Maybe you have time at some point to brainstorm it a bit?
BrainzBot
MBS-6741: Include subdivision and country in the webservice when returning an area
reosarevok
I was thinking of an inc= for this, but we're not even passing includes when serializing the artist areas for example :)
Would it make sense to just do it always?
Pratha-Fish
Hey reosarevok I was wondering if I'd really need to setup the full version of musicbrainz-db for this particular project?
The only issue is that I don't have a ton of storage on my laptop. Plus, I wont have access to wifi for a week, starting from today 6PM to download the DB🥲
If it's necessary, I can try to get it downloaded somewhere on an external HDD maybe
reosarevok
Hmm. Wonder if it would be possible to make a sample + areas dump
Since you clearly don't need all the recordings for example, just areas and urls and l_area_area plus l_area_url I guess
Pratha-Fish
_exactly_
If there's a way, I would be happy to do it.
OR if possible, we could maybe setup a mirror on one of MeB servers. Especially given the fact that the actual bot would be running the same way
reosarevok
Yes, that's also an option
Ideally we'd eventually do that, anyway, but a dump wouldn't hurt...
There's the JSON dumps where area data is all separate, but given we want to run this from the DB, it might not make sense to develop using the JSON dump, even as a test :)
That said, it's only 32 MB for the areas so you should have a copy just in case
lucifer
reosarevok: it should be simple to dump the tables with COPY command or pg_dump fwiw. if you want i can do that
reosarevok
That'd be nice - I *think* the following list would be enough:
(ideally we'd filter link to only l_area_area and l_area_url linked entries to make it smaller, not sure how big that is... but maybe it's not too bad)
Pratha-Fish
lucifer: yes, it would be a lot of help if you could help me make the dumps.
Here's the steps that are apparent to me:
Log onto wolf, fire up psql, and find a command to dump the tables reosarevok mentioned. Then I can just copy the dumps locally with sshcopy
Do I have to do anything else apart from this?
ZaphodBeeblebrox is now known as ApeKattQuest
lucifer
Pratha-Fish: `pg_dump postgresql://musicbrainz:musicbrainz@127.0.0.1:5432/musicbrainz_db -a -t area -t area_type > dump.sql` so on, before each table name you have to specify -t
then yes copy it to using sshcopy and import in your database
I usually eat only stuff with at least 50% cocoa but I will also bring some milk chocolate for those who prefer it sweet.
mayhem
milka zartherb!
and yes, for the GSoc Summit bound batch, make it 50/50 milk/dark.
reosarevok
No white? dats racist
kellnerd
aerozol: Do you eat regular chocolate or should I have a look for vegan chocolate as well?
reosarevok
(sure, I know that's *barely* chocolate, but)
mayhem
only Green & Blacks white with vanilla bits.
reosarevok
I don't think I've tried that, but sounds promising
mayhem
we should see if atj can find it and bring some. its just about the only white chocolate worth eating.
kellnerd
Ok, so I can bring a 23 kg case and an 8 kg hand luggage. Not sure how many kg of chocolate I will fit in there, it's been a while since my last flight and I don't remember the weight of my case for a week.
I wish they did more of these for students as well... But nah, they're in talks of ending GSoC 💀
Seeing all the hype for GSoC in India lately, I really wish they just end up expanding it instead, or at least hold local meetups 🫠
mayhem
its not the GSoC team's choice. they would. but the beancounters at google dont understand GSoC because it has no measurable ROI, so it must go away.
Pratha-Fish
mayhem: When would you be visiting India again? :D
kellnerd
Ok, so the lower chocolate bound is 15 * 0.1 kg + 1 kg = 2.5 kg and the upper bound is 30 * 0.1 kg + 2 kg = 5.0 kg 🧑🔬
Pratha-Fish
I'd argue GSoC has negative ROI for google at this point 💀
mayhem
Pratha-Fish: there will be a MeB mini-summit in Delhi Nov 17/18.
Pratha-Fish
mayhem: noiceeee. I'll try my very best to attend :))
mayhem
kellnerd: make the upper bound 6Kg. :)
I need to start planning the Delhi summit too. too many summits, really. :)
Pratha-Fish
mayhem: what do you like to eat? I'll see if I can bring something from Pune. It's a different state all together so it has some variety in food as well hehe
mayhem
Pratha-Fish: yes. the more off the beaten path and spicy! the better. :)
Pratha-Fish
I can already visualize misal, but I doubt it would last 🫠