16:06 PM
ishaanshah
I like the design
2020-06-08 16046, 2020
16:06 PM
ishaanshah
But I agree with ruaok's point
2020-06-08 16002, 2020
16:07 PM
ishaanshah
If I want to look at my release stats
2020-06-08 16010, 2020
16:07 PM
ruaok
don't get me wrong, I do like the design too. :)
2020-06-08 16024, 2020
16:07 PM
ishaanshah
It is not obvious that I click the dropdown
2020-06-08 16041, 2020
16:07 PM
ishaanshah
and then select releases
2020-06-08 16042, 2020
16:07 PM
rdswift has quit
2020-06-08 16000, 2020
16:08 PM
ishaanshah
2020-06-08 16010, 2020
16:08 PM
iliekcomputers
Could we use a gray pill over the text
2020-06-08 16030, 2020
16:08 PM
ishaanshah
So for know I switched back to original tablist
2020-06-08 16047, 2020
16:08 PM
iliekcomputers
2020-06-08 16000, 2020
16:09 PM
iliekcomputers
I think it's the orange that's the problem in this specific case
2020-06-08 16013, 2020
16:09 PM
ishaanshah
But when we switch to shivam-kapila's dashboard design, we can solve this issue with the subtab concept
2020-06-08 16017, 2020
16:09 PM
iliekcomputers
Oh, OK that sounds good to me
2020-06-08 16021, 2020
16:09 PM
iliekcomputers
👍🏽
2020-06-08 16026, 2020
16:09 PM
ishaanshah
2020-06-08 16002, 2020
16:10 PM
ishaanshah
similar to the playlist subtabs
2020-06-08 16029, 2020
16:10 PM
iliekcomputers
Yeah, makes sense.
2020-06-08 16000, 2020
16:11 PM
ishaanshah
Cool, wanted to ask for your opinion on this
2020-06-08 16021, 2020
16:11 PM
ishaanshah
so ig we should stick with the nav pills for now?
2020-06-08 16033, 2020
16:11 PM
iliekcomputers
Yeah, let's do that.
2020-06-08 16044, 2020
16:11 PM
ishaanshah
Ok
2020-06-08 16015, 2020
16:12 PM
ishaanshah
Other than that I started on recording calcs
2020-06-08 16032, 2020
16:12 PM
ishaanshah
two minor frontend issues are remaining
2020-06-08 16041, 2020
16:12 PM
iliekcomputers
Yeah, I skimmed through the PR. It's pretty cool how little we have to change now :)
2020-06-08 16052, 2020
16:12 PM
ishaanshah
I can do those tomorrow
2020-06-08 16032, 2020
16:13 PM
ishaanshah
I wanted to ask one thing in the stat calculatio
2020-06-08 16003, 2020
16:14 PM
iliekcomputers
Mhmm
2020-06-08 16011, 2020
16:14 PM
ishaanshah
Why do we select unique msids too
2020-06-08 16019, 2020
16:14 PM
ishaanshah
I mean unique mbid makes sense
2020-06-08 16032, 2020
16:14 PM
iliekcomputers
What do you mean?
2020-06-08 16043, 2020
16:14 PM
iliekcomputers
Link to the code that we're talking about
2020-06-08 16044, 2020
16:14 PM
iliekcomputers
?
2020-06-08 16011, 2020
16:15 PM
ishaanshah
2020-06-08 16027, 2020
16:15 PM
ishaanshah
Over here why do we have msid columns?
2020-06-08 16049, 2020
16:15 PM
ishaanshah
msid depends on other factors than name of artist and recording right?
2020-06-08 16007, 2020
16:16 PM
ishaanshah
so if the metadata of two listens from same artist is different
2020-06-08 16013, 2020
16:16 PM
ishaanshah
the msid is different
2020-06-08 16024, 2020
16:16 PM
ishaanshah
leading to multiple entries in the stats
2020-06-08 16006, 2020
16:17 PM
sumedh has quit
2020-06-08 16038, 2020
16:17 PM
iliekcomputers
hmm, the idea was supposed to be that we'll work with msids as index initially and eventually match them to mbids
2020-06-08 16054, 2020
16:17 PM
iliekcomputers
which is why the current code uses msids as source of truth for uniqueness.
2020-06-08 16055, 2020
16:17 PM
sumedh joined the channel
2020-06-08 16032, 2020
16:18 PM
ishaanshah
Hmm so suppose track a and track b are the same
2020-06-08 16051, 2020
16:18 PM
ishaanshah
but their msid is different because of their source
2020-06-08 16004, 2020
16:19 PM
ishaanshah
like lastfm vs spotify or something
2020-06-08 16014, 2020
16:19 PM
ishaanshah
then I will have two entries in my stats
2020-06-08 16024, 2020
16:19 PM
iliekcomputers
right
2020-06-08 16026, 2020
16:19 PM
ishaanshah
How will we combat this issue?
2020-06-08 16041, 2020
16:19 PM
ishaanshah
assuming we have mapping available
2020-06-08 16049, 2020
16:19 PM
iliekcomputers
the matching should eventually point the two msids to the same mbid
2020-06-08 16014, 2020
16:20 PM
iliekcomputers
and once we're comfortable with the matching and it's accuracy, we start using matched mbids as the source of truth for uniqueness instead
2020-06-08 16018, 2020
16:20 PM
iliekcomputers
that is the plan, afaik
2020-06-08 16047, 2020
16:20 PM
ishaanshah
Ohk so the workflow is
2020-06-08 16015, 2020
16:21 PM
ishaanshah
calc stats with msid -> merge if pointing to same mbid -> send to LB?
2020-06-08 16017, 2020
16:21 PM
rdswift joined the channel
2020-06-08 16034, 2020
16:21 PM
iliekcomputers
this is all still up in the air
2020-06-08 16039, 2020
16:21 PM
iliekcomputers
but i envision it as follows
2020-06-08 16001, 2020
16:22 PM
iliekcomputers
we'll eventually start using the matching to fill in mbids for all listens
2020-06-08 16015, 2020
16:22 PM
iliekcomputers
then the spark cluster will have mbids for basically all listens
2020-06-08 16025, 2020
16:22 PM
sumedh has quit
2020-06-08 16026, 2020
16:22 PM
iliekcomputers
then we'll just move the query to use mbids as index
2020-06-08 16054, 2020
16:22 PM
ishaanshah
Ohk, that makes sense
2020-06-08 16022, 2020
16:23 PM
ishaanshah
i was looking at the msb code this weekend thats why I asked
2020-06-08 16022, 2020
16:23 PM
ruaok
iliekcomputers: the plan you just outlined is right.
2020-06-08 16042, 2020
16:23 PM
ruaok
ideally we can put MBIDs directly into the listens in timescale.
2020-06-08 16040, 2020
16:24 PM
ishaanshah
Another thing
2020-06-08 16017, 2020
16:25 PM
ishaanshah
Can I modify the query so as to turn empty strings into null while calculating stats
2020-06-08 16026, 2020
16:25 PM
shivam-kapila
> ideally we can put MBIDs directly into the listens in timescale.
2020-06-08 16026, 2020
16:25 PM
shivam-kapila
This sound better
2020-06-08 16006, 2020
16:26 PM
ishaanshah
2020-06-08 16017, 2020
16:26 PM
ishaanshah
Some of the entries here are null
2020-06-08 16020, 2020
16:27 PM
ishaanshah
If we set empty strings to null, they will automatically get deleted while entering in the db
2020-06-08 16030, 2020
16:27 PM
iliekcomputers
right, that makes sense to me
2020-06-08 16046, 2020
16:27 PM
Lotheric_ has quit
2020-06-08 16018, 2020
16:28 PM
ishaanshah
Cool, I will open a PR for that
2020-06-08 16031, 2020
16:28 PM
iliekcomputers
awesome, thanks
2020-06-08 16042, 2020
16:28 PM
iliekcomputers
i'll try to take a look at PRs today, but i am a little swamped
2020-06-08 16045, 2020
16:28 PM
ishaanshah
Thats all I had for today
2020-06-08 16054, 2020
16:28 PM
ishaanshah
do you have anything else on mind
2020-06-08 16005, 2020
16:29 PM
iliekcomputers
nah, i don't have anything.
2020-06-08 16022, 2020
16:29 PM
ishaanshah
Cool, can you have a look at #900 today if you get time
2020-06-08 16032, 2020
16:29 PM
iliekcomputers
one piece of good news is that ruaok has drastically reduced the dump times so we should soon be able to update stats basically daily if we want to
2020-06-08 16039, 2020
16:29 PM
iliekcomputers
once it gets merged
2020-06-08 16049, 2020
16:29 PM
iliekcomputers
ishaanshah: sure, i'll try to.
2020-06-08 16003, 2020
16:30 PM
iliekcomputers
shivam-kapila: yeah, i
2020-06-08 16029, 2020
16:30 PM
ishaanshah
> one piece of good news is that ruaok has drastically reduced the dump times so we should soon be able to update stats basically daily if we want to
2020-06-08 16029, 2020
16:30 PM
ishaanshah
:tada:
2020-06-08 16052, 2020
16:30 PM
iliekcomputers
shivam-kapila: yeah, we only get the 20-30 min each day to chat, so it's become more formal. not sure if there's a good way around it to be honest.
2020-06-08 16058, 2020
16:30 PM
ruaok shudders
2020-06-08 16004, 2020
16:31 PM
ruaok
we should just implement streaming. :)
2020-06-08 16045, 2020
16:31 PM
iliekcomputers
ishaanshah: if you feel the meetings are a bit too much, feel free to tell me. I personally find them really helpful in keeping track.
2020-06-08 16001, 2020
16:32 PM
iliekcomputers
ruaok: so I don't think that's the spark way to do it to be honest.
2020-06-08 16059, 2020
16:32 PM
ruaok
not even the incremental dumps?
2020-06-08 16055, 2020
16:33 PM
iliekcomputers
the incremental dumps should be fixed and used, i don't think we need the level of detail the streaming thing gives us.
2020-06-08 16005, 2020
16:34 PM
ruaok
agreed.
2020-06-08 16016, 2020
16:34 PM
ruaok
just lets not import 340M listens each day.
2020-06-08 16033, 2020
16:34 PM
iliekcomputers
agree with that.
2020-06-08 16041, 2020
16:34 PM
BrainzGit
2020-06-08 16042, 2020
16:34 PM
BrainzBot
2020-06-08 16044, 2020
16:34 PM
travis-ci joined the channel
2020-06-08 16044, 2020
16:34 PM
travis-ci
2020-06-08 16044, 2020
16:34 PM
travis-ci has left the channel
2020-06-08 16016, 2020
16:35 PM
ishaanshah
*ishaanshah spilled milk on his keyboard 😥
2020-06-08 16002, 2020
16:36 PM
ishaanshah
iliekcomputers: Not at all, I find them extremely useful
2020-06-08 16008, 2020
16:39 PM
yvanzo
bitmap, reosarevok: finally fixed indexed search on test.mb.o (amqp.broker was not correctly set)
2020-06-08 16021, 2020
16:39 PM
reosarevok
Yay
2020-06-08 16006, 2020
16:40 PM
reosarevok
bitmap: around?
2020-06-08 16027, 2020
16:40 PM
BrainzGit
2020-06-08 16027, 2020
16:40 PM
BrainzBot
2020-06-08 16033, 2020
16:42 PM
reosarevok
Left a comment on the PR for now
2020-06-08 16015, 2020
16:44 PM
Mr_Monkey
Cyna[m]: Hi ! I was wondering if you have any code laying around regarding annotations on BookBrainz, since you were looking into it. Wouldn't want your efforts to go to waste :)
2020-06-08 16046, 2020
16:53 PM
sumedh joined the channel
2020-06-08 16028, 2020
16:59 PM
shivam-kapila
ishaanshah: I jroke my laptop charger 😐
2020-06-08 16032, 2020
16:59 PM
shivam-kapila
broke*
2020-06-08 16007, 2020
17:00 PM
Mr_Monkey
2020-06-08 16016, 2020
17:00 PM
Freso
<BANG>
2020-06-08 16016, 2020
17:00 PM
Freso
It’s World Oceans Monday!
2020-06-08 16016, 2020
17:00 PM
Freso
2020-06-08 16016, 2020
17:00 PM
Freso
And with that said… reviews!
2020-06-08 16016, 2020
17:00 PM
Freso
First off we have the mail-ins, which consist of pristine__ this week, so…
2020-06-08 16016, 2020
17:00 PM
Freso
pristine__: Go!
2020-06-08 16016, 2020
17:00 PM
Freso
pristine__ says…
2020-06-08 16017, 2020
17:00 PM
Freso
"""
2020-06-08 16017, 2020
17:00 PM
Freso
I mostly worked on updating FTP mapping downloader and resolved spark error due to temporary files
2020-06-08 16018, 2020
17:00 PM
Freso
"""
2020-06-08 16018, 2020
17:00 PM
Freso
Other people up for reviews: alastairp, bitmap, Mr_Monkey, ruaok, yvanzo, Freso, zas, reosarevok, diru1100, CatQuest, shivam-kapila, Cyna[m], ishaanshah[m], _lucifer – anyone else, please let me know ASAP.
2020-06-08 16019, 2020
17:00 PM
Freso
alastairp: Go!
2020-06-08 16005, 2020
17:01 PM
Freso
OR… bitmap, go!
2020-06-08 16009, 2020
17:01 PM
bitmap
hey
2020-06-08 16029, 2020
17:01 PM
bitmap
2020-06-08 16042, 2020
17:01 PM
bitmap
that should resolve some issues with our release process
2020-06-08 16044, 2020
17:01 PM
shivam-kapila
(Mr_Monkey: Exactly same expressions)
2020-06-08 16014, 2020
17:02 PM
bitmap
also did a little bit of code review and helped reosarevok with making a dump of the puid edits to put on ftp
2020-06-08 16036, 2020
17:02 PM
bitmap
2020-06-08 16054, 2020
17:02 PM
bitmap
fin! Mr_Monkey go!
2020-06-08 16058, 2020
17:02 PM
Mr_Monkey
Hi !
2020-06-08 16003, 2020
17:04 PM
Freso
>_>
2020-06-08 16028, 2020
17:04 PM
Freso
Did we lose Mr_Monkey or is he typing a very long message?
2020-06-08 16036, 2020
17:04 PM
Mr_Monkey
Last week I worked with alastairp to figure out a better way for Flask webservers and APIs to interact with React components, which had quite good results.
2020-06-08 16036, 2020
17:04 PM
Mr_Monkey
We ended up with strong typescript typings automatically exported from type definitions in pythons
2020-06-08 16040, 2020
17:04 PM
Mr_Monkey
The second option :)
2020-06-08 16012, 2020
17:05 PM
Mr_Monkey
I also continued the never ending merging tool, which is now finally ready to be published 🎉
2020-06-08 16034, 2020
17:05 PM
ruaok
woo!
2020-06-08 16048, 2020
17:05 PM
Mr_Monkey
I wanted to fix a couple of other pesky issues (non merging related) before deploying so I was implementing those fixes and continued today.
2020-06-08 16008, 2020
17:06 PM
Mr_Monkey
I also did PR reviews for LB and BB
2020-06-08 16036, 2020
17:06 PM
Mr_Monkey
And finally, on Friday and a tiny bit over the weekend, I implemented another datasource for ListenBrainz: SoundCloud
2020-06-08 16002, 2020
17:07 PM
shivam-kapila
Wow
2020-06-08 16006, 2020
17:07 PM
Mr_Monkey
Their embedded player and javascript library to control it seemed pretty simple, so why not? The PR is open and working, just need to write some tests.
2020-06-08 16007, 2020
17:07 PM
Freso
Does that also look for the `origin_url`?
