#metabrainz

/

      • alastairp
        interesting comment about sorting input in the tar file before compressing in order to increase local-compressability
      • 2021-03-16 07513, 2021

      • ruaok
        woah. 🤯
      • 2021-03-16 07558, 2021

      • ruaok
        actually, it explains it.
      • 2021-03-16 07559, 2021

      • alastairp
        I'm thinking about listens, json keys are the same no matter what order we sort the input in
      • 2021-03-16 07505, 2021

      • chaban
        smart 7z does something similar 7zip: https://www.tc4shell.com/en/7zip/smart7z/
      • 2021-03-16 07506, 2021

      • ruaok
        it really relies on the underlying data.
      • 2021-03-16 07522, 2021

      • ruaok
        "By sorting the files by name, two subsequent files have little different characters between them. Apparently this is very favourable for the compression efficiency."
      • 2021-03-16 07533, 2021

      • alastairp
        listen data might be more locally compressable if you put all stuff from the same artist/album next to each ohter (for example)
      • 2021-03-16 07554, 2021

      • ruaok
        which makes sense since a lot of compression algs use sliding windows. so if your data in your window is more compressable...
      • 2021-03-16 07558, 2021

      • alastairp
        yep
      • 2021-03-16 07505, 2021

      • ruaok
        but that wont work for listens.
      • 2021-03-16 07531, 2021

      • ruaok
        unless you sort the dump by MBID, then maaaaaybe
      • 2021-03-16 07558, 2021

      • alastairp
        right. I don't think the changing parts of listens are stable enough that it'll give us a significant decrease in size
      • 2021-03-16 07520, 2021

      • alastairp
        but anyway, could be interesting to try a few experiments - randomly add files, or sort by a few different fields
      • 2021-03-16 07527, 2021

      • _lucifer
        Mr_Monkey, updated the PR.
      • 2021-03-16 07515, 2021

      • Mr_Monkey
        Yep, saw that, thanks :)
      • 2021-03-16 07535, 2021

      • Mr_Monkey
        _lucifer: I think the endpoints for /follow and /unfollow should follow the same treatment, no?
      • 2021-03-16 07554, 2021

      • _lucifer
        yeah, makes sense to move those to the api as well.
      • 2021-03-16 07507, 2021

      • _lucifer
        will do.
      • 2021-03-16 07535, 2021

      • Mr_Monkey
        Fanks
      • 2021-03-16 07522, 2021

      • reosarevok
        ruaok, yvanzo, bitmap: started with the post, https://musicbrainz.wordpress.com/?p=8722&pre…
      • 2021-03-16 07504, 2021

      • reosarevok
        Comments / description writing appreciated :D
      • 2021-03-16 07542, 2021

      • reosarevok goes have lunch
      • 2021-03-16 07512, 2021

      • alastairp
        > MBS-11451: Support ratings for places. Places are the only entity that can be reviewed in CritiqueBrainz, but do not support ratings in MusicBrainz.
      • 2021-03-16 07513, 2021

      • BrainzBot
        MBS-11451: Support ratings for places https://tickets.metabrainz.org/browse/MBS-11451
      • 2021-03-16 07544, 2021

      • alastairp
        with the position of the comma it kind of implies that CB only exists to review places
      • 2021-03-16 07557, 2021

      • reosarevok
        I guess it does :D
      • 2021-03-16 07502, 2021

      • reosarevok
        Will amend
      • 2021-03-16 07536, 2021

      • reosarevok
        "Places are the only entity that can be reviewed in CritiqueBrainz yet cannot be rated in MusicBrainz" ?
      • 2021-03-16 07538, 2021

      • alastairp
        maybe even remove "are the only entity". "Places can be reviewed in CritiqueBrainz but cannot be rated in MusicBrainz. All other entities that can be reviewd in CB can be rated in MB"
      • 2021-03-16 07549, 2021

      • Mr_Monkey
        ruaok: similarity score currently looks to be on a scale from 0->0.1 rather than 0->1 that I was (perhaps erroneously) expecting. Is that expected? https://api.listenbrainz.org/1/user/mr_monkey/sim…
      • 2021-03-16 07522, 2021

      • alastairp
        but it's such a small sentence that I don't think it matters either way
      • 2021-03-16 07530, 2021

      • ruaok
      • 2021-03-16 07510, 2021

      • reosarevok
        alastairp: ok :)
      • 2021-03-16 07511, 2021

      • ruaok
        its that so far most of our users are not super compatible.
      • 2021-03-16 07512, 2021

      • alastairp
        poor Mr_Monkey, no one is similar to him
      • 2021-03-16 07514, 2021

      • Mr_Monkey
        Hm. So it is 0->1 and I'm just isolated?
      • 2021-03-16 07524, 2021

      • ruaok
        Mr_Monkey: most of us appear to be isolated.
      • 2021-03-16 07526, 2021

      • Mr_Monkey
        Ok, that's fine then, thanks :)
      • 2021-03-16 07500, 2021

      • ruaok
        but there should be at least one 1.0 score in the DB. in theory.
      • 2021-03-16 07537, 2021

      • alastairp
        ruaok: there should be a 1.0 score because you normalise all similarities?
      • 2021-03-16 07545, 2021

      • ruaok
        yes
      • 2021-03-16 07549, 2021

      • ruaok
        I should write a query to find those top users.
      • 2021-03-16 07501, 2021

      • RikkoM joined the channel
      • 2021-03-16 07533, 2021

      • ruaok
        :wq
      • 2021-03-16 07535, 2021

      • ruaok
        gah.
      • 2021-03-16 07536, 2021

      • CatQuest
        huh. [12:39] <reosarevok> May 17? is norways national day 🤔
      • 2021-03-16 07501, 2021

      • ruaok
        zas: do you know how to generate the keys in docker-server-configs/ssh ?
      • 2021-03-16 07520, 2021

      • ruaok
        I need to add another one, but I have no idea what the magic incantation is for creating more keys.
      • 2021-03-16 07516, 2021

      • _lucifer
        you can do that using ssh-keygen from terminal
      • 2021-03-16 07543, 2021

      • ruaok
        yes, but these keys need special features, like setting a home dir and things like that.
      • 2021-03-16 07526, 2021

      • _lucifer
        oh ok, ignore me then :)
      • 2021-03-16 07501, 2021

      • _lucifer
        alastairp, here https://github.com/metabrainz/listenbrainz-server…. shouldn't it be ${COMPOSE_PROJECT_NAME} ?
      • 2021-03-16 07533, 2021

      • _lucifer
        i mean ${COMPOSE_PROJECT_NAME}_listenbrainz instead of $COMPOSE_PROJECT_NAME_listenbrainz
      • 2021-03-16 07543, 2021

      • nifemi joined the channel
      • 2021-03-16 07548, 2021

      • sumedh has quit
      • 2021-03-16 07526, 2021

      • RikkoM has quit
      • 2021-03-16 07532, 2021

      • Nizar joined the channel
      • 2021-03-16 07542, 2021

      • Nizar has quit
      • 2021-03-16 07502, 2021

      • Nizar joined the channel
      • 2021-03-16 07521, 2021

      • RikkoM joined the channel
      • 2021-03-16 07521, 2021

      • nifemi
        Hello everyone, my name is Precious Oluwanifemi a third year undergraduate of computer engineering in Obafemi Awolowo University. I have a good experience working with ReactJs,NextJs and Typescript. I am interested on working on the Music-Reviews and Pin-My Jam ideas for ListenBrainz during the period of gsoc
      • 2021-03-16 07507, 2021

      • _lucifer
        hi nifemi, welcome to the community!
      • 2021-03-16 07549, 2021

      • nifemi
        Thank you so much @_lucifer, looking forward to it
      • 2021-03-16 07515, 2021

      • yvanzo
        reosarevok: Maybe some other cols could be moved to SMALLINT but it's a bit short in time to check potential implications and to decide now. There will be other schema changes.
      • 2021-03-16 07520, 2021

      • reosarevok shrugs
      • 2021-03-16 07529, 2021

      • reosarevok
        Cannot see any implications, but maybe I'm missing something :)
      • 2021-03-16 07547, 2021

      • alastairp
        _lucifer: good catch. perhaps we should also run shellcheck over this script
      • 2021-03-16 07551, 2021

      • reosarevok
        Anyway, I'll be back in a while and we can maybe finish the blog once bitmap is around :)
      • 2021-03-16 07521, 2021

      • _lucifer
        alastairp: yup, will do that.
      • 2021-03-16 07504, 2021

      • _lucifer
        also, we have two docker-compose.integration.yml files one in docker and second in docker/jenkins.
      • 2021-03-16 07517, 2021

      • _lucifer
        do we need both or can they be merged?
      • 2021-03-16 07552, 2021

      • ruaok
        zas: you about today?
      • 2021-03-16 07551, 2021

      • alastairp
        _lucifer: yeah, there's a trick about that. in local development we mount . into /code so that we can run tests without rebuilding, but this doesn't work in jenkins due to our docker-in-docker setup. so we need a separate file with no source volume configured
      • 2021-03-16 07526, 2021

      • alastairp
        however, we could try and improve this by using docker-compose overlay files instead, so we have just 1 file and add another config file in one case (either jenkins or locally, I'm not sure which one is better)
      • 2021-03-16 07512, 2021

      • _lucifer
        oh ok, i'll open a ticket for now and look into again later
      • 2021-03-16 07522, 2021

      • BrainzGit
        [listenbrainz-server] mayhem merged pull request #1344 (master…notify-similar-user-update): Add a notification that similar users imported or failed https://github.com/metabrainz/listenbrainz-server…
      • 2021-03-16 07539, 2021

      • bharat1999 joined the channel
      • 2021-03-16 07548, 2021

      • nifemi has quit
      • 2021-03-16 07528, 2021

      • c1e0 joined the channel
      • 2021-03-16 07510, 2021

      • ruaok
        iliekcomputers: do you remember how to add ssh keys for the FTP dumps to docker-server-configs ?
      • 2021-03-16 07555, 2021

      • BrainzGit
        [bookbrainz-site] akashgp09 opened pull request #574 (master…image-path): FIX(BB-590): Display Publisher Achievement Badges https://github.com/bookbrainz/bookbrainz-site/pul…
      • 2021-03-16 07535, 2021

      • bharat1999 has quit
      • 2021-03-16 07550, 2021

      • Nizar has quit
      • 2021-03-16 07544, 2021

      • iliekcomputers
      • 2021-03-16 07545, 2021

      • MRiddickW joined the channel
      • 2021-03-16 07508, 2021

      • _lucifer
        alastairp: https://github.com/metabrainz/listenbrainz-server… does seem to fail on not finding images
      • 2021-03-16 07532, 2021

      • ruaok
        iliekcomputers: ahh, I see. so the SSH keys in the server configs are just bog standard and the installation gives its the restrictions for what it can access and all that. good to know.
      • 2021-03-16 07541, 2021

      • _lucifer
        alastairp, adding `|| true` at the end of that statement seems to work
      • 2021-03-16 07522, 2021

      • alastairp
        _lucifer: ah, right. I think that's fine
      • 2021-03-16 07531, 2021

      • alastairp
        because `docker image rm` requires a non-empty input, right?
      • 2021-03-16 07547, 2021

      • alastairp
        and if the filter doesn't return anything then it fails to run?
      • 2021-03-16 07554, 2021

      • _lucifer
      • 2021-03-16 07508, 2021

      • alastairp
        mmm, but that's weird
      • 2021-03-16 07521, 2021

      • alastairp
        because the `docker images` should return actual images that exist
      • 2021-03-16 07528, 2021

      • _lucifer
        i am not sure what's happening
      • 2021-03-16 07531, 2021

      • alastairp
        why does it return an image that docker rm can't remove?
      • 2021-03-16 07558, 2021

      • alastairp
        there are some cases where rm doesn't work (for example, if the same image hash is used by two different labels), but that seems different
      • 2021-03-16 07502, 2021

      • bitmap
        reosarevok: I'll see if I have time to do MBS-10962 tooo. and no, I'll check the edit status one
      • 2021-03-16 07503, 2021

      • BrainzBot
        MBS-10962: Speed up listing artist's releases https://tickets.metabrainz.org/browse/MBS-10962
      • 2021-03-16 07538, 2021

      • _lucifer
        alastairp: the first line of the output comes from the command inside $()
      • 2021-03-16 07506, 2021

      • alastairp
        _lucifer: ah, mmm. so that's theoretically the image that we just built
      • 2021-03-16 07519, 2021

      • alastairp
        in which case might that image not be there?
      • 2021-03-16 07546, 2021

      • _lucifer
        yes, i think that's the case
      • 2021-03-16 07500, 2021

      • _lucifer
        how about filtering based on time?
      • 2021-03-16 07545, 2021

      • _lucifer
        ah nvm that's what we are currently doing
      • 2021-03-16 07537, 2021

      • _lucifer
        alastairp, i opened LB-839 recently, thoughts?
      • 2021-03-16 07538, 2021

      • BrainzBot
        LB-839: Categorize tests and optionally add common cases to test.sh https://tickets.metabrainz.org/browse/LB-839
      • 2021-03-16 07533, 2021

      • _lucifer
        the rationale is that say i only touched the dump code, i don't want to wait for other tests to run. it would be nice to have something like `./test.sh dumps`
      • 2021-03-16 07530, 2021

      • c1e0 has quit
      • 2021-03-16 07502, 2021

      • c1e0 joined the channel
      • 2021-03-16 07559, 2021

      • c1e0_ joined the channel
      • 2021-03-16 07528, 2021

      • c1e0 has quit
      • 2021-03-16 07528, 2021

      • alastairp
        _lucifer: interesting idea. Do you know that you can already specify a file/test class/pytest argument to the end of test.sh?
      • 2021-03-16 07553, 2021

      • alastairp
        so this would be a way of further grouping multiple files together into logical categories?
      • 2021-03-16 07515, 2021

      • _lucifer
        yes, i specify the file name usually but specifying the entire file path and multiple files is a bit cumbersome
      • 2021-03-16 07531, 2021

      • _lucifer
        i was thing using pytest marks
      • 2021-03-16 07535, 2021

      • _lucifer
        *thinking
      • 2021-03-16 07549, 2021

      • c1e0_ has quit
      • 2021-03-16 07506, 2021

      • _lucifer
        like we specify the database tests in BU so that those are not run on python 3
      • 2021-03-16 07508, 2021

      • alastairp
        one consideration here is that it requires someone to keep the marks up to date
      • 2021-03-16 07510, 2021

      • c1e0_ joined the channel
      • 2021-03-16 07513, 2021

      • _lucifer
        yeah
      • 2021-03-16 07511, 2021

      • c1e0 joined the channel
      • 2021-03-16 07534, 2021

      • alastairp
        could we improve this by better folder organisation? (just thinking out loud)
      • 2021-03-16 07558, 2021

      • alastairp
        that is to say, I'm not opposed to the ideas of marks, and we could try and add a few
      • 2021-03-16 07515, 2021

      • _lucifer
        better organisation would be a step better than what we have now.
      • 2021-03-16 07516, 2021

      • alastairp
        perhaps just use the existing pytest -m xx functionality for now
      • 2021-03-16 07558, 2021

      • alastairp
        related to this, I also want to merge integration tests and unit tests into a single docker-compose file, and maybe the same pytest run
      • 2021-03-16 07510, 2021

      • alastairp
        in which case, marks might be a good idea to separate these anyway
      • 2021-03-16 07509, 2021

      • c1e0_ has quit
      • 2021-03-16 07550, 2021

      • RikkoM has quit
      • 2021-03-16 07512, 2021

      • c1e0_ joined the channel
      • 2021-03-16 07553, 2021

      • c1e0 has quit
      • 2021-03-16 07527, 2021

      • sumedh joined the channel
      • 2021-03-16 07543, 2021

      • c1e0 joined the channel
      • 2021-03-16 07504, 2021

      • c1e0_ has quit
      • 2021-03-16 07509, 2021

      • c1e0_ joined the channel
      • 2021-03-16 07533, 2021

      • c1e0 has quit
      • 2021-03-16 07546, 2021

      • dpmittal joined the channel
      • 2021-03-16 07535, 2021

      • c1e0_ is now known as c1e0
      • 2021-03-16 07546, 2021

      • Rohan_Pillai joined the channel
      • 2021-03-16 07527, 2021

      • sumedh has quit
      • 2021-03-16 07559, 2021

      • reosarevok
      • 2021-03-16 07500, 2021

      • BrainzBot
        MBS-10647: Add [no label] to b_del_label_special trigger for labels
      • 2021-03-16 07513, 2021

      • bitmap
        makes sense to me
      • 2021-03-16 07537, 2021

      • reosarevok
      • 2021-03-16 07538, 2021

      • BrainzBot
        MBS-10208: Allow merging collections
      • 2021-03-16 07550, 2021

      • reosarevok
        I guess it makes sense to add a redirect table or?
      • 2021-03-16 07555, 2021

      • reosarevok
        The code is otherwise written, but