For the past few months I have been working on upgrading Listenbrainz Cli Tools to use a SQLite database, and now that the (release) dust as settled, time for a community annoucement!
The caching system as been completely redone using SQLite, which brings better performance, less used disk space, and most importantly, easier development.
A new command has been introduced, which allows you to see what your favorite Monstercat releases of this year are, and have an easier time voting for this yearās Best of 2024!
Note: Plenty of releases are missing from Musicbrainz still. So go add them/map your listens. The command checks for the labels āMonstercatā, āMonstercat Instinctā, āMonstercat Uncagedā and " Monstercat Silk".
This command is under the āunstableā command. Unstable commands are commands which are either buggy, unpolished, need user testing, or too hardcoded. The plan being to finish them later, while still letting users experiment with them.
Some things were changed during the migration:
- The work stats do not recurse anymore. Next update I plan to add a new command to restore the previous behaviour.
- The underrated radio has been temporary disabled, as listenbrainz is still broken, and updating it without testing will lead to bugs (and Iām lazy. This update is long overdue already)
- The listen mapper wasnāt upgraded, as Iām not still certain if Iām keeping it or it will be replaced by another way to map listens.
- Importing data dumps is disabled due to the new importer not providing MBID mapping anymore.
Lastly, new name! Itās more like it has a proper name nowā¦
Iāll try to only use one topic (this one) to bring updates on the app. So subscribe if you are interested
I havenāt touched a mac in my life so not that sure if itās possible but:
Open a terminal in the folder where you put the file, so if you put it into your download folder, right click, open in terminal(?).
Or you can open your terminal, put cd <full path of your folder>, aka something like cd /home/rustynova/download
Then you can simply use alitral by typing alistral-macos <command>
The list of commands are here, or you can add --help to get a list of commands
And tip: rename alistral-macos to alistral to make it shorter to type~
Oh and good luck. 140000 listens must be a lot of recording to fetchā¦ And the app is limited to one recording/second. My puny 25k already takes an hour to get only the recordings
Radios: You can now restrict the seed listens to a shorter period with the --seed-listen-range argument. This will only use seed listens from this period. There is also --min-seed-listens to keep a minimum of listens (default 3) from before the period if there is not enough
Changes:
lookup now refresh the recording. So if you know thereās changes on the data, you can use lookup to refresh it
lookup: Overdue score is now from 0. 1 now means that the recording is overdue. Fixes weird behaviours with bumps
Added listens wrong-mapping. This tool finds listens that have been submitted with a different name than the mapped recording. This can be useful to find listens mapped to te wrong recording.
Radios
Added --at-listening-time argument to the overdue radio. This force the score of the recording to be calculated after the previous track have been listened to, instead of the current time. This makes the radio a lot more accurate for long playlist, but may slow down the generation a lot
Future
As Brainzplayer might loose the ability to play youtube videos, the next major release will focus on youtube export of playlists / radio generation. It will probably come in january.
hi, thanks for your work, Itās really useful. Is there a way to achieve something like a playlist that mixes songs listened to by 2 users? Something similar to Spotify Blend
i think you can already do something like that with LB Radio.
if you use stats:
āstats:volpit::weekā ā will create a playlist with songs from your statistics from last week.
and you can add multiple things together:
āstats:volpit::week stats:sanojjonas::weekā will create a playlist with songs that you played and songs that i played.
but i guess what you want is the songs that we both played and not a playlist with some songs from me and some songs from you
It would also need to care about the amount of common listens. You donāt really want to listens to songs that you listened once and disliked it. It would be more of something youāll both jam to.
I did work on it, and will be there next version, but I got severely distracted and tried remaking the whole radio generator and failed. Oops.
You can now see the fresh releases of the last 7 days.
Stats:
Reimplemented the recursive work stats. This add the parent works of the listened works, so you can see the stats number of listens of the original works and/or the number of listens of a soundtrack.
Use alistral stats work-recursive
If the counts are way off, please run alistral cache refresh-data.
Radios
Radios now have a number after their name to disambiguate them.
Added shared radio. This generate a playlist of recordings both listened by the two provided users.
Config
Added the option to set the listenbrainz server url. Set listenbrainz_url in the config file to point to the new api (Should look like this: āListenBrainzā)
Cache
Added top level argument --no-cleanup to prevent cleanup between runs
Added cache refresh-data. This reload all the listened recordings
Musicbrainz:
No changes.
Builds are baking, and should be there soon. A new release should come in the next few days when the listenbrainz update and fixes the listen exports
This update is probably the second biggest refactor since 0.3.0, so watch out for regressions
Listenbrainz
Radios
You can now export your radios to Youtube. This require bringing your own api key. Set up can be found here
Config
You can now set a custom URL for the musicbrainz server.
Meta
Added a wiki at stats - Alistral Manual. This is pretty barebone right now, but will get improved as I go
Improved the logging, Progress bars should be a bit more stable now. A new verbosity flag as been added for debuging release builds
The app has been split into more library crates again, making parts more reusable.
alistral_core: The common utilities of the crates, and everything that doesnāt revolve around config and interface.
interzic: Translation layer for music apps and services. It query multiple providers to link tracks together, and if able to export playlists
tuillez: Utilities for the interface / logging
Others
Plans
Next update should bring a total refactor of the radios (again), to allow creating radios out of a few lines of json. This will allow sharing presets / creating your dream radio without learning to learn rust (Youāll still need to learn the syntax, but itās easier).
Iāll continue to split the app more, notably to move musicbrainz clippy to its own library crate (symphonize), and radios as well (no name for now. I accept suggestions)
I wonāt add any new music services to interzic for now, as Listenbrainz already have Spotify and Apple Music covered. However I donāt mind helping if someone wants to help expand it.
Added manual listen submission. Useful when you listened to a track yet you canāt add it in LB due to it not being in MB. (This was not made out of spite. No sir). Usage: Help - Alistral Manual
Interzic
Added commands to interact with the interzic database
get-mapping: get the mapping of a track
reload: reload the mappings from Musicbrainz
reverse-mapping: find what MB recording is associated to this track
