I believe the dots are the relevance score as calculated by Fresh Releases > For You. So you can still see that info when viewing sorted by release date
2023-09-20 26349, 2023
wisecrackingwaln joined the channel
2023-09-20 26308, 2023
aerozol
Date range I think the same as we have in stats seems like a good move
2023-09-20 26303, 2023
ansh
So should I keep it 1 week / 1 month / 3 months?
2023-09-20 26311, 2023
wisecrackingwaln has quit
2023-09-20 26301, 2023
SothoTalKer has quit
2023-09-20 26355, 2023
SothoTalKer joined the channel
2023-09-20 26301, 2023
aerozol
I think it's week, month, year. But three months could be a good addition for fresh releases
2023-09-20 26302, 2023
texke joined the channel
2023-09-20 26341, 2023
wisecrackingwaln joined the channel
2023-09-20 26341, 2023
wisecrackingwaln has quit
2023-09-20 26301, 2023
lucifer
mayhem: i'll look into automating the server setup thing. makes sense.
(I released beta yesterday and I did do the pot stuff from master, then merged master into beta, since I guess otherwise the new strings wouldn't be added?)
2023-09-20 26329, 2023
reosarevok
(but maybe it was the wrong thing :) )
2023-09-20 26351, 2023
fletchto99 has quit
2023-09-20 26308, 2023
fletchto99 joined the channel
2023-09-20 26306, 2023
q3lont joined the channel
2023-09-20 26355, 2023
q3lont has quit
2023-09-20 26314, 2023
monkey
aerozol: about the regression: yes it reverted the add listens dropdown as well as the follow button…
2023-09-20 26314, 2023
monkey
Time to add some tests !
2023-09-20 26305, 2023
reosarevok
yvanzo, bitmap: see MBS-13279 - do you think it'd make sense to change the editors with email '' to NULL?
hi everyone! I'm pleased to announce that Spotify is once again (for the 3rd time) a supporter of MetaBrainz! https://metabrainz.org/supporters 🦄
2023-09-20 26329, 2023
monkey
Weee !
2023-09-20 26343, 2023
yvanzo
reosarevok: When we have patches in beta that are not in master, it makes sense to create the pot from beta, otherwise it doesn’t make any difference indeed.
2023-09-20 26311, 2023
reosarevok
But we should still merge master into beta first, and then create them, no?
2023-09-20 26320, 2023
reosarevok
(if we want to put out a beta from master)
2023-09-20 26328, 2023
reosarevok
Right now it says to first create the pot, then merge master
2023-09-20 26326, 2023
yvanzo
reosarevok, bitmap: for email address, theoretically it would make more sense to have NULL rather than sometimes '' and sometimes NULL, but I can’t say from memory how it is related to missing/verified/changed email address for existing accounts, it requires digging a bit. But most of display issues reported in this ticket are unrelated.
2023-09-20 26355, 2023
yvanzo
reosarevok: “is it really intended to create pot from beta *before* merging master”: it’s not what is written.
2023-09-20 26336, 2023
yvanzo
it's not creating any pot
2023-09-20 26319, 2023
yvanzo
it updates existing *.pot files, and updates/adds *.po files
2023-09-20 26349, 2023
yvanzo
but it doesn't updates from the code
2023-09-20 26330, 2023
reosarevok
"Run ./po/update_pot.sh to generate new .pot files from the database and templates"
2023-09-20 26355, 2023
yvanzo
Ok, now you are on point 2. which doesn't create them either, it updates from the code/database.
2023-09-20 26325, 2023
reosarevok
1. ii says "update from the code" but this is talking about the beta branch, no?
2023-09-20 26345, 2023
reosarevok
But until 2. you don't update the beta branch if you follow these instructions explicitly :)
2023-09-20 26300, 2023
reosarevok
So the master code gets merged in 2. so the new strings are not available in 1. ii
2023-09-20 26306, 2023
yvanzo
1. is about the beta branch, 2. is not specific about it.
2023-09-20 26317, 2023
reosarevok
Oh, ok, so we should be specific :)
2023-09-20 26347, 2023
yvanzo
Yes
2023-09-20 26347, 2023
reosarevok
Is there a benefit of merging translations into beta specifically, and of doing translation stuff at all before merging master?
2023-09-20 26316, 2023
reosarevok
Or could we merge master always first (if we are going to) then do all translation stuff in beta?
2023-09-20 26322, 2023
yvanzo
what is "doing translation stuff"?
2023-09-20 26340, 2023
reosarevok
Basically: can we switch the order of the points for beta and always do 2. before 1.
2023-09-20 26341, 2023
yvanzo
Yes, you can get translations that are contributed but discarded due to the update.
2023-09-20 26309, 2023
reosarevok
(and then we can always do the entirety of current 1. from beta, since either it's beta-only strings. or master strings already in beta)
2023-09-20 26310, 2023
yvanzo
Sorry, "yes" there is this benefit
2023-09-20 26327, 2023
yvanzo
(so no we should probably not switch the order)
2023-09-20 26343, 2023
reosarevok
Ok, so options are: merge translations into beta, then switch to master, generate pots from there, push, merge master into beta
2023-09-20 26359, 2023
reosarevok
Or: merge translations into beta, then merge master into beta, generate pots from beta, push beta
2023-09-20 26314, 2023
reosarevok
2 seems possibly better?
2023-09-20 26304, 2023
reosarevok
I did not consider the chance of losing some translations in the meantime :)
2023-09-20 26358, 2023
yvanzo
Is your only issue having to switch back and forth to master/beta?
2023-09-20 26308, 2023
reosarevok
No, I just thought that we might want to not have updated pots in master but updated translations only in beta
2023-09-20 26322, 2023
reosarevok
But it might not matter in which case there's no difference between 1 and 2 :)
2023-09-20 26351, 2023
yvanzo
pot files are not actually translations, and translations are already updated only in beta.
2023-09-20 26306, 2023
reosarevok
So just update the current docs to say to change to master if the updates are in master in 1.ii, and push the pots to master? Works for me
2023-09-20 26323, 2023
yvanzo
Thank you.
2023-09-20 26308, 2023
yvanzo
Hopefully this is a temporary setup until there aree sufficient quality checks in Weblate to make it update beta directly.
2023-09-20 26313, 2023
reosarevok
yvanzo: re: the email thing, "But most of display issues reported in this ticket are unrelated"
2023-09-20 26328, 2023
reosarevok
It's not, the "test" account mentioned in the email has a '' email address :)
2023-09-20 26300, 2023
yvanzo
Well it seems to be, I didn’t look into the code.
2023-09-20 26318, 2023
reosarevok
(which is what causes it to display differently, since it passes a "has_email_address" test but not a "email_confirmation_date" test)
2023-09-20 26317, 2023
yvanzo
It might be an issue with the display code making too much assumptions too.
2023-09-20 26338, 2023
reosarevok
Yes, the other possible fix is to just make it so that empty string does not satisfy "has_email_address"
2023-09-20 26356, 2023
reosarevok
It just feels like having an empty string is probably not intentional, that's why I thought making them NULL might be better
