Can someone using Safari test the authentication from Picard? We had a report yesterday, and it seems to me user's Safari refuses to use http://localhost:8000 because that's HTTP (perhaps HSTS set for localhost, dunno).
2025-02-28 05901, 2025
julian45[m]
sure, let me give it a quick spin
2025-02-28 05903, 2025
julian45[m]
sign-in to MB account worked fine, i presume since it uses auth code flow rather than a native callback
2025-02-28 05934, 2025
julian45[m]
(though i'm on 2.12.3 right now and maybe it's native in dev builds -- should i grab one and try it?)
2025-02-28 05949, 2025
julian45[m]
ope, there is definitely something broken there -- if I slap ?tport=8000 onto a release URL and try to use that to get the release into the right pane of picard, it works on firefox but definitely not in safari
2025-02-28 05925, 2025
julian45[m]
* ope, there is definitely something broken there -- if I slap ?tport=8000 onto a release URL and try to use the tagger button to get the release into the right pane of picard, it works on firefox but definitely not in safari
2025-02-28 05956, 2025
julian45[m]
looked around and there doesn't seem to be a way around this :/ browserstack has it as a known issue for them, but their workarounds aren't applicable to us: https://www.browserstack.com/support/faq/local-te…
2025-02-28 05924, 2025
julian45[m]
on a different note, it looks like the jira server is not managed at all by ansible; it's not in the hosts list in metabrainz-ansible, and when i tried SSHing in (like i would with another server subject to ansible) to look at the environment, the errors i got suggested that an account hasn't been provisioned for me (or anyone else) there
2025-02-28 05939, 2025
julian45[m]
zas: would you be able to do at least a little integration work on the jira server (e.g., run the users role so that i have an account and key there), or, if accounts need to remain separately managed, could you please create an account for me there?
2025-02-28 05944, 2025
zas[m]
julian45: can you open a ticket for Picard explaining the issue with Safari? We need to look into that, that's rather annoying. Not sure what the solution could be atm, but a description of the exact issue is a good starting point
Hi monkey... (full message at <https://matrix.chatbrainz.org/_matrix/media/v3/download/chatbrainz.org/RKGtkwRSpuwXVNhdwFLzLclx>)
2025-02-28 05931, 2025
bitmap[m]
<reosarevok[m]> "bitmap: can you see if MBS-13954..." <- it's a one-line change, but I'm not sure which way was intentional (I could imagine the comments make the previews harder to check if the join phrases are meant to have very particular spacing, or too unwieldy if there are many artists with comments?)
