aerozol: we still need to work on the profile page for LB.
aerozol
That was my git practice - it doesn't interfere with any design updates
I'm currently catching up on tickets btw! Should be to the MB redesign next. Let me know if you want me to do something else first!
akshaaatt: oh wait, is there going to be a completely new profile page? I just want to get to my page with my listens, it's not obvious that 'home' should point there
PICARD-2446: Create Picard banners for store fronts
skelly37 joined the channel
outsidecontext
aerozol: thanks a lot. I totally love this. I commented on the use, which basically means I think the approach is right for the way these images are used.
Also from my side nothing to say about the design itself. Maybe someone else has something to criticize :D
akshaaatt
Yes aerozol, LB will have a new profile page soon. We should wait to point to the profile page until that’s done.
skelly37 has quit
lucifer
akshaaatt: hi! did you get an appointment for the visa yet?
monkey: sure will do
vardan: youtube api tests sounds good to me.
riksucks: 👍 will do
akshaaatt
Hi lucifer! I have an agent who says we can get the appointment in a day or two. Will schedule for it this Saturday.
lucifer
akshaaatt: ah cool, i was looking at the VAC's website form and there is no slot available till 11th Aug. (slots for later dates haven't opened yet). let me know if you are able to get one.
atj
aerozol: <3
Sophist-UK joined the channel
Sophist-UK has quit
skelly37 joined the channel
skelly37
rdswift: I've added a warning about the current undefined pipe behavior, please take a look in free time if it's formatted correctly and nicely for our docs :)
Also the section about single-instance should be extended when we agree on the urls format
atj
skelly37: can you link to the PR?
skelly37
atj: Docs, current URLs or merged single-instance that causes the behavior?
Pratha-Fish comes back from the internet loadshading
alastairp
lucifer: hi
lucifer
alastairp: hi! while updating docs today, i saw that there were some warnings in the build again (locally) and that the rtd builds are failing for some weeks due to an incompatible jinja2 version. to improve the situation i was thinking to add another workflwo to build docs.
alastairp
that sounds like a great idea
lucifer
however, running it on every commit seems a bit too much. and running it after merging pr is also not great.
alastairp
and we can't have it trigger on docs/ because it's also affected by code + depdenencies
lucifer
so i was thinking of adding a manual workflow with possibility of running it a comment is added to PR.
indeed.
another thing i was thinking of was to reduce test runtime somehow. will probably look into it next week onwards.
alastairp
manual workflows are problematic because you can forget to run them anyway. and even if we used a cron job, they're hidden in the "actions" tab
I only look at actions which occur related to a PR
lucifer
true that.
alastairp
what about a middleground of linking it to the prod release workflow?
again, not great (in fact, we already have the prod build sub-job exactly to verify that it's working before we do a release)
lucifer
prod release workflow runs for all commits though.
alastairp
I mean the action that builds and pushes the prod image
lucifer
oh i see.
alastairp
that runs on release create
but that's also problematic because a failure in docs will prevent us from releasing
lucifer
we could release first and then build docs.
any doc fixes come later but at least we are aware of the issue
alastairp
that being said, what are you objections to running on each PR/push?
from a selfish perspective, it's not costing us anything extra to just do that... and it makes us clearly aware of any problems that may occur due to a change
in fact, we could also publish the docs as an artifact, allowing us to build the docs and view them after making changes
lucifer
not much, only that most commits won't affect docs so seems wasteful but then no other options seem good either.
publishing docs as artifact sounds great
i'll look into adding this next week then. we can start at each commit and then revise if something better comes up later
alastairp
we could do a daily or weekly cron job for now, and see if we pay attention to failures. if we realise in 6 months that it's happened again, we can come back and decide if we want to link them to a PR
lucifer
that sounds good as well.
alastairp
lucifer: we talked a while back about moving messybrainz tables into LB. did we open a ticket for this?
lucifer
not sure, let me look.
alastairp
mayhem and I came up with some further ideas yesterday where we could reuse some of the metadata that we have there. We'd like to bundle these tickets together
ansh
alastairp: Hi! If you have some time, can you please review CB#441?
another thing to talk about in person at the summit
ansh: sure thing
lucifer
we should also decide on the hash to use going forward at that time.
alastairp
yes, good remembering
lucifer
no i don't see a ticket for that currently.
alastairp
I can make one at the same time
lucifer
i was actually thinking on working on that as well as part of the tech debt stuff i have planned for august. we can discuss it some time next week if time permits and then i can start work on it as well.
thanks!
alastairp
👍
next week I'm travelling at a conference but technically only unavailable due to presenting on Thursday. I should be available for at least part of the day mon-weds
lucifer
sounds good.
mayhem, alastairp: can we plan a meeting for early next week to discuss some PRs/tickets which I intend to work on for upcoming weeks?
alastairp
fine by me, monday would be best (before meeting?) though I'm unsure when/if mayhem is travelling
lucifer
or tomorrow maybe if that is better?
mayhem
monday is good. not traveling until wed
zas: atj : blocked IP request in support@
lucifer
cool Monday 1 hr before meeting then?
mayhem
sure
alastairp
great
rdswift
skelly37, I added a comment on your documentation PR.
what if I get a 10 trip ticket for free and go on only one trip?
well, typical spain in the ass.
lucifer
yeah i am not sure either. maybe some spanish news outlet has more info.
mayhem
they do. it doesn't make any more sense.
Sophist-UK joined the channel
lucifer
lol
Sophist-UK has quit
mayhem
there are some peculiar things about living here.
skelly37
rdswift: thanks for the feedback, i've pushed your suggestions. When it comes to the discussion about wrapping, I leave it up to you and if you think it should be changed, just open a PR
rdswift
skelly37: 👍
skelly37
rdswift: Also I think we should talk about the further single-instance improvements, i.e. starting some operations in the existing instance via cli with zas and outsidecontext. Since URLs went so smoothly.
I'd like to hear what all of you'd like to include
And what are your ideas about invoking such commands
My idea is to add `-e` (as "execute") flag and there we could send signals like `picard -e quit`
skelly37, what commands did you have in mind to initiate from the command line? Doesn't most of the processing depend on selecting something in the cluster pane or the album pane first?
skelly37
rdswift: Actually, this wasn't my idea, so I'll just copy outsidecontext's message: "afterwards we could also extend the pipe communication to pass not only paths but commands + data and we could add command line parameters to trigger picard to cluster, lookup or scan unmatched files."