That way we don't need to finish the event art bits
reosarevok hides
CatQuest
no.
reosarevok no hides
Dijia has quit
idgaf about event archive. but I *do* care about getting books-coverart in
like i want that this year already
hey bitmap that reminds me, who do we telk to on archive to get this started?
talk even*
reosarevok
yvanzo: your change and the comments both seem sensible to me :)
If you want to submit more commits with your ideas, I'll happily take a look and make sure it all looks good - if not, I can try to make some changes myself :)
yvanzo
reosarevok: Ok, thanks. I will continue in this direction then :)
reosarevok
Thanks!
bitmap
CatQuest: I'm not sure who the correct contact is for starting a new project tbh
though I do feel we should deploy the event art archive first before asking them to do more work for us, since they did a bunch of work on the EAA for us and we never ended up launching it
I'd like the focus on that after the schema change
zas, outsidecontext: Hello, the application time has just started. So, do you still think my application is ready to be sent? :)
zas
I do
v6lur joined the channel
asymmentric has quit
nviroepah9832 has quit
asymmentric joined the channel
asymmentric has left the channel
outsidecontext
skelly37: me too. Go for it 😃
yuzie
hi lucifer: If we use a dropdown for users to select their preferences, we can limit the selection field to IANA timezone identifiers that PG supports. What constraints for the timezone name do you suggest we need? For now, I put the timezone setting in the existing settings page as monkey suggested, and use a general setting table for reuse. For the API endpoint, I think one endpoint for one pref would be more flexible.
we cannot solely rely on client side validation because users can find ways to work around that. so the timezone name needs to be validated server side. whether we want to do it in api using some python lib or in PG using a snipped like above is the open question.
personally i am in favour of validating at the PG level but alastairp and mayhem have different opinion on this. also feel free to suggest which one you think is better.
yuzie
Yeah, I was thinking before that we can validate it in api to catch some exceptions
odnes has quit
I'm not quite familiar with how PG works with check
lucifer
it basically would boil down to PG erroring if you try to insert something that does not pass the check. for example you could `x INT CHECK (x != 42)` and then if you try to insert 42, it'll throw an error which translates to a python psycopg2 or sqlaclhemy exception depending on what you are using. you could catch the python exception and show the user an error. but if check is met then it inserts successfully.
I see, I think having a pg_timezone_names table and enforce correctness through foreign keys will be good. I will figure out how python deal with exceptions.
yvanzo
FWIW there is a timezone setting in MB editor preferences. Server side, it is checked against a list of allowed timezones from a Perl package. Client side, it is implemented in React.
lucifer
oh that's great if MB store timezones. we can follow the same thing.
yuzie
oh, thanks yvanzo. I didn't look at the code of MB yet, will take a look at that.
I see. I noticed javascript has specific way to handle the timezone feature. But LB basically chooses the flask framework. I'm not sure if we should follow react+Perl framework.
yvanzo
I’m not suggesting to use React+Perl at all :D Just showing an existing similar feature in another project.
yuzie
Yeah, that makes sense. Thanks yvanzo. That's a very nice reference for me.
skelly37
outsidecontext: I see you're still online, so one more question. Should I include info about me in the PDF for Google? Or was it just for you?
but this is actually invalid according to our XML schema, which says count should be a nonNegativeInteger
we recently changed this for user tags, but looks like we should do the same for aggregate tags too
skelly37
Also, how about my personal data like name etc.? Should I worry only about the technicals before the project gets accepted?
yvanzo
bitmap: Do you mean returning 0 instead of a negative value?
bitmap
I mean filtering out all tags with negative votes. we could still output them in the JSON WS, though
yvanzo
Is there any known use of downvoted tags through the webservice anyway? Picard?
bitmap
not sure, IIRC there's a setting to specify the minimum vote count. I suppose you could set that negative, but I don't know why you would
we could also change the XML schema
ikroop joined the channel
aerozol
alastairp: MORE scary than the twilight zone??
Thanks, I'll check it out today
(CB admin panel)
yvanzo: I think I remember reading on the forums that Picard returns your own tags even if they're vote count is 0 or under (or there's a setting for it)
yvanzo
bitmap ^
bitmap
we used to but we changed the web service a few months ago to stop returning those
