<mtrbot-ml>
[mattermost] <pkulik> @sb10q hey, did you do stub port for metlino?
mauz555 has joined #m-labs
cedric has quit [Ping timeout: 245 seconds]
cedric has joined #m-labs
cedric has quit [Changing host]
cedric has joined #m-labs
<mtrbot-ml>
[mattermost] <sb10q> @astro I would rename mattermost-github-integration/pkg.nix to mattermost-github-integration/default.nix
<mtrbot-ml>
[mattermost] <sb10q> @pkulik not yet
<mtrbot-ml>
[mattermost] <astro> @sb10q note that that is not the top-level file of the subdir but is only imported by uwsgi-config.nix
<mtrbot-ml>
[mattermost] <astro> hence renaming to default.nix may no be too logical :)
<lkcl>
whitequark: ping, i'm going to ask someone on my team to raise this as a bugreport, just wanted to give you a heads-up: use of type() instead of isinstance is extremely bad practice
<lkcl>
not many people are aware, type() has 3 arguments and is actually a dynamic meta-class-constructor function.
<lkcl>
the first argument is the "usual" one, the second argument is a list of multiple-inheritance base classes, and the third argument is a list of methods and properties.
<lkcl>
it creates a *huge* processing overhead, *dynamically* constructing a class each time just for the purposes of calling hash() on the result
<lkcl>
and it prevents and prohibits derivatives from being used in any place where it's called.
<lkcl>
one of the places was detected last week (type(x) == Record replaced with isinstance(x, Record)
<sb0>
whitequark: do you have an ecp5 versa board?
<lkcl>
whitequark: "the key thing to retain about ABCs .... s that they make isinstance (and issubclass) much more attractive ... make checking type an even worse practice than it already used to be"
<lkcl>
that's not to say that there aren't good reasons for using it (the need to match *only* one class rather than its inherited subclasses)
<lkcl>
however the way to do that is either "duck typing" (a string/enum in the constructor, assigned to a known-variable), or, i've simply used "x.__class__.__name__ == 'foo'" in the past, or, better (apparently), ABCs.
mauz555 has quit []
cr1901_modern has quit [Ping timeout: 250 seconds]
<mtrbot-ml>
[mattermost] <sb10q> ping from m-labs/artiq
<mtrbot-ml>
[mattermost] <github> github joined the team.
<mtrbot-ml>
[mattermost] <github> ping from m-labs/artiq
<mtrbot-ml>
[mattermost] <nakengelhardt> ok that didn’t work since mattermost tried to interpret it
<mtrbot-ml>
[mattermost] <sb10q> ok i see
<mtrbot-ml>
[mattermost] <sb10q> whitequark: does filtering the markup in the bridge sound like an acceptable compromise?
<attie>
oh interesting the photo didn't get through the bridge
<whitequark>
sb: does that mean the notifications will not have clickable links?
<whitequark>
that makes the noficiations almost entirely useless functionally
<whitequark>
since i'd have to go through the web interface to do anything
<attie>
I still don't see what's wrong with not showing the user icon in the notice frankly, why is it so essential?
<attie>
if you remove the other two links and leave just the link to the actual issue, it should look ok on both sides I think
<mtrbot-ml>
[mattermost] <sb10q> the other option is to run notifico and mattermost-github-integration at the same time, and disable bridging for those bots
<whitequark>
I'd say that would be much preferable
<mtrbot-ml>
[mattermost] <sb10q> does notifico work well with uwsgi?