Apache Allura / Chat is hosted on FreeNode IRC channel #allura
a little lengthy but most of those changes are trivial
You could use pop('TicketMonitoringEmail', None) instead of try / except / pass. But there isn't any fallout from cahnging results['tracker_config'] from the actual config instance to the json() dict?
catching KeyError does also protect if the 'options' isn't there for some reason. that'd be unusual though
since the method is @expose('json') the json() is called by TG
True. I had considered that but figured it shouldn't really come up. I'd leave it with the except / pass, though, just to be safe
.. called automatically, now its explicit
Fair enough. So, I guess if all those many test changes pass, seems ok
yuriyarhipovua:  ticket:428 Tracker API include config & milestones & saved bins
dbrondsema:  don't expose TicketMonitoringEmail in API, and associated TestRestApiBase changes
i'm pretty sure the cron that pulls new git code isn't triggering the refresh
but i don't have access to the cron, or don't know where it is
yeah, didn't see anything in the access logs from curl or wget etc
i assume that's in the crontab for the allura or root user?
yeah. i should figure out how to get my access set up on the VM
updated the domain, presumably it'll work now
cool, thanks :)
np, should've checked that when you mentioned it before, didn't even occur to me
tvansteenburgh: can we expand that cron to pull & push all branches & tags?
ok, i changed it to
already a generic
git pull since it's using the update.sh script
hm, guess i need
git push --tags instead?
iirc --tags does only tags
git push --all ?
--all --tags ?
i thought --all was just heads
yeah, both it looks like
both params are needed
ok, cron updated
for pull, doesn't a generic
git pull just do the current branch?
there's an --all but docs indicate that's only for fetching, not merging into local
getting confused now
do we need to locally track all branches?
not necessarily, but we want to show them all on the web I think
it seems a little complicated to me since we're pulling into one repo location and then pushing into the one we use for web browsing
ok, i think i've got it sorted
will see when the repo refresh finishes
man, if only we didn't have this ridiculously long analyzing step :)
must be looking at a lot of new branches
now the page takes forever to load, but still only shows master
i'll look into it tomorrow, don't have time right now, time for dinner
tags are there though