<dwrensha>
asheesh: do we have a way to explicitly request that `spk publish` go to the experimental app market?
jemc has joined #sandstorm
jemc has quit [Client Quit]
jemc has joined #sandstorm
<asheesh>
It does by default! If no one "approves" or "rejects" it, it stays in experimental.
<dwrensha>
it's immediately put in the experimental app market ?!
<asheesh>
As far as I know, yes absolutely.
<dwrensha>
I guess that means one could do `spk publish --embargo` to do accomplish my goal
tantalum has joined #sandstorm
aeviator has joined #sandstorm
aeviator has quit [Ping timeout: 265 seconds]
amyers has quit [Remote host closed the connection]
amyers has joined #sandstorm
aeviator has joined #sandstorm
mnutt_ has joined #sandstorm
jemc has quit [Quit: WeeChat 1.4]
jemc has joined #sandstorm
aeviator has quit [Ping timeout: 240 seconds]
mnutt_ has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
mnutt_ has joined #sandstorm
amyers has quit [Ping timeout: 255 seconds]
amyers has joined #sandstorm
amyers has quit [Remote host closed the connection]
amyers has joined #sandstorm
mnutt_ has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
mnutt_ has joined #sandstorm
<asheesh>
win 1
<asheesh>
...
<mrdomino>
apparently my crazy RENAME_EXCHANGE hack is non-portable -- it breaks on Void
<mrdomino>
also every aspect of file io in that app is really dumb right now, including that i don't just assume small-ish appends are atomic and that i don't keep any fds open
aeviator has joined #sandstorm
jemc has quit [Ping timeout: 255 seconds]
ThePurgingPanda has joined #sandstorm
ThePurgingPanda_ has quit [Ping timeout: 244 seconds]
afuentes has quit [Ping timeout: 240 seconds]
<dwrensha>
mrdomino: time to start using sqlite?
<mrdomino>
maybe. i'll need it for the next app, but i sorta like the aggressive (militant?) simplicity of a flat file for this one
jemc has joined #sandstorm
<mrdomino>
i was actually considering just taking out the file store entirely and letting it be like an unlogged irc channel, but i'm pretty sure that would be bad and surprising
<mrdomino>
you really don't want a plain irc connection -- you want irc with tmux, which the file store sorta emulates
Telesight has joined #sandstorm
<crw>
congrats on sandstorm for work launch.
indiebio has joined #sandstorm
amyers has quit [Read error: Connection reset by peer]
amyers has joined #sandstorm
<kentonv>
crw: thanks!
<pod>
has anyone noticed that when they use gitlab they're being asked constantly to allow notifications?
<pod>
seems that w/ each new subdomain chrome needs a fresh authorization from the user
<dwrensha>
pod: weird, when I start gitlab grains they don't ask me for permission to send desktop notifications
<dwrensha>
pod: to get desktop notifications to play nice with Sandstorm, GitLab would need to go through Sandstorm's activity/events API
<dwrensha>
(as rocket.chat does)
<dwrensha>
...but I'm puzzled as to how you got gitlab to send desktop notifications in the first place
<pod>
to clarify it hasn't sent any AFAIK
<pod>
but chrome is asking me to authorize it anyway
<pod>
I'll screenshot the next time it happens
mnutt_ has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<zarvox>
pod: oh! That will happen if any of your grains are reporting activity events. So probably a rocket chat grain is trying to notify you about chat messages.
<zarvox>
We should probably make the permission prompt be user-directed in the notifications dropdown...
<kentonv>
zarvox: it sounds to me like he is actually describing gitlab requesting access (from many different subdomains)
<zarvox>
Hmmm, perhaps.
<kentonv>
maybe gitlab has a notification feature we didn't know about
<zarvox>
sound like it!
<kentonv>
that only triggers after some specific actions have taken place
mnutt_ has joined #sandstorm
isd has joined #sandstorm
mnutt_ has quit [Client Quit]
<mrdomino>
interesting, "server returned error on upload: Error: remote exception: remote exception: Peer disconnected.
<mrdomino>
"
<mrdomino>
on spk publish
<kentonv>
mrdomino: hmm, did you just try this like five times?
<mrdomino>
yeah, i tried --embargo and some other stuff
<kentonv>
something about your package is crashing the app index app... -_-
<kentonv>
can you share your spk somewhere so I can debug?
mnutt_ has joined #sandstorm
aeviator has quit [Ping timeout: 252 seconds]
amyers has quit [Read error: Connection reset by peer]
<dwrensha>
does the app index app need a license or a screenshot?
<kentonv>
well.... it shouldn't _crash_ regardless.
<pod>
dwrensha: you maintain gitlab on sandstorm yes?
<dwrensha>
pod: yeah
<pod>
if so have a potential bug report, if you have 30 seconds to hear it before it goes on GH
<dwrensha>
ok
<pod>
the "merge request" action never seems to finish
<pod>
rather, the button continues to say "merge in progress" indefinitely, even though the underlying action completes (as verified by checking the activity history in another tab)
Telesight has quit [Quit: Leaving.]
<pod>
huh maybe it's just really slow? worked for me when I attempted to reproduce it
<dwrensha>
weird
<gwillen>
Congrats on the SFW release! :D
<kentonv>
goddammit, I sent that mail out with a localhost link at the top
<kentonv>
:(
<mrdomino>
is there any way to detect when a WebSession has ended?
<dwrensha>
mrdomino: Sandstorm should drop it once it is no longer in use
<gwillen>
kentonv: whoops :-(
<dwrensha>
mrdomino: so you could add something to the destructor of your object that implements WebSession
<mrdomino>
ok, great
<mrdomino>
i suppose that'd time out after 90 seconds or so
tantalum has quit [Quit: Leaving.]
<isd>
So, with offer templates, it generates a new key every time you load the page. How do those get collected? Do I have to do anything to keep from leaking them, and what can I rely on as far as them still being valid?
<dwrensha>
isd: they get automatically deleted if they are not used within five minutes of closing the page
<dwrensha>
one they are used, they are persistent
biotim has joined #sandstorm
<isd>
dwrensha: thanks.
<kentonv>
dwrensha: is it really 5 minutes? I think we reduced the time?
<kentonv>
or maybe we only reduced the time after which the app is killed if not in use
<dwrensha>
maybe it's more like 60 seconds?
<isd>
60 seconds seems awfully short
<isd>
Well, I guess if it's after the page is *closed*
<dwrensha>
the idea is the usually you will keep the page open
<kentonv>
right, while you have it open, Sandstorm automatically sends keep-alives