These are chat archives for rosshinkley/nightmare

18th
Apr 2016
Ross Hinkley
@rosshinkley
Apr 18 2016 03:56
oh man
@Mr0grog at first blush, that looks super elegant
(referring to segmentio/nightmare#579)
Rob Brackett
@Mr0grog
Apr 18 2016 04:01
Tried to keep it simple
Ross Hinkley
@rosshinkley
Apr 18 2016 04:02
goal accomplished
Rob Brackett
@Mr0grog
Apr 18 2016 04:06
Oh, just realized I forgot to mention in the notes section that respondTo can only hold one responder per name and calling it a second time with a new name will silently bump the old responder
rosshinkley @rosshinkley nods
Ross Hinkley
@rosshinkley
Apr 18 2016 04:07
i saw that in the source
man. i need to look at this with fresher eyes than what i have right now
Rob Brackett
@Mr0grog
Apr 18 2016 04:08
ha, yeah, no rush on it
Ross Hinkley
@rosshinkley
Apr 18 2016 04:08
oh, i'm in a hurry ;)
Rob Brackett
@Mr0grog
Apr 18 2016 04:08
has big implications, so it should stew for a bit
Ross Hinkley
@rosshinkley
Apr 18 2016 04:08
more to the point, i want to pull it down and play with it a bit
Rob Brackett
@Mr0grog
Apr 18 2016 04:08
hopefully garner other eyes, too
rosshinkley @rosshinkley nods
Ross Hinkley
@rosshinkley
Apr 18 2016 04:09
i take it you don't like the rob n' ross show? ;)
Rob Brackett
@Mr0grog
Apr 18 2016 04:09
hahahaha
well, the thing I’d like to see
is to bring people who’ve cared enough about things to fork look in
Ross Hinkley
@rosshinkley
Apr 18 2016 04:10
i agree.
Rob Brackett
@Mr0grog
Apr 18 2016 04:10
and maybe slowly bring their own changes back in if you stick around and be active ;)
e.g. @fritx
Ross Hinkley
@rosshinkley
Apr 18 2016 04:10
<== not going anywhere anytime soon
Rob Brackett
@Mr0grog
Apr 18 2016 04:11
:thumbsup:
Ross Hinkley
@rosshinkley
Apr 18 2016 04:11
this is much cooler than what i was puttering with today
i've been monkeying with removing the make
and realizing in the process that i didn't understand spawn as well as i thought i did
Rob Brackett
@Mr0grog
Apr 18 2016 04:12
ah-ha
lots of options with child processes there
oh, nice
Ross Hinkley
@rosshinkley
Apr 18 2016 04:13
yeah, and what i ended up running into was xvfb-run creating child processe
s
and i probably would not have caught it, but i don't have xrandr set up on this chroot
and xvfb-run balks at that, citing the servernum
... which kept incrementing
Rob Brackett
@Mr0grog
Apr 18 2016 04:14
oh
Ross Hinkley
@rosshinkley
Apr 18 2016 04:14
meaning the kill call wasn't getting honored
anyway. I think i have it sorted, but i need to do some cleanup
and testing
and i'm realizing now your goto fix is still sitting :|
(a thousand apologies)
that should be the first thing i do tomorrow
Rob Brackett
@Mr0grog
Apr 18 2016 04:18
enh
Ross Hinkley
@rosshinkley
Apr 18 2016 04:18
okay, second. Coffee first.
Rob Brackett
@Mr0grog
Apr 18 2016 04:18
ha
Ross Hinkley
@rosshinkley
Apr 18 2016 04:19
;)
Rob Brackett
@Mr0grog
Apr 18 2016 04:19
honestly if I were going to make noise about something sitting, I'd point at the crasher first :P
Ross Hinkley
@rosshinkley
Apr 18 2016 04:20
yyyyeah, i just looked at that + the webcontents hang item
i suppose the crash might be patch-hacked around with a magic wait
in your example, if you change the navigate wait to something obnoxious like 5s, does it still crash intermittently?
Rob Brackett
@Mr0grog
Apr 18 2016 04:26
which example?
the one in the Electron bug?
Ross Hinkley
@rosshinkley
Apr 18 2016 04:26
yeah, the one in 5180
Rob Brackett
@Mr0grog
Apr 18 2016 04:26
hmmm, I never tried turning it up that high
I suspect yes
I think the issue is just that devtools is open at the time of the navigation
which is why it seemed to happen less without the delay
Ross Hinkley
@rosshinkley
Apr 18 2016 04:28
mmmmmm that makes me also wonder if it's less likely/doesn't happen after initial nav
Rob Brackett
@Mr0grog
Apr 18 2016 04:28
I haven’t dug deep enough in the Electron/Chromium source to fully understand where the reference goes bad
oh, interesting
Ross Hinkley
@rosshinkley
Apr 18 2016 04:28
electron gets squirrely if you don't navigate to something first
Rob Brackett
@Mr0grog
Apr 18 2016 04:28
I actually didn't try
just getting the hostWebContents before any navigation at all
Ross Hinkley
@rosshinkley
Apr 18 2016 04:30
might be worth it for grins
"don't nagivate to something"
you mean have a webcontents loaded? sure
rosshinkley @rosshinkley sighs.
Ross Hinkley
@rosshinkley
Apr 18 2016 04:32
i should get to bed
i'll catch you on good morrow, if you're about
Rob Brackett
@Mr0grog
Apr 18 2016 04:35
:thumbsup:
Ross Hinkley
@rosshinkley
Apr 18 2016 16:45
goooooood day
happy to see 5191 done, hopefully that gets released soon
Ross Hinkley
@rosshinkley
Apr 18 2016 17:10
@Mr0grog I think as time permits, i'll put together a pr with loglevel or something to try and address 421