00:16:45  <devsnek>does anyone know of an eslint plugin that will enable parsing natives syntax
00:36:21  * plutoniixjoined
00:38:14  * AtumTquit (Remote host closed the connection)
01:52:04  * Garbeequit (Quit: Connection closed for inactivity)
02:18:01  * xaxxonquit (Ping timeout: 256 seconds)
04:01:39  * xaxxonjoined
06:34:12  * xaxxonquit (Quit: xaxxon)
08:45:50  <trungl-bot>Tree closed by [email protected]: closed (https://build.chromium.org/p/client.v8/builders/V8%20Linux%20-%20debug/builds/19794 from 9a0ccf4f49a4de40ba6c1c2dc8aaca3a4c311d88)
08:57:54  * microcolonelchanged nick to microcolon
09:13:02  <trungl-bot>Tree opened by [email protected]: open
10:25:34  * mylesborinsquit (Quit: farewell for now)
10:26:04  * mylesborinsjoined
11:29:44  * AtumTjoined
11:33:40  * plutoniixquit (Quit: Leaving)
11:36:15  * Garbeejoined
13:47:48  * bradleymeckjoined
13:55:41  * trevnorrisquit (Ping timeout: 276 seconds)
13:57:02  * trevnorrisjoined
14:07:31  <devsnek>how is all this debug stuff like MakeMirror consumed and what by
14:07:38  <devsnek>i don't see it being use anywhere in the v8 source
14:10:03  <bradleymeck>devsnek: you used to be encouraged to get the debug context and run stuff against it, the context was bootstrapped by debug.js
14:10:20  <devsnek>yea i know there used to be the DebugContext api
14:10:21  <devsnek>but it was removed
14:10:46  * bradleymeckdigs about
14:10:49  <devsnek>i just can't find what is still using it
14:11:14  <devsnek>i see there are still tests for the internal debug_context thing
14:12:19  <bradleymeck>i think the devtools protocol impl still uses it
14:12:25  <bradleymeck>but it will take a sec for me to dig it up
14:12:29  <devsnek>is that in a separate repo
14:13:44  <bradleymeck>lots of it is so I'm trying to cross ref places
14:13:51  <devsnek>ah ok thx
14:16:16  <bradleymeck>devsnek: it looks like mirrors were slated to be phased out (at least reading some 2016 stuff), see RemoteObject for how to use things today
14:16:21  <bradleymeck>so nothing should be using it anymore
14:16:35  <devsnek>RemoteObject sounds very uh
14:16:44  <devsnek>suggestive of the object not being in the same process
14:16:50  <devsnek>or isolate
14:17:02  <bradleymeck>not in the same heap, yea
14:17:18  <bradleymeck>idk if you can get them into the original isolate, would be interesting to know
14:17:51  <devsnek>i'm just messing around shrug
14:18:29  <devsnek>lol here's some interesting code `accumulator->Add(global_object ? "<RemoteObject>" : "<RemoteObject>");`
14:59:00  * bradleymeckquit (Read error: Connection reset by peer)
14:59:31  * bradleymeckjoined
15:02:36  <trungl-bot>Tree closed by [email protected]: closed - maintenance
15:31:47  * AtumTquit (Quit: No Ping reply in 180 seconds.)
15:33:30  * AtumTjoined
15:37:51  <trungl-bot>Tree opened by [email protected]: open
16:04:48  * bradleymeckquit (Quit: )
17:02:52  * RT|Chatzillaquit (Quit: ChatZilla 0.9.86.1 [Firefox 2.0.0.22pre/2010030309])
17:03:23  * bradleymeckjoined
17:27:46  * seventhjoined
17:40:18  * Wes-quit (Remote host closed the connection)
18:20:48  * Wes-joined
18:26:31  <Wes->Hey, folks - I am evaluating a script with Script::Compile->run. The script ends with a string. Normally, the script terminates normally, and it evaluates to the string. If it throws, a different string is evaluated. Sometimes, though.......the script just stops running and it evaluates to undefined. Any idea how I can figure out what's going on in that case??
18:30:05  <caitp>Wes-: you mean it's always the same script but you get different results sometimes?
18:34:33  <Wes->caitp: almost....it's a script that loads other scripts, which are eval'd with eval(). These are nearly identical (running math of some kind). But the the container is always the same and has a global try..catch block.
18:36:32  <caitp>well, there's probably some path which is not capturing the completion properly
18:37:33  <caitp>hard to say what exactly, but you can look at src/rewriter.cc to get hints... tracing the path through bytecode will also be helpful in figuring it out
18:39:59  <caitp>the way v8 captures completions for scripts/eval is kind of gross, I'm hopeful we can change it to something easier to follow (like JSC's approach), but for now you have to deal with AST rewriting
18:40:06  <caitp>which is a bit hard to reason about
18:41:44  <Wes->caitp: That was my thinking too... Trying to figure out what code paths could possibly not be capture by try { everything; } catch { "unhandled exception" }. I'll take a boo through rewrite.cc etc in a bit. Right now I'm trying to figure out if it's possible to attach the remote debugger to an embedding without too much work.
18:44:04  <Wes->Hmm, I wonder if I'm hitting OOM
18:45:24  <caitp>well, OOM tends to fail more like "just crashing"
18:45:54  <caitp>unless you have OOM handlers set up that prevent hard crashes
18:46:03  <Wes->caitp: Does it? Darn. On spidermonkey, it's essentially an uncatchable exception. Was hoping it was similar on v8.
18:46:26  <Wes->caitp: Nothing like that.......still throwing C++ exceptions when script exceptions are uncaught. This embedding is in its absolute infancy.
18:46:32  <caitp>by default it should be a hard crash
18:46:49  <Wes->Okay, I can at least strike that off my investigation list.
18:48:30  <caitp>oh, I guess you can't prevent the hard crash, you can just have code run before the hard crash
20:01:40  * trevnorrisquit (Ping timeout: 276 seconds)
20:05:12  * trungl-botquit (Ping timeout: 276 seconds)
20:05:42  * trevnorrisjoined
20:07:20  * trungl-botjoined
20:41:55  * bradleymeckquit (Quit: bradleymeck)
20:52:33  * seventhquit (Remote host closed the connection)
22:17:41  * bradleymeckjoined
22:31:40  * RT|Chatzillajoined
23:24:57  * bradleymeckquit (Quit: bradleymeck)
23:53:08  <devsnek>bradleymeck: i went full crazy https://github.com/devsnek/v8-debug