00:06:41  * jugglinmikequit (Quit: Leaving.)
00:09:44  * plutoniixjoined
00:42:30  * bradleymeckjoined
00:45:59  * bradleymeckquit (Client Quit)
01:00:20  * bradleymeckjoined
01:16:55  * Cube8joined
01:23:50  * Cube8quit (Ping timeout: 255 seconds)
02:36:24  * plutoniixquit (Quit: Leaving)
02:43:45  * bradleymeckquit (Quit: bradleymeck)
02:55:12  * kenansulaymanquit (Ping timeout: 260 seconds)
03:01:34  * plutoniixjoined
07:43:07  * rmcilroyquit (Ping timeout: 246 seconds)
07:54:50  * rmcilroyjoined
09:01:24  * plutoniixquit (Quit: Leaving)
09:22:49  * plutoniixjoined
10:25:10  * mylesborinsquit (Remote host closed the connection)
10:25:36  * mylesborinsjoined
10:55:41  * plutoniixquit (Remote host closed the connection)
12:27:16  * Cube8joined
12:57:12  * bradleymeckjoined
13:59:33  * jugglinmikejoined
14:08:10  * plutoniixjoined
14:20:57  * jugglinmikequit (Ping timeout: 240 seconds)
14:36:28  * jugglinmikejoined
15:44:15  * kenansulaymanjoined
15:45:43  * zvquit (Ping timeout: 255 seconds)
16:27:45  * ebarrettquit (Quit: WeeChat 1.7)
17:03:04  * RT|Chatzillaquit (Quit: ChatZilla [Firefox])
17:27:34  * zvjoined
17:42:09  <trungl-bot>Tree closed by [email protected]: closed (https://build.chromium.org/p/client.v8/builders/V8%20Linux64%20TSAN/builds/16737 from 4af9cfccf601f512b0fa6d9d5042684d66e2e9ca)
17:49:50  * seventhjoined
17:51:12  <trungl-bot>Tree opened by [email protected]: open (ulan is fixing TSAN failure)
18:07:47  <aklein>caitp: any opinion on https://chromium-review.googlesource.com/c/609365/1/src/parsing/parser.cc#b2991?
18:08:07  <caitp>don't care
18:12:01  <aklein>k :)
18:12:13  <aklein>we can always add it back if it actually does something
18:12:20  <aklein>I removed it since it was always passing Parser::scope()
18:12:50  <aklein>thanks
18:20:24  <bradleymeck>gsathya: it did come to my mind that Node has to create *lots* of throw away v8::Modules as well, so idk if that is relevant w/ the GC concern
18:21:06  <bradleymeck>2 Modules per CJS, 1 per ESM
18:25:22  <gsathya>bradleymeck: how does one overload the resolution behavior? what are the hooks you mention?
18:25:56  <bradleymeck>gsathya: put on hold but we do have some use cases that need to be filled by them
18:26:14  <bradleymeck>API isn't as important but guy bedford was the person in charge
18:26:56  <bradleymeck>roughly the idea is you get [Origin: Symbol/Object representing dependent, specifier: string]
18:27:09  <bradleymeck>and you return a Module wrapper of some kind
18:27:39  <bradleymeck>in the PR this is doable by changing the `.link` property of a module wrapper if I remember right
18:28:09  <bradleymeck>but we haven't exposed that in any way to userland since lack of dynamic import makes it hard to judge what works
18:30:22  <gsathya>bradleymeck: i don't see why the Origin can't be a string? (assuming the Origin is equivalent of the referrer, here)
18:30:32  <bradleymeck>cause we don't always have strings
18:30:40  <gsathya>right, so it's undefined in those cases
18:30:50  <gsathya>it's up to the custom loader to deal with those cases
18:30:58  <bradleymeck>and then we still need to figure out the resolve function for that Origin
18:31:10  <bradleymeck>so idk how I can do that when 2 things have undefined but diff resolve functions
18:31:28  <bradleymeck>how can I determine the origin at all to differentiate?
18:35:40  <gsathya>so I can see two solutions when Origin is undefined (a) look at specifier and try to load it if it's an absolute path (b) throw an error. Is there any other acceptable behavior?
18:39:32  <gsathya>other custom behavior could be out of scope for now, until the hooks are more fleshed out.
18:48:06  <bradleymeck>gsathya: with userland loaders pretty much the main use case is to change how the resolve algorithm works. I can't say I know all possible ways in which people desire to change the algorithm
18:48:36  <bradleymeck>first use case is web compatible resolution mechanics
18:48:47  <bradleymeck>what will happen if there are Scripts with duplicate file names?
18:49:42  <bradleymeck>I guess it just calls w/e the current one is hmmmm
18:50:04  <bradleymeck>(this comes up looking at things like the new userland @std/esm loader)
18:51:34  <bradleymeck>in normal CJS, when a script fails to load it is removed from the cache (and can be GC'd)
18:51:56  <bradleymeck>so you can require that file again and it will have the same name but might have different context around it
18:52:42  <trungl-bot>Tree opened by [email protected]: open
19:15:07  * bradleymeckquit (Quit: bradleymeck)
19:28:56  <trungl-bot>Tree closed by [email protected]: closed (https://build.chromium.org/p/client.v8/builders/V8%20Linux64%20TSAN/builds/16740 from a7ab836aba387050d1d0bfc33af9893f83028136)
19:29:57  <trungl-bot>Tree opened by [email protected]: open
19:31:07  * zvquit (Ping timeout: 255 seconds)
19:34:45  * bradleymeckjoined
19:46:48  <gsathya>bradleymeck: ok let me think about this a little more, i'll ping the bug once I have an update
19:47:01  <gsathya>thanks for all the explanations
19:53:25  * seventhquit (Remote host closed the connection)
20:36:24  * zvjoined
21:24:38  * bradleymeckquit (Quit: bradleymeck)
21:31:57  * bradleymeckjoined
21:59:10  * Guest59quit (Quit: My Mac has gone to sleep. ZZZzzz…)
22:13:45  * xiinotulpjoined
22:14:59  * bradleymeckquit (Quit: bradleymeck)
22:17:05  * plutoniixquit (Ping timeout: 248 seconds)
22:24:04  * zvquit (Ping timeout: 240 seconds)
22:33:25  * xiinotulpquit (Quit: Leaving)
22:39:25  * Guest59joined
22:50:16  * RT|Chatzillajoined
22:59:36  * Cube8quit (Quit: Leaving)
23:36:35  * paulfryzelquit (Quit: Connection closed for inactivity)