00:26:48  * plutoniixjoined
01:41:12  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
01:56:47  * rwlbuis_quit (Quit: Connection closed for inactivity)
02:16:22  * bradleymeckjoined
03:04:42  * bradleymeckquit (Quit: bradleymeck)
05:02:39  * plutoniixquit (Quit: Leaving)
06:09:29  * plutoniixjoined
06:10:13  * plutoniixquit (Max SendQ exceeded)
06:10:48  * plutoniixjoined
06:55:05  * Keverwjoined
07:06:18  * Keverwquit (Quit: Textual IRC Client: www.textualapp.com)
07:16:40  * plutoniixquit (Ping timeout: 260 seconds)
07:30:02  * plutoniixjoined
07:31:04  * plutoniixquit (Max SendQ exceeded)
07:31:32  * plutoniixjoined
08:00:26  <trungl-bot>Tree closed by [email protected]: closed - maintenance
08:11:29  <trungl-bot>Tree opened by [email protected]: open
08:34:58  * Net147quit (Quit: Quit)
09:20:14  * privikjoined
09:20:30  <privik>Can anyone explain to me why new wins over bind?
09:39:27  * plutoniixquit (Quit: Leaving)
09:43:46  * plutoniixjoined
09:44:23  * plutoniixquit (Max SendQ exceeded)
09:44:57  * plutoniixjoined
10:02:21  <trungl-bot>Tree closed by [email protected]: closed - maintenance
10:25:10  * mylesborinsquit (Quit: farewell for now)
10:25:41  * mylesborinsjoined
10:51:31  * plutoniixquit (Quit: Leaving)
10:59:38  <trungl-bot>Tree opened by [email protected]: open
12:51:22  * bradleymeckjoined
13:10:35  * unixpicklejoined
13:18:41  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
13:26:39  * unixpicklejoined
13:30:56  * unixpicklequit (Client Quit)
13:33:12  * bradleymeckquit (Quit: bradleymeck)
13:50:02  * trungl-botquit (Ping timeout: 246 seconds)
13:51:15  * plutoniixjoined
14:01:53  * trungl-botjoined
14:03:10  * bradleymeckjoined
14:04:31  * Net147joined
15:06:14  * unixpicklejoined
16:09:21  * bradleymeckquit (Quit: bradleymeck)
16:34:40  * RT|Chatzillaquit (Quit: ChatZilla [Firefox])
16:43:35  <littledan>privik: What do you mean?
16:44:06  <privik>littledan: A function bound to an object loses its this binding with new.
16:44:52  <privik>littledan: So there is something that makes the function returned by bind know when we are constructing a new object or when we are calling it plain.
16:45:27  <littledan>yeah, that's right, this is per spec, see https://tc39.github.io/ecma262/#sec-bound-function-exotic-objects-construct-argumentslist-newtarget
16:45:47  <littledan>the target used is the one in the bound function, even for new (which uses [[Construct]])(
16:45:58  <littledan>maybe bound functions aren't the right thing for your use case
16:47:09  <littledan>or, sorry, I'm confused--the [[BoundFunctionThis]] is only used in [[Call]]
16:47:23  <littledan>so, it looses the this binding with new
16:47:34  <littledan>(the target I was thinking of is actually the underlying function)
16:54:34  <caitp>how would it make sense if it didn't lose the binding with new?
17:09:58  * privikquit (Ping timeout: 240 seconds)
17:31:23  <littledan>well, it would make new on a bound function act just like call
17:31:28  <littledan>you can achieve this with a Proxy
17:31:37  <littledan>what does "make sense" even mean in JavaScript anyway?
17:34:02  * bradleymeckjoined
17:35:03  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
17:37:20  * bradleymeckquit (Client Quit)
17:41:07  * unixpicklejoined
17:42:45  * bradleymeckjoined
17:48:56  * bradleymeckquit (Quit: bradleymeck)
18:02:43  <caitp>I don't see how you can achieve it with a Proxy, the `this` is a new object every time
18:05:59  <caitp>i mean I guess you could avoid calling [[Construct]] without failing invariant checks, but that just makes `new` meaningless
18:23:26  <littledan>Yeah, Proxies can do a lot of meaningless things
18:35:18  * Guest59quit (Quit: My Mac has gone to sleep. ZZZzzz…)
18:36:55  * Guest59joined
18:39:08  * Guest59quit (Read error: Connection reset by peer)
19:30:49  * bradleymeckjoined
19:53:25  * Guest59joined
19:53:41  * Guest59quit (Client Quit)
20:10:58  * Kronuzquit (Ping timeout: 268 seconds)
20:15:18  * Kronuzjoined
21:20:40  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
21:33:52  <jwolfe>is the test webkit/fast/js/excessive-comma-usage specifically designed to fail in no-optimization builds? i'm getting a maximum call stack exceeded error, which is understandable, but are we making debug builds fail on purpose?
21:33:55  * bradleymeckquit (Quit: bradleymeck)
21:34:09  <jwolfe>do bots pass that test in debug builds? do they do any special configuration to increase the call stack limit?
21:34:47  <jwolfe>i guess the question is: do any bots run with optimizations turned off? i would imagine not.
21:41:13  * unixpicklejoined
21:52:25  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
22:09:35  <caitp>jwolfe: that test gets skipped on a number of slow configurations
22:10:18  * unixpicklejoined
22:10:24  <jwolfe>caitp: is there a way to skip tests using run-tests.py (...looking now)
22:10:51  <caitp>yes, everything with the SKIP flag in webkit.status (or mjsunit.status, or other status files)
22:13:16  <jwolfe>i mean from the command line.
22:13:25  <caitp>run-tests.py will skip it
22:13:28  <jwolfe>looks like --slow-tests=skip
22:13:30  <caitp>based on the status files
22:13:45  <caitp>if you use the msan variant on an arm64 build for instance, it will skip it
22:14:29  <caitp>so, if something changed and it's now too slow on architectures that aren't configured to skip it, lets see if trybots can still do it
22:16:03  <jwolfe>when i `python tools/run-tests.py bot_default tests262` with a gn configuration v8_optimized_debug=false, then that test runs and fails.
22:38:59  <caitp>you'd have to use a configuration that gets skipped
22:39:12  * bradleymeckjoined
22:39:47  <caitp>I don't believe there's a way to skip arbitrary tests other than the status files, but could be wrong
22:40:22  <caitp>now, if the test wasn't too slow to pass before, and there's a good reason for it, you may want to add a status line
22:51:41  * RT|Chatzillajoined
22:52:20  * plutoniixquit (Quit: Leaving)
22:56:53  <jwolfe>the dry run on my WIP CL passed even though i didn't do any `#ifdef V8_I18N_SUPPORT`. i assume this means none of the default bots test a no-i18n configuration. is there a CQ_INCLUDE_TRYBOTS= line i can add to the CL that will test some no-i18n configurations?
22:56:59  <jwolfe>this is the CL for reference: https://codereview.chromium.org/2728763006/
22:58:16  <caitp>not aware, you should ask jungshik I think
22:58:23  <caitp>or dan may know
23:09:51  <aklein>jwolfe: from a quick git grep, looks like CQ_INCLUDE_TRYBOTS=master.tryserver.v8:v8_linux_noi18n_rel_ng
23:11:59  <jwolfe>aklein: thanks i'll try that.
23:12:48  * bradleymeckquit (Quit: bradleymeck)
23:21:30  * bradleymeckjoined
23:24:33  * bradleymeckquit (Client Quit)
23:25:41  * bradleymeckjoined
23:25:46  * bradleymeckquit (Client Quit)
23:29:28  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
23:48:57  * bradleymeckjoined