00:16:32  * Fishrock123joined
00:42:42  <devsnek>i really like reflection just cuz being about to toy with stuff is epic
00:43:02  <devsnek>like devtools being able to introspect all the values and stuff is so cool
00:43:50  <Bakkot>Devtools can always do whatever, of course; the important question is whether it's part of the language.
00:43:58  <devsnek>indeed
00:44:31  <devsnek>in terms of what is built into the language i will tend to side with encapsulation
00:46:00  * isiahmeadowsjoined
00:56:54  * Fishrock123quit (Quit: Leaving...)
01:06:21  * AtumT_joined
01:09:09  * AtumTquit (Ping timeout: 256 seconds)
01:10:34  * AtumT_quit (Remote host closed the connection)
01:23:43  * not-an-aardvarkjoined
01:37:09  * isiahmeadowspart
02:00:38  * keith_millerjoined
02:17:19  * jwaldenquit (Quit: back later)
02:46:43  * gibson042joined
02:50:28  * tornodojoined
02:51:13  <tornodo>hi
02:52:12  * tornodoquit (Client Quit)
04:06:48  * jmdyckquit (Remote host closed the connection)
04:16:06  * keith_millerquit (Quit: My MacBook has gone to sleep. ZZZzzz…)
04:37:14  * gibson042quit (Ping timeout: 276 seconds)
07:43:10  * not-an-aardvarkquit (Quit: Connection closed for inactivity)
07:45:55  * keith_millerjoined
08:10:22  * keith_millerquit (Quit: My MacBook has gone to sleep. ZZZzzz…)
08:29:25  * keith_millerjoined
08:38:09  * keith_millerquit (Quit: My MacBook has gone to sleep. ZZZzzz…)
08:40:23  * keith_millerjoined
09:44:53  <ljharb>dev tools already have reflection superpowers, you can even synchronously inspect a promise
09:45:05  <ljharb>it seems obvious that private fields would be inspectable via dev tools as well
09:48:28  * JaseWjoined
09:48:32  * JaseWchanged nick to Jayflux
10:14:00  * keith_millerquit (Quit: My MacBook has gone to sleep. ZZZzzz…)
10:25:09  * mylesborinsquit (Quit: farewell for now)
10:25:40  * mylesborinsjoined
10:27:04  <littledan>"obvious" things require a lot of hard work by engineers who are plugged into our process, by the way
10:27:20  <littledan>whether that's through reading documents we produce or talking with us directly
10:28:34  <littledan>for example, seems like the notes show that it was "obvious" in specifying Annex B 3.3 that browsers should do something further than just sloppy mode hoisting of functions in blocks in functions, but then when I was implementing this in V8, I didn't get the memo and tried to ship just what was in the spec (and Firefox did the same)
10:28:59  <littledan>we should make sure to be in touch with DevTools people through this process, as well as in future proposals where these issues come up
10:32:17  <littledan>another "obvious" thing could be that BigInt implementations would be optimized in the 64-bit case; this is turning out a little different than the committee expected for V8, but seems like they have solid reasons and a plan for making future changes as needed. Fortunately, they are communicating openly with us about all of this.
10:33:30  * Jayfluxquit (Remote host closed the connection)
11:14:55  * AtumTjoined
11:32:38  * keith_millerjoined
11:40:07  * jmdyckjoined
11:58:24  * bradleymeckjoined
12:33:18  * bradleymeckquit (Quit: bradleymeck)
12:39:14  <devsnek>I thought v8 already optimised for thay
12:39:30  <devsnek>that
12:44:07  * gibson042joined
13:07:47  * bradleymeckjoined
13:36:21  * gibson042quit (Ping timeout: 252 seconds)
13:41:43  <littledan>optimized for what?
13:42:00  <littledan>oh, BigInt in the 64-bit case? not to my knowledge
13:44:42  <bradleymeck>we should also be aware of the editor workflows concern. most editors don't provide the devtools superpowers that we are speaking of
13:45:13  <bradleymeck>they often integrate debuggers but don't have the time and/or knowledge of working around things
13:52:03  * gibson042joined
14:15:44  * sebmarkbage|awayquit (Quit: ZNC - http://znc.in)
14:30:46  <littledan>bradleymeck: What kind of non-devtools debugger integration do they provide? (I use vim)
14:31:07  <littledan>I mean, I don't see a good way to get around that besides going and integrating them into devtools better...
14:31:26  <bradleymeck>vscode integrates with the chrome devtools protocol but does not provide as full featured workflows as chrome devtools
14:32:06  <bradleymeck>it pops up a docked segment of the editor for the debugger console and uses its own UI for step/continue
14:32:56  <bradleymeck>you can still perform call stack evals to get around some things, but various features are missing
14:35:00  <bradleymeck>though some things are hard to do even with debugger protocols
14:35:22  <bradleymeck>like changing the values of variables inside functions that are not on the stack
15:04:03  * JaseWjoined
15:05:02  * JaseWquit (Quit: Leaving...)
15:05:45  * JaseWjoined
15:24:25  * bradleymeckquit (Quit: bradleymeck)
15:42:14  * bradleymeckjoined
16:38:37  <devsnek>linclark: i think you made an error in your blog post yesterday about cjs
16:38:53  * bradleymeckquit (Quit: bradleymeck)
16:39:03  <devsnek>you said the entire export object is copied?
16:39:51  <linclark>devsnek: I did say the export object is copied, based on something from Axel's modules work
16:39:59  <linclark>devsnek: is that not the case?
16:40:04  * bradleymeckjoined
16:40:11  <devsnek>i don't think i've ever seen a cjs implementation that does that
16:40:23  <devsnek>maybe i'm wrong
16:40:40  <linclark>devsnek: so I know that the use case that I describe works as I say it does because I tested it
16:41:03  <devsnek>ok
16:41:12  <linclark>I didn't step through the code, but I do think I've seen more people than just Axel saying that's how it works
16:41:26  <linclark>let me find where he describes it real quick...
16:42:03  <linclark>devsnek: here's where he describes it: http://2ality.com/2015/07/es6-module-exports.html
16:42:39  <devsnek>hmm ok
16:42:57  <devsnek>i was under the impression that the exports object is shared
16:43:35  <devsnek>actually his example uses node.js hold on
16:44:50  <devsnek>oh wait you mean binding the value
16:45:20  <devsnek>linclark: i see where i tripped up now hehe
16:46:11  <linclark>ah, cool, I'm glad that I didn't misrepresent :)
16:47:01  <devsnek>the wording made me think it was like return deepCopy(exports) for every import
16:47:07  <devsnek>which didn't sound right
16:51:06  <linclark>oh yeah, that would make things slow
17:46:29  * JaseWquit (Quit: Leaving...)
18:38:58  * keith_millerquit (Quit: My MacBook has gone to sleep. ZZZzzz…)
18:41:27  * keith_millerjoined
18:42:04  <ljharb>littledan: yeah that's fair; i didn't mean to imply that it was something that people would automatically know to implement, more that it seems like the sort of thing people will certainly want to do, and browsers, once aware, will be willing to do.
18:43:27  * keith_millerquit (Client Quit)
18:53:53  * bradleymeckquit (Quit: bradleymeck)
18:56:21  * bradleymeckjoined
19:00:53  <devsnek>littledan: i did some investigating, i think i confused bigint's capi having a fast path for creating one from a uint64_t
19:02:42  * keith_millerjoined
20:02:58  * caridy_joined
20:06:31  * caridyquit (Ping timeout: 265 seconds)
21:21:13  * bradleymeckquit (Quit: bradleymeck)
22:27:53  * bradleymeckjoined
22:35:53  * jwaldenjoined
22:47:19  * bradleymeckquit (Quit: )
22:51:22  * bradleymeckjoined
22:57:22  * bradleymeckquit (Quit: bradleymeck)