#racket on 2020-05-25 — irc logs at freenode.irclog.whitequark.org

2020-05-03 17:48 samth changed the topic of #racket to: Racket v7.7 has been released: https://blog.racket-lang.org/2020/05/racket-v7-7.html -- Racket -- https://racket-lang.org -- https://pkgs.racket-lang.org -- Paste at http://pasterack.org

00:04 orivej has quit [Ping timeout: 256 seconds]

00:05 orivej has joined #racket

00:12 corpix has quit [Remote host closed the connection]

00:12 orivej has quit [Ping timeout: 264 seconds]

00:13 orivej has joined #racket

00:13 corpix has joined #racket

00:13 corpix has quit [Remote host closed the connection]

00:16 catonano_ has joined #racket

00:16 tilpner_ has joined #racket

00:19 corpix has joined #racket

00:19 tilpner has quit [Ping timeout: 265 seconds]

00:19 tilpner_ is now known as tilpner

00:19 catonano has quit [Ping timeout: 260 seconds]

00:19 catonano_ is now known as catonano

00:26 orivej has quit [Ping timeout: 272 seconds]

00:26 orivej has joined #racket

00:40 orivej has quit [Quit: No Ping reply in 180 seconds.]

00:41 orivej has joined #racket

00:52 orivej has quit [Quit: No Ping reply in 180 seconds.]

00:52 sinsnare has joined #racket

00:53 orivej has joined #racket

00:53 sinsnare has quit [Remote host closed the connection]

00:59 sleepnap has joined #racket

01:10 KindOne has quit [Ping timeout: 272 seconds]

01:15 lockywolf has joined #racket

01:24 lockywolf_ has joined #racket

01:26 lockywolf has quit [Ping timeout: 260 seconds]

01:29 KindOne has joined #racket

01:32 sleepnap has quit [Quit: Leaving.]

01:41 orivej has quit [Ping timeout: 240 seconds]

01:42 orivej has joined #racket

01:43 wingsorc has joined #racket

02:12 orivej has quit [Ping timeout: 272 seconds]

02:12 orivej_ has joined #racket

02:18 wingsorc has quit [Quit: Leaving]

02:32 orivej_ has quit [Ping timeout: 272 seconds]

02:35 lavaflow has quit [Ping timeout: 256 seconds]

02:36 lavaflow has joined #racket

02:39 orivej has joined #racket

02:40 lavaflow has quit [Ping timeout: 260 seconds]

02:51 YuGiOhJCJ has joined #racket

02:58 orivej has quit [Ping timeout: 240 seconds]

02:58 orivej_ has joined #racket

03:09 FreeFull has quit []

03:15 ArneBab has joined #racket

03:17 KindTwo has joined #racket

03:18 KindOne has quit [Ping timeout: 256 seconds]

03:21 KindTwo is now known as KindOne

03:29 orivej_ has quit [Ping timeout: 272 seconds]

03:29 orivej has joined #racket

03:34 dddddd has quit [Ping timeout: 258 seconds]

03:50 nebunez has quit [Ping timeout: 265 seconds]

03:52 orivej_ has joined #racket

03:54 orivej has quit [Ping timeout: 264 seconds]

03:57 orivej_ has quit [Ping timeout: 240 seconds]

04:10 FreeFull has joined #racket

04:33 lockywolf_ has quit [Remote host closed the connection]

04:34 lockywolf_ has joined #racket

04:35 lockywolf_ has quit [Max SendQ exceeded]

04:36 lockywolf_ has joined #racket

04:37 lockywolf_ has quit [Max SendQ exceeded]

04:38 lockywolf_ has joined #racket

04:39 lockywolf_ has quit [Max SendQ exceeded]

04:39 lockywolf_ has joined #racket

04:43 SGASAU has quit [Remote host closed the connection]

04:48 aidalgol has quit [Read error: Connection reset by peer]

04:59 ArneBab has quit [Read error: Connection reset by peer]

05:07 Lowl3v3l has quit [Quit: Leaving.]

05:29 aidalgol has joined #racket

05:29 SGASAU has joined #racket

05:38 SGASAU has quit [Remote host closed the connection]

05:53 SGASAU has joined #racket

05:57 rgherdt has joined #racket

06:13 SGASAU has quit [Remote host closed the connection]

06:13 SGASAU has joined #racket

06:25 andrei-n has joined #racket

06:37 SGASAU has quit [Remote host closed the connection]

06:38 SGASAU has joined #racket

07:00 nullcone has joined #racket

07:01 SGASAU has quit [Ping timeout: 256 seconds]

07:27 orivej has joined #racket

07:34 Sgeo has quit [Read error: Connection reset by peer]

07:35 lockywolf_ has quit [Remote host closed the connection]

07:36 lockywolf_ has joined #racket

07:39 SGASAU has joined #racket

07:40 FreeFull has quit [Quit: rebooting]

07:40 bkhl_ has joined #racket

07:42 FreeFull has joined #racket

07:47 bkhl_ is now known as bkhl

07:52 orivej_ has joined #racket

07:54 orivej has quit [Ping timeout: 258 seconds]

07:59 srandon111 has quit [Ping timeout: 256 seconds]

08:00 orivej_ has quit [Ping timeout: 264 seconds]

08:01 orivej has joined #racket

08:08 orivej has quit [Quit: No Ping reply in 180 seconds.]

08:09 orivej has joined #racket

08:29 orivej has quit [Ping timeout: 256 seconds]

08:29 orivej_ has joined #racket

08:34 bkhl has left #racket [#racket]

08:36 libertyprime has joined #racket

08:41 orivej_ has quit [Ping timeout: 256 seconds]

08:42 orivej has joined #racket

08:49 orivej has quit [Quit: No Ping reply in 180 seconds.]

08:51 orivej has joined #racket

08:59 orivej has quit [Ping timeout: 240 seconds]

09:00 orivej has joined #racket

09:12 orivej has quit [Ping timeout: 265 seconds]

09:13 orivej has joined #racket

09:25 orivej has quit [Quit: No Ping reply in 180 seconds.]

09:26 orivej has joined #racket

09:26 SGASAU has quit [Remote host closed the connection]

09:26 SGASAU has joined #racket

09:44 orivej has quit [Ping timeout: 256 seconds]

09:44 orivej has joined #racket

09:52 jellie has joined #racket

09:55 sleepnap has joined #racket

09:55 sleepnap has left #racket [#racket]

10:02 orivej has quit [Quit: No Ping reply in 180 seconds.]

10:04 orivej has joined #racket

10:05 <dzoe> Ok, now the futures are slightly faster, but apparently after some continuation frames are GC'd, the futures start to allocate memory in chunks of 16 bytes after each iteration.

10:05 <dzoe> If I could get a rid of this allocation, there would be a huge performance boost.

10:06 <dzoe> Is there a way to confirm what is the source of this allocation?

10:10 true-grue has joined #racket

10:24 urbanslug has joined #racket

10:24 jellie has quit [Ping timeout: 260 seconds]

10:25 real-grue has joined #racket

10:28 true-grue has quit [Ping timeout: 256 seconds]

10:28 <und-l-t-d[m]> (a quick style question) if I want to represent a value that is unset (function argument or struct field that should be ignored if unset) - is #f or (void) better? I think I've seen #f used more often in main distribution...

10:28 lockywolf_ has quit [Ping timeout: 240 seconds]

10:31 iyzsong has joined #racket

10:34 nullcone has quit [Quit: Connection closed for inactivity]

10:39 <und-l-t-d[m]> foing to use #f for now since it is not a valid value for the field anyway and doesn't require anything around it in an if expression

10:39 <und-l-t-d[m]> *going

10:48 <dzoe> That's what most of us do - you can (if value ...) easily.

10:52 orivej has quit [Quit: No Ping reply in 180 seconds.]

10:54 orivej has joined #racket

10:58 catonano has quit [Read error: Connection reset by peer]

11:01 urbanslug is now known as jellie

11:01 catonano has joined #racket

11:02 YuGiOhJCJ has quit [Quit: YuGiOhJCJ]

11:32 jellie has quit [Ping timeout: 265 seconds]

11:33 jellie has joined #racket

11:34 lockywolf_ has joined #racket

11:38 jellie has quit [Ping timeout: 260 seconds]

11:39 orivej has quit [Ping timeout: 256 seconds]

11:39 orivej has joined #racket

11:39 jellie has joined #racket

11:40 dddddd has joined #racket

11:44 jellie has quit [Ping timeout: 260 seconds]

11:48 orivej_ has joined #racket

11:48 orivej has quit [Ping timeout: 260 seconds]

11:51 <und-l-t-d[m]> thanks

11:53 jellie has joined #racket

11:55 davidl_ has quit [Ping timeout: 258 seconds]

11:55 orivej_ has quit [Ping timeout: 240 seconds]

11:56 orivej has joined #racket

11:58 jellie has quit [Ping timeout: 264 seconds]

11:59 jellie has joined #racket

11:59 davidl_ has joined #racket

12:01 tlcu has joined #racket

12:04 jellie has quit [Ping timeout: 256 seconds]

12:05 jellie has joined #racket

12:07 sz0 has quit [Ping timeout: 260 seconds]

12:07 terrorjack has quit [Ping timeout: 260 seconds]

12:08 orivej has quit [Ping timeout: 264 seconds]

12:08 ^[_ has quit [Ping timeout: 260 seconds]

12:08 orivej has joined #racket

12:08 krono has quit [Ping timeout: 260 seconds]

12:08 krono has joined #racket

12:10 sz0 has joined #racket

12:10 jellie has quit [Ping timeout: 260 seconds]

12:11 terrorjack has joined #racket

12:11 jellie has joined #racket

12:12 aidalgol has quit [Read error: Connection reset by peer]

12:13 TCZ has joined #racket

12:16 jellie has quit [Ping timeout: 260 seconds]

12:18 jellie has joined #racket

12:22 jellie has quit [Ping timeout: 256 seconds]

12:24 jellie has joined #racket

12:29 jellie has quit [Ping timeout: 272 seconds]

12:30 jellie has joined #racket

12:34 <samth> dzoe: there's not great memory profiling support, but you could use `dump-memory-stats` periodically and look at what changes

12:35 jellie has quit [Ping timeout: 256 seconds]

12:36 jellie has joined #racket

12:40 <wingo> samth: fwiw something i've found useful in the past in guile at least is a way of driving the sampling profiler not from SIGITIMER but from gc

12:40 <wingo> the sampling rate is much lower tho so you have to run your test for longer

12:40 jellie has quit [Ping timeout: 256 seconds]

12:42 jellie has joined #racket

12:46 lavaflow has joined #racket

12:47 jellie has quit [Ping timeout: 260 seconds]

12:48 orivej has quit [Read error: Connection reset by peer]

12:48 orivej_ has joined #racket

12:48 <samth> wingo: thanks, that's a clever idea

12:57 iyzsong has quit [Quit: ZNC 1.7.1 - https://znc.in]

12:57 orivej_ has quit [Ping timeout: 256 seconds]

12:58 orivej has joined #racket

13:00 SGASAU has quit [Remote host closed the connection]

13:00 SGASAU has joined #racket

13:00 iyzsong has joined #racket

13:10 libertyprime has quit [Remote host closed the connection]

13:13 TCZ has quit [Quit: Leaving]

13:15 jellie has joined #racket

13:18 orivej has quit [Ping timeout: 272 seconds]

13:18 orivej_ has joined #racket

13:31 orivej_ has quit [Ping timeout: 265 seconds]

13:31 orivej has joined #racket

13:35 srandon111 has joined #racket

13:39 orivej has quit [Ping timeout: 246 seconds]

13:39 orivej has joined #racket

14:01 <dzoe> samth: I suspect that actually some call frames need to be allocated and I have no idea how to confirm that.

14:01 <dzoe> The only hint is that all allocations are 16 bytes.

14:02 <dzoe> And I do remember that the function call tree should not be very deep in futures.

14:02 <dzoe> I'll look into inlining one variant and check the difference.

14:03 <dzoe> Although more information from future-event 'sync with "allocate memory" would be really beneficial :-/

14:03 <samth> dzoe: that `'sync` should only happen for a whole new page

14:04 <samth> so unfortunately I don't think there's much more info at that point

14:12 tlcu has quit [Ping timeout: 258 seconds]

14:14 tlcu has joined #racket

14:15 <dzoe> Interesting.

14:15 <dzoe> Maybe I'll add some debug info to custom racket build.

14:15 <dzoe> Btw, CS has absolutely no syncs there.

14:16 <dzoe> Futures in CS work like a charm.

14:16 <dzoe> But there are boxed flonums :)

14:16 <samth> i expect mflatt will fix that relatively soon

14:17 <samth> more significant imo is that racket doesn't have a good way to work with 64-bit integers efficiently

14:17 <samth> although i don't know if you run into that

14:18 jellie has quit [Quit: WeeChat 2.8]

14:21 samlamamma has joined #racket

14:21 <samlamamma> Does Racket have any built-in procedures for tree traversal? I'm hopelessly in need of SUBST-IF from CL :-/

14:28 <samth> samlamamma: I don't know of anything like htat

14:31 <samlamamma> samth:That's too bad! I wrote it myself in the end, I should've done that from the beginning he-he :-)

14:39 <dzoe> samth: and what are fixnums all about?

14:40 <dzoe> I know it is roughly 62bits or whatever, but that is enough :)

14:41 <samth> fixnums (being shorter by a few bits) are not enough for some uses

14:41 <dzoe> You are not entirely right here :)

14:41 <dzoe> I am using full 32bit numbers on 32bit Racket builds without any performance impact :)

14:42 <dzoe> With 32bit = #x0 .. #xffffffff

14:42 <dzoe> I assume the same trick can be used with 64bit numbers.

14:44 iyzsong has quit [Quit: ZNC 1.7.1 - https://znc.in]

14:45 <samlamamma> dzoe:How do you do that? Doesn't Racket do bit tagging?

14:54 orivej has quit [Quit: No Ping reply in 180 seconds.]

14:55 orivej has joined #racket

14:59 <dzoe> samlamamma: racket/unsafe/ops and you construct the numbers, it is easy to write "conversion" functions for that

15:00 juanfra has quit [Ping timeout: 256 seconds]

15:01 <samth> dzoe: i'm interested to see that

15:01 <dzoe> samth: just do not write #xff000000 but rather (unsafe-fxior (unsafe-fxlshift #xff 24) #x000000) for example :)

15:02 <dzoe> But of course, you must not EVER touch that value using anything else than unsafe fixnum operations.

15:02 <dzoe> A simple display or format can bring the whole program down, of course ...

15:02 endformationage has joined #racket

15:03 <samlamamma> Could you `deref` that number by putting it into function position?

15:03 <dzoe> For some numbers I'd be afraid, yes ...

15:03 <dzoe> Not a good idea.

15:03 <samlamamma> He-he :-)

15:03 <samlamamma> https://i.kym-cdn.com/entries/icons/facebook/000/025/817/Screen_Shot_2018-03-30_at_11.34.27_AM.jpg is this you while coding? dzoe

15:04 <dzoe> This is by far the least unsafe thing I do with Racket ...

15:11 Lowl3v3l has joined #racket

15:16 orivej has quit [Ping timeout: 258 seconds]

15:16 orivej has joined #racket

15:21 samlamamma has quit [Ping timeout: 260 seconds]

15:23 juanfra has joined #racket

15:25 samlamamma has joined #racket

15:27 orivej has quit [Ping timeout: 240 seconds]

15:27 orivej has joined #racket

15:27 SGASAU has quit [Remote host closed the connection]

15:28 SGASAU has joined #racket

15:34 nebunez has joined #racket

15:38 Sgeo has joined #racket

15:56 orivej has quit [Ping timeout: 256 seconds]

15:56 orivej has joined #racket

16:03 efm has quit [Ping timeout: 264 seconds]

16:18 efm has joined #racket

16:24 orivej has quit [Ping timeout: 256 seconds]

16:25 orivej has joined #racket

16:46 orivej has quit [Ping timeout: 256 seconds]

16:46 orivej_ has joined #racket

16:56 orivej_ has quit [Ping timeout: 258 seconds]

16:57 orivej has joined #racket

16:58 juanfra has quit [Ping timeout: 256 seconds]

16:59 juanfra has joined #racket

17:05 orivej_ has joined #racket

17:05 orivej has quit [Ping timeout: 272 seconds]

17:12 epony has quit [Quit: upgrades]

17:22 epony has joined #racket

17:22 sleepnap has joined #racket

17:25 orivej has joined #racket

17:25 orivej_ has quit [Ping timeout: 256 seconds]

17:33 efm has quit [Ping timeout: 265 seconds]

17:40 orivej has quit [Ping timeout: 256 seconds]

17:41 orivej has joined #racket

17:42 ArthurStrong has quit [Ping timeout: 265 seconds]

18:01 orivej has quit [Ping timeout: 246 seconds]

18:02 orivej has joined #racket

18:32 orivej has quit [Ping timeout: 256 seconds]

18:32 orivej_ has joined #racket

18:38 KindTwo has joined #racket

18:39 KindOne has quit [Ping timeout: 272 seconds]

18:42 KindTwo is now known as KindOne

19:17 orivej has joined #racket

19:17 orivej_ has quit [Ping timeout: 264 seconds]

19:22 SGASAU has quit [Remote host closed the connection]

19:22 SGASAU has joined #racket

19:25 stagboot has joined #racket

19:28 <stagboot> Seems I misunderstand require/typed's #:opaque argument. I'd think that (require/typed net/http-client [#:opaque SSL base-ssl?-tnl/c]) (define x : SSL #f) would be OK, but it retorts, "expected SSL. given False"?

19:29 <samth> stagboot: if you use `#:opaque` the only way to produce such a value is to import it with `require/typed` or to check for it with `base-ssl?-tnl/c`

19:34 <stagboot> samth: thanks. I'm producing a new value, so I can't import it. How can I "check for it?"

19:35 <samth> in that case, opaque probably isn't the right thing. but how are you creating it?

19:36 <samth> looking at the docs, I recommend translating the definitions of `base-ssl?` and `base-ssl?-tnl/c` directly

19:37 <samth> ie, (define-type Base-SSL (U Boolean Symbol SSL-Client-Context))

19:37 <samth> (define-type Base-SSL-TNL (U Base-SSL (List Base-SSL Input-Port Output-Port (-> Port Void)))

19:37 <samth> and use #:opaque for SSL-Client-Context

19:39 <stagboot> isn't that hacky, though? what if the type were more complex, and made of many non-standard types?

19:44 <stagboot> oh! wait, i think i see: i was trying to use #:opaque to derive a type from a contract, but it expects a boolean.

19:46 <stagboot> which would mean that, unless there's a function to derive types from contracts, i would need to define the types as you've outlined.

19:57 even4void has joined #racket

19:58 <samth> the problem is not that it got a boolean, but that typed racket doesn't know what's in `base-ssl?-tnl/c`.

20:00 aidalgol has joined #racket

20:00 even4void has quit [Remote host closed the connection]

20:04 orivej_ has joined #racket

20:05 orivej has quit [Read error: Connection reset by peer]

20:08 samlamamma has quit [Ping timeout: 260 seconds]

20:23 even4void has joined #racket

20:26 even4void_ has joined #racket

20:27 even4void_ has quit [Client Quit]

20:33 even4void has quit [Remote host closed the connection]

20:33 orivej_ has quit [Ping timeout: 246 seconds]

20:33 orivej has joined #racket

20:43 jmiven has quit [Quit: bye]

20:44 jmiven has joined #racket

20:56 orivej has quit [Ping timeout: 256 seconds]

20:56 ArneBab has joined #racket

20:57 orivej has joined #racket

21:05 stagboot has quit [Quit: Leaving]

21:08 tlcu has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]

21:12 <dzoe> Sometimes I think I should have never started this project.

21:12 <dzoe> Apparently there is a way to remove all allocations and as a nice side effect reduce GC frequency.

21:13 <dzoe> The downside is I need to maintain a pool of special structures for passing arguments to "procedures" the futures "call" and just reference these in the "futures-friendly" queue.

21:13 <dzoe> Now I am really curious what other bugs are lurking in there ...

21:18 TCZ has joined #racket

21:28 nebunez has quit [Ping timeout: 256 seconds]

21:35 zmt00 has quit [Quit: Leaving]

21:37 zmt00 has joined #racket

21:56 orivej has quit [Ping timeout: 256 seconds]

21:56 orivej has joined #racket

22:00 andrei-n has quit [Quit: Leaving]

22:05 corpix_ has joined #racket

22:05 corpix has quit [Ping timeout: 240 seconds]

22:05 ermo has quit [Ping timeout: 256 seconds]

22:07 ermo has joined #racket

22:09 orivej has quit [Ping timeout: 256 seconds]

22:09 orivej_ has joined #racket

22:17 jellie has joined #racket

22:23 ArthurStrong has joined #racket

22:36 orivej_ has quit [Ping timeout: 264 seconds]

22:37 orivej has joined #racket

22:46 TCZ has quit [Quit: Leaving]

22:57 orivej has quit [Quit: No Ping reply in 180 seconds.]

22:58 orivej has joined #racket

23:02 libertyprime has joined #racket

23:03 ArthurStrong has quit [Read error: Connection reset by peer]

23:04 real-grue has quit [Read error: Connection reset by peer]

23:06 orivej has quit [Ping timeout: 264 seconds]

23:06 orivej has joined #racket

23:09 ArthurStrong has joined #racket

23:14 orivej has quit [Read error: Connection reset by peer]

23:15 orivej has joined #racket

23:31 orivej has quit [Ping timeout: 256 seconds]

23:31 orivej has joined #racket

23:40 orivej has quit [Ping timeout: 260 seconds]

23:40 orivej has joined #racket

23:45 rgherdt has quit [Ping timeout: 260 seconds]

23:51 orivej has quit [Quit: No Ping reply in 180 seconds.]

23:52 orivej has joined #racket

23:54 dmiles has quit [Read error: Connection reset by peer]

23:58 dmiles has joined #racket