Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6-dev, 2017-11-05

| Channels | #perl6-dev index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:45 evalable6 joined #perl6-dev
00:57 japhb .ask [Tux] Where is the historical file for your newly-expanded suite of CSV perf tests?  I'm thinking of expanding the buggable speed grapher to graph any of them.
00:57 yoleaux japhb: I'll pass your message to [Tux].
00:59 Zoffix Tux said "not yet"
01:00 Zoffix https://irclog.perlgeek.de/perl6-dev/2017-10-29#i_15371662
01:01 japhb Zoffix: Hmmm, that was almost a week ago.  Seems ripe for a re-ask.
01:36 timotimo greetings japhb
01:36 timotimo how are you today?
01:56 TimToady joined #perl6-dev
02:22 evalable6 joined #perl6-dev
02:56 ilbot3 joined #perl6-dev
02:56 Topic for #perl6-dev is now Perl 6 language and compiler development 2.0 | Logs at http://irclog.perlgeek.de/perl6-dev/today | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
03:18 BenGoldberg joined #perl6-dev
03:42 hoelzro joined #perl6-dev
07:17 yoleaux AlexDaniel`: RT/GH weekly
08:33 robertle joined #perl6-dev
09:18 fkpenta joined #perl6-dev
09:34 [Tux] Rakudo version 2017.10-125-g360eb2289 - MoarVM version 2017.10-58-gad8618468
09:34 [Tux] csv-test-xs-20      0.454 -  0.455
09:34 [Tux] csv-ip5xs           1.110 -  1.122
09:34 [Tux] test-t              3.302 -  3.315
09:34 [Tux] test               12.052 - 12.061
09:34 [Tux] csv-parser         12.201 - 12.736
09:34 yoleaux 00:57Z <japhb> [Tux]: Where is the historical file for your newly-expanded suite of CSV perf tests?  I'm thinking of expanding the buggable speed grapher to graph any of them.
09:35 [Tux] csv-ip5xs-20       13.069 - 13.245
09:35 [Tux] test-t-20 --race   21.415 - 22.174
09:35 [Tux] test-t-20          60.028 - 60.618
09:35 [Tux] https://tux.nl/Talks/CSV6/speed-all.log
09:41 Geth ¦ roast: fc36d3dde2 | (Stefan Seifert)++ | 2 files
09:41 Geth ¦ roast: Make spurious timeouts in test runs less likely
09:41 Geth ¦ roast: review: https://github.com/perl6/roast/commit/fc36d3dde2
10:07 evalable6 joined #perl6-dev
10:27 gfldex lizmat: is there a way to stop snapper from snapping?
10:28 evalable6 joined #perl6-dev
10:35 lizmat gfldex: I have been thinking about that
10:35 lizmat snapper(0) would perhaps be logical in a way ?
10:35 lizmat or snapper(Inf) probably better
10:36 gfldex snapper(:stop);
10:36 lizmat that too  :-)
10:37 gfldex m: sleep ∞;
10:37 camelia rakudo-moar 360eb2289: OUTPUT: «(timeout)»
10:39 lizmat gfldex: should the period between a stop / start also be considered a period or not ?
10:40 gfldex lizmat: I don't think so. If I want to stop something, I likely don't want the result of the operation anyway.
10:41 lizmat so you would run a report after stopping and before starting again ?
10:42 lizmat or just discard ?
10:42 lizmat hmmm
10:42 gfldex no report
10:42 gfldex one can ask for the report by hand
10:44 lizmat yeah, I wasn't going to auto-generate that for stopping the snapper
10:49 El_Che who telemeters the telemetry?
10:51 Geth ¦ rakudo: 9344d35de8 | (Elizabeth Mattijsen)++ | lib/Telemetry.pm6
10:51 Geth ¦ rakudo: Add :stop / :reset parameters to snapper()
10:51 Geth ¦ rakudo:
10:51 Geth ¦ rakudo: - :reset allows you to forget about the snaps made so far
10:51 Geth ¦ rakudo: - :stop allows you to stop the snapper
10:51 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/9344d35de8
10:51 lizmat El_Che: the snapper of snappers
10:52 lizmat gfldex: mind you, it won't check if it's stopped until after the sleep comes out
10:52 lizmat I don't really want to use Promises in there because of the overhead involved
10:53 gfldex lizmat: I agree. The last snapping wont hurt.
10:54 El_Che snap is a popular term it seems. Yesterday I had a look at Ubuntu's snap to see if it was feasable to provide rakudo snaps. And I reorganized my backups moving to ZFS snapshots and the sync'er (remote of other drive) for the snapshots is calles simplesnap and the autosnapshortter zfSnap.
10:54 El_Che this must be what facebook must feel like
10:55 El_Che or amazon. "Hey you bough a vacuum cleaner. Here have a look at a zillion other vacuuum cleaners"
10:55 El_Che :)
11:02 Geth ¦ rakudo: d5fc6cbb9e | (Elizabeth Mattijsen)++ | lib/Telemetry.pm6
11:02 Geth ¦ rakudo: Don't start a snapper if not running and :stop
11:02 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/d5fc6cbb9e
11:23 gfldex lizmat: https://github.com/gfldex/perl6-telemetry-httpd
11:24 lizmat gfldex: cool!!
11:26 lizmat gfldex: I was also thinking about adding a :csv option
11:26 gfldex that would be handy
11:27 gfldex would make it easy to output proper html
11:28 gfldex lizmat: the module is so tiny, I wonder if it could be shipped with Rakudo.
11:29 lizmat something to consider, afaiac
11:30 lizmat so, :csv would just output the values, whitespace separated ?
11:30 lizmat no header, footer, legend
11:30 lizmat ?
11:30 gfldex :csv should have a header
11:31 lizmat the names of the columns ?
11:31 gfldex yes
11:31 lizmat I mean more the stuff about "Telemetry Report of Process #44910 (2017-11-05T11:23:04Z)",
11:31 lizmat etc.
11:31 lizmat the pre-amble if you like
11:31 gfldex I would not add the pre-amble in csv
11:33 lizmat ack
11:35 El_Che gfldex++
11:35 El_Che and lizmat++
11:36 lizmat gfldex: also no total, right ?
11:39 stmuk joined #perl6-dev
11:55 Geth ¦ rakudo: 3e4ef2e0fc | (Elizabeth Mattijsen)++ | lib/Telemetry.pm6
11:55 Geth ¦ rakudo: Add :csv option to report
11:55 Geth ¦ rakudo:
11:55 Geth ¦ rakudo: - which defaults to RAKUDO_REPORT_CSV setting
11:55 Geth ¦ rakudo: - also made report() itself hopefully a bit less racy
11:55 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/3e4ef2e0fc
11:56 lizmat gfldex: I chose to use whitespace as a delimiter, because it works better with .words
11:58 lizmat (which is about 2.5x as fast as .split(',')
11:58 lizmat )
12:33 lizmat gfldex: feels to me that https://github.com/gfldex/perl6-telemetry-httpd/blob/master/lib/Telemetry/httpd.pm6 could be a "server" sub in Telemetry
12:34 lizmat and then have a helper module "snapserver" (name to be bikedhedded) that would just do: 'snapper(1); httpd' in its mainline
12:50 evalable6 joined #perl6-dev
13:05 BenGoldberg joined #perl6-dev
13:16 gfldex lizmat: I would be careful with exporting generic names in a telemetry module.
13:19 lizmat So always force something like "use Telemetry <snap>" ?
13:20 lizmat note that imports are lexical in Perl 6
13:20 lizmat m: { use Telemetry; snap }; snap
13:20 camelia rakudo-moar 3e4ef2e0f: OUTPUT: «5===SORRY!5=== Error while compiling <tmp>␤Undeclared routine:␤    snap used at line 1␤␤»
13:20 lizmat m: { use Telemetry; snap }
13:20 camelia rakudo-moar 3e4ef2e0f: OUTPUT: «wallclock util% max-rss gw gtc tw ttc aw atc␤202 25.247525 0 0 0 0 0 0 0␤»
13:21 Geth ¦ rakudo: 4d21ad67da | (Elizabeth Mattijsen)++ | lib/Telemetry.pm6
13:21 Geth ¦ rakudo: Oops, the default for :csv should be off!
13:21 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/4d21ad67da
13:24 Geth ¦ rakudo: 0f2f0cd3ab | (Elizabeth Mattijsen)++ | lib/Telemetry.pm6
13:24 Geth ¦ rakudo: Introduce Telemetry 'safe-ctrl-c'
13:24 Geth ¦ rakudo:
13:24 Geth ¦ rakudo: - when called, installs a signal handler for CTRL-c that just exits
13:24 Geth ¦ rakudo: - makes sure that if we're snapping, we're going to see a report even
13:24 Geth ¦ rakudo:   if we're control-c'ing out of the program, e.g. in case of a hang
13:24 Geth ¦ rakudo:   or an infini-loop.
13:24 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/0f2f0cd3ab
13:27 Geth ¦ rakudo: 824a5dc298 | (Elizabeth Mattijsen)++ | lib/snapper.pm6
13:27 Geth ¦ rakudo: Make -Msnapper set the safe-ctrl-c feature
13:27 Geth ¦ rakudo:
13:27 Geth ¦ rakudo: - so even perl6 -Msnapper -e 'loop {}' will give you a report
13:27 Geth ¦ rakudo:   - after you ctrl-c it of course
13:27 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/824a5dc298
13:28 Geth ¦ roast: 2437d1f795 | (Zoffix Znet)++ (committed using GitHub Web editor) | packages/Test/Util.pm
13:28 Geth ¦ roast: Fix inconsistent doesn't-hang defaults
13:28 Geth ¦ roast: review: https://github.com/perl6/roast/commit/2437d1f795
13:30 Geth ¦ roast: 6aa1ce3b5c | (Zoffix Znet)++ (committed using GitHub Web editor) | S28-named-variables/init-instant.t
13:30 Geth ¦ roast: Respect ROAST_TIMING_SCALE in time-sensitive test
13:30 Geth ¦ roast: review: https://github.com/perl6/roast/commit/6aa1ce3b5c
13:30 Geth ¦ roast: 694b2e7abf | (Zoffix Znet)++ (committed using GitHub Web editor) | S28-named-variables/init-instant.t
13:30 Geth ¦ roast: Fix inaccurate test description
13:30 Geth ¦ roast: review: https://github.com/perl6/roast/commit/694b2e7abf
13:31 Geth ¦ roast: 008843bc96 | (Zoffix Znet)++ (committed using GitHub Web editor) | S28-named-variables/init-instant.t
13:31 Geth ¦ roast: Fix typo
13:31 Geth ¦ roast: review: https://github.com/perl6/roast/commit/008843bc96
13:35 buggable joined #perl6-dev
13:36 buggable joined #perl6-dev
13:37 Zoffix buggable: speed :4
13:37 buggable Zoffix,    ▆     ▅     ▆  ▂█   dates: 2017-10-26–2017-11-04
13:37 buggable Zoffix,   ▁█ ▄   █   ▂▃█  ██ ▂ range: 19.341s–21.820s
13:37 buggable Zoffix,   ██▆█ ▁ █▇▇▇███ ▁██▅█ speed: 3% slower (widths: 10/3)
13:37 buggable Zoffix, ▁▃████▅█████████▇█████
13:37 Zoffix (uses test-t-20 --race now)
13:51 Geth ¦ roast: 3ad163d3a1 | usev6++ | 2 files
13:51 Geth ¦ roast: [jvm] Skip two tests that hang during spectest
13:51 Geth ¦ roast: review: https://github.com/perl6/roast/commit/3ad163d3a1
14:33 Ben_Goldberg joined #perl6-dev
14:45 ugexe I have to admit that having a http server in the core, lib/ or not, seems out of place
14:47 Zoffix client, sure, but server? :/
14:49 * Zoffix came up with another marketing blurb: "It's not magic. It's Perl 6."    Both shows the language as awesome and powerful, while distancing from bad rep of "too magical" Perl sometimes gets
14:51 AlexDaniel Zoffix++
14:57 BenGoldberg joined #perl6-dev
15:14 Geth ¦ nqp: 17ce8f83d7 | usev6++ | 3 files
15:14 Geth ¦ nqp: [jvm] Implement 'getport' for ServerSocketHandle
15:14 Geth ¦ nqp: review: https://github.com/perl6/nqp/commit/17ce8f83d7
15:16 AlexDaniel buggable: tag MOLD
15:16 buggable AlexDaniel, There are 233 tickets tagged with MOLD; See http://fail.rakudo.party/t/MOLD for details
15:18 jnthn What's MOLD?
15:20 committable6 joined #perl6-dev
15:21 Zoffix jnthn: tickets that haven't seen any updates in last 2 years
15:22 AlexDaniel (the tag is added automatically)
15:36 Zoffix Fixing breakage due to readonly $/ being touched by .subst... And really feels like a bad direction. This is one set of changes that needs to be done to a method the second a .subst or .match call is added: https://gist.github.com/zoffixznet/c68579b7cad03b286d735dbe91e529f1
15:36 Zoffix IMO we should not set $/ with method matching forms
15:38 Zoffix Reverting https://github.com/rakudo/rakudo/commit/738908be4d79f1472821cd363ec82a1ef0bc3f32
15:39 Zoffix I think we need to go into the opposite direction from that. Eliminate all the caller's-$/ fondling. Prolly make stuff faster too
15:39 llfourn joined #perl6-dev
15:44 Geth ¦ roast: b0b3844c7d | (Zoffix Znet)++ | S32-str/substr.t
15:44 Geth ¦ roast: Revert "Test .subst complains when it can't use $/"
15:44 Geth ¦ roast:
15:44 Geth ¦ roast: This reverts commit 80a49c4324bc4a166c311e3eb01b7bae8814893c.
15:44 Geth ¦ roast:
15:44 Geth ¦ roast: Such behaviour has too much unwanted impact in grammar actions
15:44 Geth ¦ roast: where $/ is the common name for the parameter. Having to change
15:44 Geth ¦ roast: $/ to something else the second you introduce a .subst call
15:44 Geth ¦ roast: (or other methods that set $/) is a bit too much.
15:44 Geth ¦ roast:
15:44 Geth ¦ roast: https://irclog.perlgeek.de/perl6-dev/2017-11-05#i_15404670
15:44 Geth ¦ roast: review: https://github.com/perl6/roast/commit/b0b3844c7d
15:47 AlexDaniel “You can only reassign tickets that you own or that are unowned”… Really?
15:51 Geth ¦ roast: 76e1f94768 | (Zoffix Znet)++ | S05-grammar/action-stubs.t
15:51 Geth ¦ roast: Remove new unwanted test
15:51 Geth ¦ roast:
15:51 Geth ¦ roast: Part of revert for https://github.com/perl6/roast/commit/b0b3844c7d
15:51 Geth ¦ roast: review: https://github.com/perl6/roast/commit/76e1f94768
15:51 Geth ¦ rakudo: 1c387153ea | (Zoffix Znet)++ | src/core/Str.pm
15:51 Geth ¦ rakudo: Revert ""Do. Or do not. There is no try.""
15:51 Geth ¦ rakudo:
15:51 Geth ¦ rakudo: This reverts commit 738908be4d79f1472821cd363ec82a1ef0bc3f32.
15:51 Geth ¦ rakudo:
15:51 Geth ¦ rakudo: Such behaviour has too much unwanted impact in grammar actions
15:51 Geth ¦ rakudo: where $/ is the common name for the parameter. Having to change
15:51 Geth ¦ rakudo: $/ to something else the second you introduce a .subst call
15:51 Geth ¦ rakudo: (or other methods that set $/) is a bit too much.
15:51 Geth ¦ rakudo:
15:52 Geth ¦ rakudo: https://irclog.perlgeek.de/perl6-dev/2017-11-05#i_15404670
15:52 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/1c387153ea
15:56 AlexDaniel buggable: NYI
15:56 AlexDaniel buggable: tag NYI
15:56 buggable AlexDaniel, There are 58 tickets tagged with NYI; See http://fail.rakudo.party/t/NYI for details
16:00 AlexDaniel I'm not happy with merging tickets on RT because it creates a mess, but closing duplicates and simply linking them back and forth kinda sucks too
16:01 AlexDaniel (because creates an impression that the issue was fixed in one of the tickets and therefore all others should be closed)
16:11 nine AlexDaniel: what's so bad about merging?
16:11 Geth ¦ roast: 5360b920d0 | (Zoffix Znet)++ | packages/Test/Util.pm
16:11 Geth ¦ roast: Improve doesn't-hang() failure mode
16:11 Geth ¦ roast:
16:11 Geth ¦ roast: Add diagnostic message telling the user to try increasing timing
16:11 Geth ¦ roast: scale when test detects a hang.
16:11 Geth ¦ roast: review: https://github.com/perl6/roast/commit/5360b920d0
16:13 Zoffix nine: IME: because RT interleaves one ticket into another instead of appending it so it quickly becomes a mess of incoherent replies.
16:22 AlexDaniel yeah
16:30 Geth ¦ 6.d-prep: 9d645a8025 | (Zoffix Znet)++ (committed using GitHub Web editor) | TODO/FEATURES.md
16:30 Geth ¦ 6.d-prep: Add R#1235 to [potential] TODO list
16:30 Geth ¦ 6.d-prep: review: https://github.com/perl6/6.d-prep/commit/9d645a8025
16:30 synopsebot R#1235 [open]: https://github.com/rakudo/rakudo/issues/1235 [6.d][RFC] Caller-$/-setting is too invasive
16:31 Zoffix AlexDaniel: you should've made your last issue about testing :P
16:31 Zoffix And then say. " What about testing? R#1234 " and people would not know whether that's an actual Issue number :P
16:31 synopsebot R#1234 [open]: https://github.com/rakudo/rakudo/issues/1234 [RFC] Whitespacey things in «» (say «a\nb».perl)
16:33 AlexDaniel we can do it in R#12345 :)
16:35 AlexDaniel that should take just a few years I think
16:39 Zoffix m: say (12345-1234)/50/12
16:39 camelia rakudo-moar 1c387153e: OUTPUT: «18.518333␤»
16:39 Zoffix like 18 years :)
16:54 [Tux] japhb, is https://tux.nl/Talks/CSV6/speed-all.log filling your needs?
16:55 Zoffix buggable: speed
16:55 buggable Zoffix, ▁▁▅█▄▅▂▃▃▇▄▄▄▅▅█▃▃▆█▄▅ dates: 2017-10-26–2017-11-04; range: 19.341s–21.820s; speed: 3% slower (widths: 10/3)
16:55 lizmat .
16:55 yoleaux 14:42Z <AlexDaniel> lizmat: weekly report https://gist.github.com/AlexDaniel/5abca8af46308661c3d30cdad53d17bf
16:55 Zoffix ^ buggable already uses it. [Tux]++
16:55 AlexDaniel Zoffix: does not look right. It feels like you forgot to count pull requests
16:56 AlexDaniel (and that we are still fixing RT bugs, but that will be happening less and less I think)
16:56 Zoffix AlexDaniel: IIRC we get like 30 tickets a month + I figure about 20 PRs
16:57 Zoffix releasable6: status
16:57 releasable6 Zoffix, Next release in 13 days and ≈2 hours. No blockers. 0 out of 132 commits logged
16:57 releasable6 Zoffix, Details: https://gist.github.com/52dacd51054c6a8c3f8faf53df5bb688
16:57 Zoffix NeuralAnomaly: status
16:57 NeuralAnomaly Zoffix, Use releasable6 bot instead. e.g. try: releasable6: status
17:00 AlexDaniel Zoffix: according to the gist above we resolved 17 tickets in the last week (some of them were created and resolved immediately), while the total number of tickets is around the same (1720→1718)
17:00 AlexDaniel assuming the same pace that's around 68 tickets per month not including pull requests
17:02 AlexDaniel hm, maybe “New tickets (7)” section should include resolved tickets
17:18 AlexDaniel done. Same link: https://gist.github.com/AlexDaniel/5abca8af46308661c3d30cdad53d17bf
17:22 Geth ¦ rakudo: 22939bc819 | (Elizabeth Mattijsen)++ | lib/Telemetry.pm6
17:22 Geth ¦ rakudo: Telemetry report now removes columns without any data
17:22 Geth ¦ rakudo:
17:22 Geth ¦ rakudo: - only for the human readable case, though: :csv output unchanged
17:22 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/22939bc819
17:32 stmuk_ joined #perl6-dev
17:44 jnthn (from backlog) fwiw, I agree with ugexe and Zoffix - the point Telemetry involves a HTTP server and stuff, that feels a bit beyond the scope of things I'd like in the Rakudo repo. In that case, it's better it lives on the ecosystem and can depend on one of the many options for building a HTTP server in there.
17:45 jnthn Or the HTTP aspects of it could live there, at least.
17:46 Geth ¦ rakudo: 77142fdb29 | (Elizabeth Mattijsen)++ | lib/Telemetry.pm6
17:46 Geth ¦ rakudo: Fix some Telemetry / snapper race condition
17:46 Geth ¦ rakudo:
17:46 Geth ¦ rakudo: Sometimes, if you would use snapper and say your own report at the
17:46 Geth ¦ rakudo: end, another (almost) empty report would be shown as well.
17:46 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/77142fdb29
17:47 * lizmat notes jnthn ugexe Zoffix' comments
17:48 jnthn I guess sine there's bits of Telemetry that depend on nqp::getrusage and maybe poke into scheduler internals, we'd rather keep those in the Rakudo repo
17:48 jnthn As we do with NativeCall
17:50 lizmat jnthn: yes, that was my intent
17:51 lizmat but also open it up for more general usage / mixes
17:54 jnthn *nod*
18:38 lizmat hmmm.. I'm considering dropping support for "use Telemetry :COLUMNS"
18:38 lizmat if we're going to generalize Telemetry::Instruments, then exporting stuff becomes very confusing
19:32 evalable6 joined #perl6-dev
19:49 BenGoldberg joined #perl6-dev
19:57 llfourn_ joined #perl6-dev
19:57 * Zoffix bumps brochure on size up, from 16 to 20 pages
19:57 El_Che_ joined #perl6-dev
19:58 Zoffix This section is all that's left to do: https://i.imgur.com/HSyod2Y.png and it could be done on 1 page, but I rather add more @marketing-things and code examples.
20:01 Zoffix That's without book excerpts btw
20:02 hoelzro_ joined #perl6-dev
20:02 [Coke]_ joined #perl6-dev
20:03 lizmat Zoffix++
20:04 lizmat the only thing that slightly bothers me is the lack of periods at the end of a sentence.
20:08 Zoffix lizmat: hah. I always remove periods at the end of bullets in bulleted lists, since bullets already signify breaks, but I'll add them back into this brochure :)
20:08 lizmat well, that's why I said slighly
20:08 lizmat *tly
20:11 evalable6_ joined #perl6-dev
20:11 nativecallable6_ joined #perl6-dev
20:11 bloatable6_ joined #perl6-dev
20:11 committable6_ joined #perl6-dev
20:31 benchable6 joined #perl6-dev
20:31 bisectable6 joined #perl6-dev
20:31 quotable6 joined #perl6-dev
20:31 coverable6 joined #perl6-dev
20:31 releasable6 joined #perl6-dev
20:31 unicodable6 joined #perl6-dev
20:31 statisfiable6 joined #perl6-dev
20:47 AlexDaniel Zoffix: there's a guy on #perl6 struggling with WWW module. I'm suspecting RT#131553 RT#123282 but I really don't know
20:47 synopsebot RT#131553 [new]: https://rt.perl.org/Ticket/Display.html?id=131553 IO::Socket::INET mishandles IPv6
20:47 AlexDaniel maybe you have some thoughts
20:47 synopsebot RT#123282 [new]: https://rt.perl.org/Ticket/Display.html?id=123282 [IO] IO::Socket::INET fails to connect if a resource is available over IPv6 and the client is misconfigured to use IPv6
20:50 El_Che_ Zoffix: it's a sentence, not an enumeration of words. So thx for the periods :)
20:51 El_Che_ Zoffix: I don't think that bullet point is a good match with that much text? Maybe paragraphs and bold keywords?
20:51 El_Che_ maybe not, just thinking out loud
20:53 El_Che Zoffix: the community pic you posted on twitter, are you using that? It looks like mostly p5 people
21:45 Zoffix Yes. I'm using that.
21:50 [Tux] joined #perl6-dev
21:50 SmokeMachine joined #perl6-dev
21:50 llfourn joined #perl6-dev
21:50 ugexe joined #perl6-dev
21:53 benchable6 joined #perl6-dev
21:53 statisfiable6 joined #perl6-dev
21:53 unicodable6 joined #perl6-dev
21:53 bisectable6 joined #perl6-dev
21:53 releasable6 joined #perl6-dev
22:11 evalable6 joined #perl6-dev
22:11 nativecallable6 joined #perl6-dev
22:11 bloatable6 joined #perl6-dev
22:13 japhb Zoffix: Is your change to buggable to use the -20 race version pushed to GH?
22:14 Zoffix Yes
22:15 japhb Dammit, I was looking at my fork when I thought I was looking at your copy
22:27 tadzik joined #perl6-dev
22:43 japhb Zoffix: In https://github.com/japhb/perl6-buggable/blob/master/lib/Buggable/Plugin/Speed.pm6#L75-L79 Why are you averaging the 3 or 10 entries, rather than taking the minimum within those zones?  The average can be messed up by a fluke, but min should be more stable.
22:44 japhb Zoffix, in fact I might go with default comparison of minimum of first 10 values versus single most current value (== minimum of most recent 1).
22:46 Zoffix japhb: why minimum? The "fluke" you mention would mess it up even more
22:47 Zoffix That'd just undo why the averaging was done in the first place.
22:48 Zoffix If you take a minimum of 10 values and one of them is 20% faster due to a bug, you now have all 10 of those values poisoned until the bug clears out.
22:50 ilmari[m] joined #perl6-dev
22:50 AlexDaniel` joined #perl6-dev
22:50 CIAvash[m] joined #perl6-dev
22:51 japhb Zoffix: I would think it far more common that background processing on [Tux]'s machine would a significant effect on the mean than that we had a bug that resulted in super-fast timings.
22:52 japhb Because timing is a left-truncated distribution, where outliers can *only* exist because of bugs, the far more likely outliers are towards higher values.
22:52 japhb Nevertheless, we could always just change it to "minimum of first 3" or so.
22:54 japhb (Oh, and FWIW, s/stable/robust/.  Sigh, tired brane.)
22:58 Zoffix Makes no sense to me. Why is the minimum value preferred? If we make nine 10s measurements and one 5s measurement, you say there be background processes, so you take the 5s value, but the measurements on the other end of the spectrum are overwhelmingly more likely to have the same background processes that gave us 10s 9 times out of 10 before, so you'll show a 200% speed loss in that scenario, whereas the
22:58 Zoffix pertubation will show up as merely 1% change were the averages taken.
22:59 Zoffix From what I understand you say the lower the number the more pure the measurement is, but that purity does not exist on both ends of the spectrum.
23:02 Zoffix at the same time
23:11 japhb OK, let's look at some more "realistic" data.  If you run a test several times in a row, and the results are 10.3, 10.1, 10.5, 10.2, 10.0, 10.4, 10.3, 10.2 -- how fast do you consider the code to be?  10.0 (min)?  10.25 (mean)?  10.2 or 10.3 (median/mode)?  10.5 (max)?
23:13 japhb (FWIW, I fully understand your point about poisoning, but my argument on that angle is that a half-sized measurement would poison the mean as well, though not the median).
23:14 japhb Actually, I think it's a strict improvement to change from mean to median, and that's still safe from single-value poisoning (and in fact up to 50% poisoning).  Would that be an acceptable change for you?
23:19 Zoffix ZofBot: MATH IS HARD! LET'S GO SHOPPING!
23:19 ZofBot Zoffix, well, on moar we have a native array as expected. my earlier message was a false alarm (I mixed things up)
23:20 Zoffix japhb: yeah, median is OK
23:20 japhb Zoffix: Cool, thank you.
23:23 evalable6 joined #perl6-dev
23:57 japhb [Tux]: It looks like the ' --race' is being dropped from the last couple 2x-second test-t-20 results in speed-all.log
23:57 japhb Zoffix: And ... there's your bug, which I considered highly unlikely, but is in fact happening *right now*.
23:57 japhb *sigh*
23:58 * japhb cries quietly into his Triscuits and cheddar

| Channels | #perl6-dev index | Today | | Search | Google Search | Plain-Text | summary