japhb Interesting article from ~1.5 months ago that I came across today, pertinent to inclusion in technical communities: thebias.com/2017/09/26/how-good-in...inclusion/ 01:14
It puts words to an odd feeling I've had in the past that "assume good intent" could be used as a weapon as well as a positive force. 01:15
But there's an in between there that I think the author didn't address: You can assume non-malicious intent *without* also auto-forgiving bad behavior. 01:16
Still, I see the underlying point I think. 01:17
timotimo even if it's not put in as a "rule" i suppose it conveys an atmosphere of "we'll bend over backwards to find the good intents of your bully when you come to us to complain" 01:24
timotimo has only read like a fifth of the article before writing that response 01:27
japhb Yeah, I think that's the takeaway (the rest of the article is worth reading though) 01:32
Zoffix Looks like SJW hogwash to me. 01:33
Everyone's under attack and if you disagree, you're part of the problem. 01:34
japhb I fail to understand how SJW is an insult. Like liberal, scientist, or gay, it's an adjective that has no intrinsic negative connotation for me. ("So you're saying this person fights ... for justice. And that's bad ... how exactly?") 02:40
Also, we've already decided to be inclusive with every hug we've got, so studying how to be better at that, even if some of the lessons run counter to our existing preconceptions, seems worthwhile. 02:54
Zoffix japhb: SJW is an insult because SJWs' "fight for justice" has the opposite of the intended effect and the SJWs fail to realize that. They "fight" using the same tools that create injustice and their apparent goal is to eradicate *the people* who espouse unwanted ideas rather than eradicating the ideas themselves. 02:57
japhb I think I see the theory of what you're saying, but I'd need strong evidence that's the reality, given the known prevalence of techniques such as gaslighting. As a first test, who uses SJW as an insult loudly and publicly? If they are in a group directly in conflict with the claimed SJWs, they have an incentive to misinterpret/misdirect. 03:15
Zoffix You work at google, right? I think the "Google Memo" fiasco is a good example. Granted, the onslaught of the mindless masses who never even read it probably affected the outcome, but you just gotta chuckle at the irony of firing a guy for pointing out Google's policies are creating a hostile environments for those who don't lean left politically. The fact that right-wingers and their families suffer 03:31
material damage due the ideas they hold is I think exactly the proof you're asking for. This is thought police.
japhb He wasn't fired for pointing out hostility towards right-leaning people. As I am not a lawyer, I need to be careful not to speculate, but I can say I thought "The Case for Empathy" (www.youtube.com/watch?v=T2olLcujHl8) had an interesting view on it. 03:36
MasterDuke m: use nqp; my int \i = nqp::unbox_i(1) # kind of surprising you can have sigil-less natives as routine parameters, but not as `my` variables 03:45
camelia Type check failed in binding; expected int but got Int (1)
in block <unit> at <tmp> line 1
japhb MasterDuke: That does seem a little odd, though sounds like a simple omission or NYI. 03:46
Zoffix japhb: he's just one of the examples. A quick search shows other cases; even a tubler blog called "racistsgettingfired". And my premise is this: getting a racist fired does nothing for "social justice". It doesn't address the cause why that person thought the idea of racism was a good one. It merely is an agression against a person perceived to hold an idea. It's like an Atheist being against religion and 03:49
deciding to fire every priest they could find. It'd do naught to discourage believe in that religion and will only strengthen in (the virtual martyrdom).
MasterDuke it's one of those things where it's late and i'm tired, so i could very well be forgetting some simple rule about containerization or something like that. or maybe what i'm doing should work. who knows... 03:51
Zoffix japhb: I think it can be viewed as an ultimate final goal: when the social justice "war" is complete. What's the final outcome? Are we meant to be paranoid about some hair style that's predominantly used by $culture and we can't use it because we'd be "racist"? Do I need to find out a person's sex and skin color before I complement their intelligence because of "microagressions" that enforce stereotypes I 03:52
intellectually know to be false? Or is it a world where we share our culture and care for another?
Anyway. I guess I started this OT topic :)
Zoffix shuts up and goes to be
d
japhb Zoffix: I would submit that your premise is incorrect. If someone in a position of power creates a hostile work environment, that's a bad thing in and of itself (and also illegal). It doesn't matter *why* they are hostile and in power, *that* they are is a problem in and of itself that should be solved. 04:04
As for the topic being OT, well -- it certainly did go a bit afield, but this comes back to the environment we as a community want to build. 04:05
We need to decide how to handle cases where people are being hostile to newcomers, or beginners, or people from different backgrounds, or fans of other languages, or what have you. 04:06
Hostility and intolerance are easy to simplistically debate when someone goes Godwin, but it's a real problem that we can't be flippant about. We need to work hard to create a good environment. 04:07
Zoffix japhb: exactly, it's illegal. Why is a twitter mob used where the legal instruments are available? 04:08
japhb Empathy, inclusiveness, and psychological safety are *hard work*, but oh so very worth it in the end.
Zoffix: Because the legal system is not unbiased, it is not equitable, and it is certainly not perfect. 04:09
Zoffix And a mob is? 04:11
japhb Zoffix: What tools do the disempowered have? (Also, I will call you on the term 'mob'. Not every social movement is a 'mob' -- that's a term often used to discredit people organizing for a cause.) 04:13
Zoffix japhb: depends on the country. Here (Canada) everyone's legally equal. Everyone has the same tools. 04:14
japhb: laws-lois.justice.gc.ca/eng/acts/e-...lText.html 04:17
japhb And FWIW, I'm not trying to debate all the politics and drama surrounding us right now. I'm just saying that listening and trying to find the useful bits of various thoughts on community building is a good thing. We shouldn't discredit someone's ideas just because they are activists on some subject.
Zoffix Fair enough.
Zoffix plays against self and points out Employment Equity Act does not protect based on sexual orientation 04:23
There's some protection by the Human Rights Act tho: laws-lois.justice.gc.ca/eng/acts/H-...1.html#h-5 04:24
oh man almost midnight /o\ 04:25
Zoffix drops to bed
japhb Also, it doesn't protect private sector employees in Yukon, the Northwest Territories, or Nunavut, nor employees of entities under 100 employees
Good night, Zoffix. Sleep well.
Zoffix P.S.: ^ this would be social justice to me. Plugging in these holes in legal documents. Not debating whether "white list/black list" are sexist terms :) 04:26
*racist 04:28
lizmat Files=1229, Tests=75827, 318 wallclock secs (14.75 usr 5.35 sys + 2197.03 cusr 214.91 csys = 2432.04 CPU) 08:23
so, do we have a solution for modules wanting to use atomic ints that should also work on the JVM backend? 08:43
or are those operators stubbed on the JVM ?
[Tux] Rakudo version 2017.10-138-gfc9e7241b - MoarVM version 2017.10-58-gad8618468
csv-ip5xs1.115 - 1.116
csv-ip5xs-2012.814 - 13.047
csv-parser12.352 - 12.920
csv-test-xs-200.431 - 0.465
test11.920 - 12.994
test-t3.245 - 3.503
test-t-2058.960 - 59.998
test-t-20 --race21.706 - 22.639
09:02
yoleaux 6 Nov 2017 17:47Z <japhb> [Tux]: Could you remember the lines in speed-all.log that contain "I'll try to remember to run the timings later today. System is rather busy"?
yoleaux 6 Nov 2017 17:47Z <japhb> [Tux]: s:1st/remember/remove/
japhb, removed 09:05
lizmat [Tux]: I can't help but think something has changed with the new benchmarks: the increase we see with test-t I don't see that in tests I run on my machine 09:09
could it be the machine is getting a bit hotter and thermal controls throttle the CPU ? 09:10
[Tux] I don't think so: it always builds perl6 just before the timing, so if that is heating it up, it will be the same heat-up as allways 09:11
the load is now 0.2, lots of free mem, no swapping
is there a command to tell the CPU to use max? 09:12
lizmat is the wrong person to ask that for a Linux machine :-)
[Tux] hehm mac is just BSD, also linux :P 09:13
lizmat yeah, but these types of things are generally hardware / motherboard specific, no ?
nine japhb: I guess the gist is "assume good intent" is excellent _advise_ in communication (especially non-face-to-face), while it's bad as a _rule_ for a community. 09:14
gfldex Zoffix: could you send this guy one of your brochures please? :-> blog.abevoelker.com/sick-of-ruby-d...ogramming/ 09:16
nine [Tux]: I usually set the frequency govenor to performance when benchmarking: cpupower frequency-set -g performance 09:17
[Tux] Bummer. I recognize that and even have that as a comment: time.pl:# cpupower frequency-set -g performance 09:18
nine One can also set a specific frequency with -f instead of -g 09:19
[Tux] re-running time-twice
Rakudo version 2017.10-138-gfc9e7241b - MoarVM version 2017.10-58-gad8618468
csv-ip5xs1.117 - 1.224
csv-ip5xs-2012.985 - 14.379
csv-parser12.779 - 15.101
csv-test-xs-200.421 - 0.513
test12.441 - 12.731
test-t3.280 - 3.524
test-t-2060.739 - 61.593
test-t-20 --race20.621 - 21.645
09:26
lizmat well, for an identical system, we see quite a lot of variance :-( 09:31
afk& 09:47
gfldex jnthn: do you got a test for IO::Socket::Async that checks if the Supply is closed when a socket is closed at the remote end before any data was send from the remote end? 09:58
I think I might need such a test. :-/
jnthn m: start react whenever IO::Socket::Async.listen('localhost', 33333) { .close }; sleep 0.5; given IO::Socket::Async.connect('localhost', 33333) -> $conn { react whenever $conn { } }; say 'ok' 10:03
camelia ok
gfldex m: start react whenever IO::Socket::Async.listen('localhost', 33333) { .on-close: { say 'on-close' }; .close }; sleep 0.5; given IO::Socket::Async.connect('localhost', 33333) -> $conn { react whenever $conn { } }; say 'ok' 10:17
camelia ok
gfldex should .on-close work in this case? 10:18
jnthn What are you calling .on-close on? 10:19
Looks like the socket, which isn't going to work
It'll probably be dying with a no such method exception
gfldex m: start react whenever IO::Socket::Async.listen('localhost', 33333) { .Supply.on-close: { say 'on-close' }; .close }; sleep 0.5; given IO::Socket::Async.connect('localhost', 33333) -> $conn { react whenever $conn { } }; say 'ok' 10:20
camelia ok
jnthn That's better but the Supply is never tapped 10:21
whenever .Supply.on-close({ say 'on-close' }) { } should do it 10:22
gfldex with my golfed httpd, if I hit ctrl-r in the browser it will hang after a few GETs. I'm trying to find out where it hangs. 10:23
Or better when I hold ctrl-r for a few seconds.
anyway, I need to run to work now. Debugging is for later.
gfldex runs
jnthn o/
pmurias how do we stop github for marking commits as failed because appveyor is broken? 12:58
AlexDaniel you're talking about nqp repo, right? 13:10
there is this ticket: github.com/perl6/nqp/issues/372
I think what we have now was added in an attempt to resolve it 13:11
pmurias AlexDaniel: it needs to either work or get removed 13:26
AlexDaniel pmurias: yea, that's right. I'd say the first option is more preferable :) Perhaps samcv knows more about it 13:27
pmurias AlexDaniel: we can turn it on when it works 13:33
Geth rakudo: 948a3fa479 | (Elizabeth Mattijsen)++ | lib/Telemetry.pm6
Fix typo
14:32
lizmat commute to amsterdam&
Geth rakudo: 1bc9936ad7 | (Elizabeth Mattijsen)++ | src/core/signals.pm
Make sure that safe signals are more likely handled

If general workers workers are very busy, any handling of signals would be queued with the other jobs waiting. By moving signal handling to the timer-queue, which is intended for fast delivery of small jobs, we have a large chance of signals getting delivered.
15:04
Zoffix Someone's warning against using getrusage: www.reddit.com/r/perl6/comments/7b...tablished/ 15:28
yoleaux 14:51Z <perlpilot> Zoffix: you might want to include a couple of sad-path examples in WWW
Zoffix No, thanks. `orelse` isn't rocket surgery. 15:29
lizmat Zoffix: also, the supervisor calls it from a single thread, and the snapper also from a single thread, typically only 10 times / second 15:35
Zoffix ah right
lizmat so, less chance of lock contention 15:36
afk for a bit again&
jnthn Hm, think I have a fix for #1230 16:23
Darn, nope. Fixes the bug, regresses something lese. 16:30
*else
Ah, I forgot something 16:33
Geth rakudo: 965575710c | (Jonathan Worthington)++ | 3 files
Ensure CLOSE phasers have correct outer chain

Fixes #1230.
16:43
synopsebot RAKUDO#1230 [closed]: github.com/rakudo/rakudo/issues/1230 [concurrency] CLOSE phaser can have busted outer scope
Geth roast: f910bc98d8 | (Jonathan Worthington)++ | S17-supply/syntax.t
Test for rakudo/1230
16:44
jnthn There we go
AlexDaniel dislikes the ISSUE_TEMPLATE we have now 16:46
anybody else feeling this way?
jnthn AlexDaniel: What do you dislike about it, in particular? 16:50
I think maybe "The Problem" could do away 16:51
Expected/Actual behavior are really useful to guide people to submit.
Zoffix likes ISSUE_TEMPLATE 16:52
AlexDaniel as a person reporting a bug, I don't want to fill in half of the questions. I end up deleting the whole thing anyway
(taking my latest regression ticket as an example, all required info is in there already)
Zoffix So delete it. That's what the instructions tell you to do. The people who don't know what to report, on the other hand, have the questions to guide them. 16:53
AlexDaniel It feels like we can guide people with subsequent questions when needed
Zoffix "subsequent" to what? The person might not file the report at all. 16:54
"subsequent" to what? The person might never respond again.
"subsequent" to what? The person might no longer have the problematic machine/program in their possession. 16:55
AlexDaniel you think this template will make them feel more welcome? It could be so, but I see it as annoying paperwork
Zoffix Yes. The template tells them they came to the right place. 16:56
And it specifically says the template is optional.
jnthn AlexDaniel: I suspect you're not the target audience for the template.
Suggesting to people who may not report compiler bugs week in, week out, what information is useful, is a good idea IMO 16:57
AlexDaniel we'll see how it goes 16:59
jnthn Indeed, with GitHub issues in general. So far I think it's going OK. 17:00
[Coke] Having done support in various cases for a while, we need to think of our own sanity to some degree. if the report has no expected/actual information, we can't easily fix it anyway. 17:03
so if someone has a doubt about how to report, the template is overall helpful.
jnthn bbl 17:05
lizmat csv-test-xs-20 0.425 - 0.476 18:25
csv-ip5xs 1.101 - 1.102
test-t --race 1.258 - 1.430
test-t 3.201 - 3.206
test 11.771 - 11.860
csv-parser 12.160 - 12.448
csv-ip5xs-20 12.567 - 12.718
test-t-20 --race 19.686 - 21.689
test-t-20 57.704 - 59.213
new: test-t --race :-)
timotimo oh, would you look at that 18:48
lizmat thought experiment because of a gettext presentation I'm looking at: 21:10
could this be made to work? If so, how? 21:11
m: role Deutsch { }; my $a = "one two three" but Deutsch; say $a # how to get "eins zwei drei" ??'
camelia one two three
lizmat I guess some MOP magic ?? 21:13
gfldex m: role Deutsch { method Str { <1 2 3>.Str } }; my $a = "one two three" but Deutsch; say $a
camelia one two three
gfldex m: role Deutsch { method gist { <1 2 3>.Str } }; my $a = "one two three" but Deutsch; say $a 21:14
camelia one two three
AlexDaniel m: role Deutsch { method Str { <1 2 3>.Str } }; my $a = "one two three" but Deutsch; put $a
camelia one two three
AlexDaniel m: role Deutsch { method Str { <1 2 3>.Str } }; my $a = "one two three" but Deutsch; put $a.Str
camelia 1 2 3
lizmat I guess our text in say/put should be more like type =:= Str rather than isa Str 21:15
but we can'd do that in MMD
afaik 21:16
Geth rakudo: MasterDuke17++ created pull request #1239:
Simplify INTERPOLATE_ASSERTION
21:24
synopsebot RAKUDO#1239 [open]: github.com/rakudo/rakudo/pull/1239 Simplify INTERPOLATE_ASSERTION
Zoffix m: role Deutsch { method Str { "eins zwei drei" } }; multi say(Deutsch $_) is default { say .Str }; my $a = "one two three" but Deutsch; say $a 21:36
camelia eins zwei drei
Zoffix m: role Deutsch { has str $!value is box_target = "eins zwei drei" }; my $a = "one two three" but Deutsch; say $a 21:38
camelia While composing Str+{Deutsch}: Duplicate box_target for native str: attributes 0 and 0
in block <unit> at <tmp> line 1
Geth tap-harness6: CurtTilmes++ created pull request #20:
Fix Issue #19: write-only $/
21:44
rakudo: tbrowder++ created pull request #1240:
Fix rt #128221
22:09
synopsebot RT#128221 [open]: rt.perl.org/Ticket/Display.html?id=128221 [BUG] Weird internal error when parsing some very simple Pod with '-' in the first column of a table in Rakudo
RAKUDO#1240 [open]: github.com/rakudo/rakudo/pull/1240 Fix rt #128221
Geth roast: tbrowder++ created pull request #353:
Fix rt 128221
22:17
synopsebot SPEC#353 [open]: github.com/perl6/roast/pull/353 Fix rt 128221
Geth rakudo: fa8bc54981 | (Daniel Green)++ | src/core/Match.pm
Simplify INTERPOLATE_ASSERTION

by removing some unused variables and conditionals.
22:50
rakudo: 0e50993f41 | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | src/core/Match.pm
Merge pull request #1239 from MasterDuke17/simplify_INTERPOLATE_ASSERTION_even_more

Simplify INTERPOLATE_ASSERTION
synopsebot RAKUDO#1239 [closed]: github.com/rakudo/rakudo/pull/1239 Simplify INTERPOLATE_ASSERTION
tap-harness6: 738cf16d4c | (Curt Tilmes)++ | lib/TAP.pm
Fix Issue #19: write-only $/
23:08
tap-harness6: f1a9d84ed9 | (Zoffix Znet)++ (committed using GitHub Web editor) | lib/TAP.pm
Merge pull request #20 from CurtTilmes/master

Fix Issue #19: write-only $/
tbrowder hi, lizmat! 23:54
lizmat hi tbrowder
just back from Amsterdam.pm meeting (another ~2 hour drive through dense fog half of the way)
so don't expect much from me at this time :-)
tbrowder i finally submitted my rakudo PR for pod tables, including tests in a roast PR. i hope you can check it out when you get a chance. 23:55
fog, yuck! glad you arrived safely. 23:56
lizmat yeah, me too :-) 23:58
will look at it tomorrow, unless someone beats me to it :-)
good night,
&
timotimo pod tables!
i'm glad about that
tbrowder thnx, nite!
timotimo i once tried to make them work better. unfortunately it was a two-pass-parsing thing, so TT wasn't very happy 23:59