Perl 6 - the future is here, just unevenly distributed

IRC log for #pr-challenge, 2015-01-03

| Channels | #pr-challenge index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:11 bjakubski joined #pr-challenge
00:14 ether neilb: I'd be content with just seeing hte list of dists that were assigned; if I want to contact the assignee for a dist I can go through you
00:16 neilb oh, ok. cool. I’m just working through all today’s signup emails. When I’m done, and have sent them all their jan asignment, I’ll send you a full list. ping me if I forget, had a wass of gline or two
00:19 vroom joined #pr-challenge
00:21 mwhisenhunt joined #pr-challenge
00:29 ether "wass of gline"++
01:15 vairav joined #pr-challenge
01:26 vroom joined #pr-challenge
01:28 vroom joined #pr-challenge
01:31 vroom joined #pr-challenge
01:31 neilb 245 people have now been emailed assignments. I’ll email you tomorrow ether — need sleep now!
01:38 vroom joined #pr-challenge
01:40 ether wow!
01:40 ether who says perl is dead
02:02 vroom joined #pr-challenge
02:10 vroom joined #pr-challenge
02:12 Su-Shee_ joined #pr-challenge
02:14 vroom joined #pr-challenge
02:15 vroom joined #pr-challenge
02:20 vroom joined #pr-challenge
03:54 adelq joined #pr-challenge
03:59 popl crap, I missed neilb
04:00 popl I wonder how many people signed up for the challenge. I still haven't been assigned anything.
04:07 oalders judging by the backlog, 245 people?
04:11 popl oalders: at least 245 people, at least
04:12 popl s/, at least/
04:12 popl /
04:12 * popl needs to eat
04:21 vroom joined #pr-challenge
04:32 isBEKaml joined #pr-challenge
04:38 vroom joined #pr-challenge
04:56 adelq joined #pr-challenge
05:34 estrabd popl: eat
05:34 estrabd damn, 245 people, NICE!
05:40 isBEKaml estrabd: I'm pretty sure it's more than that :-)
05:40 isBEKaml estrabd: perhaps neilb only mentioned that number as *recent* signups?
05:53 isBEKaml_ joined #pr-challenge
05:54 popl estrabd: I still haven't eaten.
05:54 popl I need to.
05:54 popl :P
05:54 popl I keep getting sucked into the horrible code from $work.
05:54 popl It's mesmerizing, like a cobra.
05:54 popl I've resorted to Variable::Magic to tell me wtf these variables are getting populated.
05:54 popl getting annoyed. ;)
05:56 isBEKaml popl: you're working? On a Saturday?
05:57 popl isBEKaml: it's Firday still here.
05:57 popl but yeah, I'll be working this weekend.
05:57 popl I told my boss I would.
05:57 popl *Friday
05:58 popl I mean, he didn't ask me but I need to finish figuring out this bug.
05:58 popl in this 20 year old Perl
05:59 popl weee
05:59 isBEKaml popl: yeah, I know what you mean. I've been sucked into that routine a few times myself :D
06:00 dfluck joined #pr-challenge
06:02 isBEKaml popl: you better figure that bug out quickly so you can still get most of your weekend to yourself :-)
06:02 popl there's another $thing I need to work on after that
06:03 popl but I do need to do stuff this weekend
06:04 popl nobody said I can't drink beer and write code at the same time, though.
06:04 popl Paulaner Salvator ftw!
06:12 vairav joined #pr-challenge
06:26 dfluck_ joined #pr-challenge
06:56 vairav joined #pr-challenge
07:33 jashank sawyer, moritz: Ask and I shall blog: http://www.rulingia.com/~jashank/blog/2015/01/02/perl-gems/
07:36 isBEKaml joined #pr-challenge
07:39 isBEKaml jashank++
07:43 isBEKaml jashank: for me, the big moment when it all dawned on me was the realization that programs are just about interactions between lines of code. That no code stands on its own in a vacuum. Then it was "it's all just data, man!" :-)
07:44 jashank isBEKaml: Right.
07:47 dfluck So Lisp? ;)
07:50 isBEKaml dfluck: well, meta-programming is talking about meta-data ;-)
07:50 jashank I admit, Lisp got a lot easier once I made that leap too.
07:50 dfluck I was always told about the zen moment when you finally "get Lisp", although I think mine was gradual.
07:50 jashank (Although I only really use Emacs-Lisp, and usually need dash.el so my brain doesn't explode.)
07:51 dfluck I learned Lisp with Emacs Lisp; it was fantastic.
07:51 dfluck Made Racket much easier in my prog lang course this past semester.
07:51 jashank I quite like the look of Racket, but it's way down my round tuit.
07:54 dfluck If you know Emacs Lisp, it's not too different.
07:54 jashank I've always found Scheme and Lisp to be sufficiently different that it takes me a moment to switch between the two.
07:55 jashank I learned MIT Scheme from gjs and hal's 6.001 videos when I was... hm, 4? 5?
07:55 dfluck Oh, wow.
07:55 dfluck Did that make picking up less Lispy languages difficult?
07:55 isBEKaml ehh, wow
07:55 isBEKaml you were coherent when you were 4?
07:58 jashank dfluck: Not really? The first non-Lisp languages I learned were C and Perl, and my experience of them was so radically different to Lisp that I didn't really closely associate them?
07:58 sivoais jashank++ # great post!
07:59 jashank isBEKaml: Even earlier, apparently.
08:00 isBEKaml jashank: wow
08:00 bjakubski joined #pr-challenge
08:00 jashank sivoais: Only rewrote it about a dozen times.
08:02 sivoais I know that feel. That's one of the barriers I'm trying to get over with my writing.
08:05 jashank sivoais: Writing is rewriting!
08:05 jashank I had it drummed into me that if you don't redraft at least once, you're doing it wrong.
08:05 jashank I have dozens and dozens of blog posts waiting for me to redraft them.
08:06 isBEKaml that's why I don't write. It's all too much effort :P
08:06 sivoais same here, but I just never get a chance to go back and say, "this is ready" :-P
08:06 dfluck I've wanted to blog for a bit. I have a post half-written about a Markov-chain generator I wrote in OCaml, but then I went on vacation and... yeah.
08:09 jashank sivoais: Well, how do you know your codebase is ready to ship?  You take a leap at the 90% complete mark, and hope for the best.
08:11 jashank Twitter's a great place to hone that, IMO; you want to tweet an idea or a concept, and you 'compress' it in several passes by word substitutions or grammatical rearrangements, etc. etc. until you get down under the 140-codepoint limit.
08:11 sivoais Never thought about it like that. I like it.
08:15 jashank I stole^Wborrowed that one, but I can't find where from.
08:15 jashank I _think_ it was from Matt Might, but I can't find the article I'm thinking of.
08:21 meredith joined #pr-challenge
08:24 mstr joined #pr-challenge
08:29 moritz jashank++
08:42 neilb joined #pr-challenge
08:47 BRAD joined #pr-challenge
09:15 diocles joined #pr-challenge
09:28 vairav joined #pr-challenge
09:39 neilb_ joined #pr-challenge
09:41 kentnl jashank: one of the problems I see with perl map/grep is they're inherently "backwards" to my way of thinking
09:42 kentnl You have 4 of those in a sequence, and I have to read it forwards to find the input, then work backwards through the chain to find where it goes
09:43 kentnl where I'd rather think in more bash style flow:   input | process | process > output
09:43 popl I love map.
09:43 kentnl Its also one of things I find bad to have dogmatic use of.
09:43 popl When I program functionally it usually feels cleaner.
09:44 kentnl For instance, if you've seen autobox which allows you to use map left-to-right like ruby, that's an improvement
09:44 kentnl but I often find if you have 4 chains, mixing map and grep, it can be better expressed as a single loop with proper flow control
09:44 kentnl grep becomes transposed as "next if $cond"
09:45 meredith elixir's pipelines are easy on the eyes for that sort of thing
09:46 kentnl That said, I have a penchant for intentionally dumbing down my code, because who knows who will read it tomorrow and not understand it :D
09:46 pink_mist don't forget the brilliant List::Gather by ether that provides gather/take semantics ... that's also awesome to use
09:46 pink_mist ether++
09:47 meredith ^aye
09:49 kentnl rafl++ # for the orignal code , ether++ for maintaining everything
09:50 * kentnl thinks we could have a CPAN built of just the things ether has comaint on
09:50 kentnl but I am shy of a good punny name for such a device ;)
09:51 pink_mist Ethereal Perl Archive Network?
09:53 popl herp
09:54 jashank kentnl: Right, and that's one of the things that I feel makes Ruby's Enumerable methods much easier to jump into---they all have quite obvious directions of flow.
09:55 kentnl jashank: Its probably also nicer in languages that comprehend the chain and optimise it, instead of perl, where I'm seeing "great, so I create and destroy several temporary lists here" every time I do it.
09:56 kentnl similarly, grep {  } 1 .. 2**(bignumber)   # gives me the spooks about creating a massive list of integers if I don't intend to use most of them and don't need them in memory all at once.
09:57 popl kentnl: languages like haskell do things lazily
09:58 popl I think Perl6 does that too
10:12 jashank popl++
10:13 neilb joined #pr-challenge
10:13 jashank Yeah, one of the things that I'd love to see as an optimisation for (Perl|Ruby|...) is the threaded handling of set transformations.
10:14 jashank Single-queue-many-consumer or whatever for small lists or for transformations that are obviously stateless.
10:14 jashank Of course, then you have the problem of determining what's a stateless operation.
10:18 pink_mist hope that's not equivalent to the halting problem :P
10:20 LLamaRider joined #pr-challenge
10:20 isBEKaml joined #pr-challenge
10:22 jashank I... hm.
10:22 jashank I don't think it is?
10:24 pink_mist I'm honestly not sure; hence me merely hoping it isn't, instead of stating whether it is or isn't :P
10:24 sue joined #pr-challenge
10:24 jashank If I have a 'kernel' (using the OpenCL sense of the word) that runs a task, I can, by spelunking a "low level" (I think I actually want the phrase 'register language'?) representation of the kernel, determine what registers it touches, and if those registers are outside the kernel, then it's clearly stateful?
10:25 jashank I guess you could stretch that intuitive feely-proof further and say that if a block in a map or a grep or whatever is accessing anything outside the block's scope, it's stateful?
10:26 jashank That may be far too conservative, though, for a worker-queue optimisation.
10:26 jashank pink_mist: I think the tl;dr is "no, I don't think it is".
10:26 pink_mist if there's an eval in there, how can you be sure?
10:27 jashank If there's an eval in there, you have bigger problems :P
10:27 jashank I suspect the optimisation I'm trying to feel towards here is only applicable in very naive cases.
10:27 pink_mist yes probably :P but this is supposed to be a general thing, and people will do crazy things just for the heck of it :P
10:28 jashank Very true.
10:28 jashank So, for the most part, ignoring the overheads of threading, a naive optimisation in that way would yield _some_ performance improvement.
10:28 jashank (Also, why am I feeling like I should be emphasising things using I<...>, not _..._?)
10:29 pink_mist too much pod, not enough irc :P
10:29 jashank Heh.
10:31 jashank ... so $self->father just wandered in with a ticking clock written in Perl, which deliberately ticks subtly irregularly.
10:31 popl well done
10:33 popl jashank: He might like this: https://www.thinkgeek.com/product/b278/
10:35 jashank popl: I've pulled a prank with one of those before.
10:37 jluis joined #pr-challenge
10:40 mstr joined #pr-challenge
10:41 vairav joined #pr-challenge
10:48 popl jashank: I've imagined pulling a prank with one of those before.
10:48 popl Unfortunately it devolved into an office-wide homicide of subservients.
10:49 popl Some bosses really don't like being pranked.
10:50 popl Okay, I think I need sleep. :P
10:54 briang joined #pr-challenge
10:55 jashank popl++
10:55 pink_mist sleep++ :P
10:56 sawyer morning
10:56 briang if I clone a dist from github and it has unactioned PRs, is there any way I can apply those PR to my clone?
10:57 isBEKaml briang: you need to pull patches off those PRs and merge them by hand into your clone
10:57 jashank "by hand" is a bit of an understatement.
10:58 jashank If you're comfortable with git, set up some remotes and merge or rebase your master.
10:58 Su-Shee_ left #pr-challenge
10:59 Su-Shee joined #pr-challenge
10:59 moritz or you can   git pull https://github.com/$forker/$poject.git $forked_branch_name
11:00 isBEKaml jashank: you mean overstatement. :-)
11:00 isBEKaml anyway, that's what I'd have done. Not that I'm not very comfortable with git.
11:00 Su-Shee good mornibg everyone.
11:00 moritz good morning Su-Shee
11:01 isBEKaml moritz: $forker -- good name :-)
11:01 jashank Github -- what the fork.
11:02 vairav joined #pr-challenge
11:02 isBEKaml Github -- fork me up!
11:02 isBEKaml ForkApp!
11:02 briang what? proper $forked?
11:03 moritz $github_user_who_issued_the_pull_request :-)
11:05 isBEKaml briang: just be prepared for some headaches with unmergeable code. That is, if the main repo is so far advanced (by commits) beyond the point where the fork was done, it's a real pain to merge them.
11:05 mulander joined #pr-challenge
11:06 briang actually,the PR I'm looking at is trivial (two lines). I'm trying to generalise :)
11:07 isBEKaml briang: or you can just choose to skip all PRs that have not been acted upon the longest
11:07 briang In this case, I can't ignore it. It fixes a deprecation
11:09 moritz if you can't ignore it, it might make sense (or not) to base your branch off of the branch from which the pull request was made
11:10 briang It looks like the fork was deleted
11:10 briang sergeyromanov wants to merge 1 commit into riemann42:master from unknown repository
11:10 moritz :(
11:11 sawyer whoa
11:11 moritz briang: then the simplest thing is probably to append .patch to the pull request URL
11:11 briang yes!!
11:11 jashank moritz++
11:11 moritz briang: and use 'git am $filename' to apply it
11:11 briang thank you moritz
11:11 briang moritz: ++
11:11 moritz which preserves meta data
11:11 briang I didn't know you could do that
11:11 moritz (at least author, date and commit message)
11:12 isBEKaml briang: yes. that's a relatively unknown feature of github
11:12 jashank I'm sure I've seen that before.
11:13 briang that works, and is exactly what I was looking for. thnaks again
11:13 briang *thanks
11:13 isBEKaml briang: if you merge that with the .patch, you preserve all author details. So, while your contribution is acknowledged in your PR, you're sure that the previous authors are also acknowledged.
11:13 * isBEKaml .oO(Too many 'the's kill the language)
11:14 briang isBEKaml: As I said, that's exactl what I was looking for
11:15 isBEKaml briang: great. If you go back through the backscrool, this is what I meant when I said "you need to pull patches off those PRs" Sorry if I wasn't very clear. :-)
11:15 isBEKaml *backscroll
11:16 briang isBEKaml: You could probably hear the whoosh of me missing your point ;)
11:19 isBEKaml briang: whoa - we have all been there :-)
11:23 jluis joined #pr-challenge
11:27 isBEKaml joined #pr-challenge
11:40 Mike_B joined #pr-challenge
11:40 isBEKaml DAE find reading code in the browser(sans synhighlight) more pleasant than a $editor? :-)
11:41 Mike_B Woah, 50+ people in the channel!
11:41 Mike_B rt.cpan.org seems to be having proxy issues: https://rt.cpan.org/Ticket/Display.html?id=33816
11:42 isBEKaml Mike_B: we have about 5x that number signed up!
11:42 Mike_B Does anybody know if the reverse proxy is handled by BestPractical or noc.perl.org?
11:42 Mike_B isBEKaml: yeah it's really amazing.
11:44 Mike_B I'm assigned DateTime::Format::Epoch https://metacpan.org/pod/DateTime::Format::Epoch - it has a couple of open bugs but most have been 'patched' in Git already or even fixed; and the other ones are not really valid as far as I can see... I'll have to look hard to think about a proper PR.
11:44 Mike_B Ah: RT is back again :D
11:45 jashank isBEKaml: On occasional, but a lot of time time I prefer to have my editor massaging code for me, so I get a sanitised view?
11:46 isBEKaml jashank: it may be just me today, but I find that reading code in the browser kind of forces me to pay attention than just skim lines.
11:47 isBEKaml jashank: with an editor, I get lazy eye syndrome :P
11:47 isBEKaml or a roving eye ;-)
11:56 jkg_ joined #pr-challenge
11:58 dfluck I think for me it's way too easy to miss something little that would've been easy to see had it been highlighted.
11:58 dfluck Helps group stuff together too, at least in my brain. :P
12:00 mstr joined #pr-challenge
12:14 daveh joined #pr-challenge
12:15 isBEKaml dfluck: to each their own, then :-)
12:15 isBEKaml dfluck: although I must say that what you have said holds true for me, occasionally
12:18 sawyer .op mstr jkg_ daveh
12:18 sawyer argh
12:27 _Dave OK, Aristototle Pagaltis said they shut off registrations on blogs.perl.org because security
12:35 diocles joined #pr-challenge
12:36 * jashank decides to head off for the night.
12:42 sawyer _Dave: is there any intention on reopening them?
12:42 _Dave i've posted a comment on the bug
12:43 _Dave https://github.com/blogs-perl-org/blogs.perl.org/issues/293#issuecomment-68503362
12:43 _Dave davorg is still trapped in Africa
12:47 sawyer _Dave: replied to it
12:47 _Dave ok, 200 was a mild  exaggeration ;)
12:49 _Dave sawyer: is the git process documented? i'm about to clone my module...
12:50 sawyer what do you mean?
12:50 _Dave the clone/fork process
12:50 _Dave i want to get it right :)
12:52 sawyer moritz set up a log, so it should be there
12:52 sawyer there's a link to it from the resources
12:52 sawyer i don't have time now to go into the exact commands
12:52 sawyer been trying to get off the computer all morning :)
12:53 _Dave ok
12:55 _Dave hm, can't find
12:56 _Dave got it
12:56 _Dave misleading link text
12:56 sawyer fix it? :)
12:58 _Dave clone and branch?
12:59 sawyer what?
12:59 _Dave to fix the text in resources?
13:00 sawyer no need for branch there
13:00 sawyer but i can just put you in the contributors group
13:00 sawyer what's your gh id?
13:01 _Dave davehodg
13:02 sawyer done
13:02 _Dave you have a pull request
13:03 isBEKaml _Dave: you have  commit bit. merge it yourself :D
13:05 _Dave merged
13:05 _Dave that was novel
13:05 _Dave never done that before
13:09 mstr sawyer: mind to help me with a git issue again?
13:10 mstr i created a branch. did some work. made some PRs.
13:10 mstr now some PRs got merged and others not (because of merge conflict)
13:10 mstr so i went ahead configured upstream in my local git repo and pulled from the current upstream master to my master
13:11 mstr then i switched to my feature branch and did a rebase from the newly pulled master
13:11 mstr i fixed the merge conflicts
13:11 mstr but then when i tried to push my feature branch to GH it got rejected
13:12 mstr so i did a "git push --force origin +branchname"
13:13 mstr now considering the documentation to "push --force" that might not be what i was supposed to do.
13:13 mstr what would be the proper way to do what i was trying to achieve?
13:16 mstr everything seems to be fine. i can see the old PR still there. i can see that my rebase worked. its just not obvious that i did all that stuff in the background.
13:26 kentnl mstr: github is a bit odd in that if you change SHA1s and you're not the owner of the pull request, then the change won't get detected.
13:27 mstr ok. i am the owner of that PR. everything seems to be detected just fine.
13:27 mstr i'm just wondering if my "push --force" was the right thing to do
13:28 kentnl yeah, it often is if you want to say "previous sequence of commits is wrong" and replace them with a new sequence.
13:28 kentnl ( which is the case of rebase )
13:29 kentnl though you probably don't need force *and* "+", + implies force
13:38 mstr thanks
13:38 mstr i'm wondering though. everybody always says that one thing git is so much better in than anything else is merges.
13:39 mstr yet if you have 2 PR both just adding to the same file (README eg) git doesn't merge as expected
13:40 mstr so i'm either forced to wait until PR#1 is merged, rebase my other branch and only afterwards issue PR#2
13:40 mstr or, i don't know, maybe combine the two PRs into one?
13:40 kentnl Don't conflate github with git. Github is ok, but it only expresses a subset of git functionality.
13:41 mstr just seems that the whole process forces a lot of work on me
13:41 kentnl and you don't *have* to work that way, it depends on the maintainer. Github itself encourages "merge" by default vs "rebase and fast forward" by default
13:41 mstr yeah i know. i was talking about git rgd the merge conflicts which i don't consider merge conflicts
13:41 kentnl Some people and projects prefer merges, others prefer rebases =)
13:42 kentnl If your two PRs share common code, you could rebase one PR to share that ancestor.
13:42 mstr well seems to me that i just did that rebase to overcome the merge conflict in the README. i guess upstream could have fixed that, but chose not to accept the PR instead
13:43 kentnl Yeah, rebasing is IMHO the "nice" way to avoid merge conflicts.
13:43 kentnl Especially as it lets the person who made the changes resolve the conflict, as opposed to the maintainer needing to resolve it.
13:44 mstr i get that. don't want to cause more work on the maintainer than necessary.
13:44 kentnl The process of doing it "The nice way" does make you do more work, its just the price you pay I guess =)
13:44 kentnl you could spend less work and do it less nicely.
13:45 mstr but if i rebase PR2 from PR1. wouldn't that mean that PR2 has also all changes from PR1? what if the maintainer doesn't like PR1? he would also have to reject PR2 then.
13:46 mstr and all that just so there is no merge conflict. :)
13:46 kentnl yeah, they would. They could also cherry-pick individual commits though, or locally rebase a subsection of PR2 themselves.
13:47 mstr oh my. i just found out about "git reflog". all my dirty secrets. ;)
13:47 kentnl Yeah, reflog is handy. Data never really gets deleted when you do things ;).
13:47 kentnl Even once things fall off the reflog, the objects still lurk in your repository
13:48 mstr but if somebody would clone my public repo- would they see all that stuff?
13:48 kentnl Nope.
13:48 mstr ha! good :)
13:49 kentnl When you push or clone, only the direct parents of commits are fetched.
13:49 kentnl ( recursively )
13:49 kentnl if the SHA1 is not in the ancestry of the clone, it won't replicate.
13:49 kentnl ( and the reflog itself doesn't clone )
13:50 mstr which means that reset actions won't ever show
13:50 kentnl I imagine if you were really clever and had a copy of somebodies reflog you could trick git into sending you the blobs referred to, but I haven't tried it.
13:51 * kentnl heartily recommends reading the Pro Git chapter on internals
13:52 kentnl you can basically see git as a KV store, and some of those V's have textual refererences to other K's .
13:52 kentnl That's about it.
13:55 BRAD joined #pr-challenge
14:03 diocles joined #pr-challenge
14:21 tinypig joined #pr-challenge
14:43 dfluck joined #pr-challenge
14:45 BRAD joined #pr-challenge
15:00 vroom joined #pr-challenge
15:04 lbrown afternoon
15:10 * jkg gets around to emailing $maintainer o/
15:13 Su-Shee haven't even looked yet.. still in vacation mode :)
15:21 vroom joined #pr-challenge
15:31 lbrown so apparently sort() in PHP modifies the original array instead of returning a sorted one
15:31 lbrown took me half an hour to work that one out
15:31 lbrown </rant>
15:40 Su-Shee we're having a PHP challenge as well?! :)
15:40 lbrown no
15:40 lbrown I just like ranting about it
15:40 lbrown do you know php?
15:41 * jrhunt posits that no one can truly "know" PHP
15:41 Su-Shee yes.
15:49 lbrown so a PHP array works kind of like a perl hash, but ordered?
15:50 bjakubski joined #pr-challenge
15:50 Su-Shee it's basically a thing with nuneric keys. but how does that matter with cpan modules?
15:52 lbrown sorry, I'm just going off topic
15:56 vroom joined #pr-challenge
15:58 Su-Shee hm did we decide what to do with modules not updated for long and not being used by anything else?
15:58 pink_mist they get lower scores, so they're not as likely to be assigned - afaik
16:07 Su-Shee hm, my module mostly looks like it's simply.. done. let's see.
16:12 upasana which is your module? :)
16:13 upasana I have not done anything yet, I am still trying to understand how to use that module, it is quite daunting!
16:18 bentglasstube joined #pr-challenge
16:21 Su-Shee upasana: mousex::nativetraits :)
16:21 Su-Shee upasana: yours?
16:22 upasana IO::Socket::SSL
16:23 Su-Shee oh fuck shit :)
16:23 Su-Shee someobe got DBI :)
16:23 upasana I had almost the same reaction :-/
16:24 upasana this will going to be fun for me even if I won't be able to contribute anything :D
16:24 Su-Shee well as sawyer pointed out: "hence challenge" :)
16:24 upasana yes! :)
16:25 Su-Shee it will titally be fun and I'll be damned if I at keast don't get more/better documentatoon in :)
16:26 Su-Shee sorry typing on my phone in a train in tge netherlands..
16:26 upasana w00t! :D
16:26 upasana Which part?
16:27 Su-Shee right now amelo
16:27 Su-Shee on my way back from bf in nijmegen
16:29 upasana Cool!
16:29 vroom joined #pr-challenge
16:29 upasana I never heard "Amelo", I know Nijmegen, have a good journey! :)
16:30 Su-Shee upasana: one of the little towns on the way to the german border
16:31 Su-Shee well.. "border" :)
16:32 upasana Now, I know :)
16:32 bjakubski joined #pr-challenge
16:32 neilb joined #pr-challenge
16:41 thrig ... day 3, installing Dist::Zilla plugins so can actually build assigned module ...
16:44 Su-Shee uhoh :)
16:44 thrig and uh geez dependencies
16:46 pink_mist cpanm to the rescue :P (iirc Dist::Zilla tells you a command to run to automagically install all the deps it needs?)
16:47 thrig yes, and it's still running ...
16:47 pink_mist heh :P
16:48 BHS_error heh Dist:Zilla :)
16:48 BHS_error it just pulls half of the CPAN to install itself
16:48 thrig no, that was yesterday. today is plugins, and more modules
16:49 pink_mist which module, thrig? :P
16:49 thrig Net::Amazon::S3
16:50 pink_mist ohh, intriguing ... but I do still prefer the one I got -- IO::All
16:50 BHS_error I got XML::SAX::Base
16:50 BHS_error had to email the maintainer to have an idea of what it was about :)
16:50 pink_mist oh lord, I hope I don't get an XML module :P
16:51 pink_mist BHS_error: haha, that should give you an idea of what kind of PR to make -- fix the documentation to describe it better :P
16:53 thrig (my first PR will be to mention how to install/build it)
16:53 pink_mist haha
16:53 BHS_error actually the doc is pretty clear, provided you know what you are doing
16:53 bjakubski joined #pr-challenge
16:54 BHS_error I mean, you don't wake up a morning and say "oh, what if I use XML::SAX::Base"
16:55 BHS_error so the maintainer provided me some background regarding the module (why it was existing, how he became maintainer), which was really helpful
16:57 thrig Parse-RecDescent? oh for pete's
16:58 * kentnl renames the channel #pr-mashocism
16:59 thrig 64 distributions and fail! Bailing out the installation for Dist-Zilla-PluginBundle-TestingMania-0.24
17:00 kentnl Ugh. That one is a bit of a cancer.
17:00 kentnl --notest #
17:00 kentnl do you have a dist.ini you can nuke plugins in easily? I'd do that there.
17:02 kentnl thrig: also, the usual "see if prove works" applies for pain avoidance :)
17:03 mstr joined #pr-challenge
17:12 dfluck joined #pr-challenge
17:20 sue joined #pr-challenge
17:23 thrig yeah, deleting the @TestMania block gets rid of that plugin and the beast builds (I'll stick with ExtUtils::MakeMaker, sheesh)
17:49 vairav joined #pr-challenge
18:04 mstr joined #pr-challenge
18:10 vairav joined #pr-challenge
18:14 bjakubski joined #pr-challenge
18:41 mstr joined #pr-challenge
19:03 BRAD joined #pr-challenge
19:07 ivan1 joined #pr-challenge
19:09 ivanwills joined #pr-challenge
19:11 ether TEstingMania is insane
19:11 ether and grossly out of date
19:11 ether dzil plugin authors really need to stay on top of things or their plugins slide into brokenness quite quickly
19:12 ether see also: plugin authors who don't write tests :(
19:12 ether dzil tests are quite easy once you have written a few
19:12 ether my minting profile has a special case for Dist::Zilla::Plugin::* dists and mints a set of template tests that are very easy to fill in :D
19:12 BRAD joined #pr-challenge
19:13 ether look at what got minted for me by default: https://github.com/karenetheridge/Dist-Zilla-Plugin-RewriteVersion-Transitional/commit/d7f9e290aaf642daa1f280847a40a79fbb706696
19:19 tinypig joined #pr-challenge
19:40 diocles joined #pr-challenge
19:48 tolid joined #pr-challenge
19:58 nbezzala joined #pr-challenge
20:12 neilb_ joined #pr-challenge
20:15 lbrown dat feeling when someone's making commits but igoring your PR
20:19 kentnl lbrown: what did you pull on? If I get bored I could pretend to read the diff and give feedback or something :)
20:20 kentnl code review can help get things merged :D
20:20 lbrown you don't want to
20:20 lbrown it's PHP
20:20 kentnl Oh. I see.
20:20 lbrown and it got merged just now anyway :)
20:21 kentnl Yeah. That's a completely different problem.
20:21 lbrown it was really atrocious though
20:22 lbrown matching up language codes with language names was just a random list of case..switch...switch..switch..switch
20:22 kentnl if you tell me you at least didn't make the whitespace worse, I'll give you a gold star for trying
20:22 * lbrown receives gold star
20:22 * lbrown beams with delight
20:23 kentnl \o/
20:23 lbrown I'll sort out the whitespace in that file when I want to practice patience sometime
20:23 kentnl Pet hate is finding myself on projects where random coders just commit stuff with randomly alternating indentation and EOL settings :/
20:23 lbrown 6 spaces some places, 4 in others, someone used tabs in half the lines as well
20:24 kentnl I fuck it up sometimes, but I at least attempt to care about that stuff!
20:24 kentnl "there is extra whitespace at EOL in this commit" # Instant shame :(
20:24 lbrown ouch
20:25 lbrown let's try and improve the German localization
20:25 lbrown using Google Translate
20:31 * kentnl donates a failboat
20:31 lbrown so I just saw '$value = md5($value);'
20:32 lbrown it looks like they're hashing passwords using md5
20:32 lbrown unsalted, too
20:33 kentnl well. that's "Secure"
20:33 lbrown what's the best algorithm now for hashing passwords?
20:33 lbrown SHA1?
20:33 kentnl SHA1 is possibly broken
20:33 lbrown oh
20:33 kentnl actually, it is broken
20:33 lbrown what should we use?
20:34 kentnl I'd be inclined to recommend whirlpool or SHA256
20:34 kentnl and of course, salting with per-user salt instead of a shared global salt.
20:35 lbrown but md4 is easier to implement
20:35 kentnl why not just use CRC16
20:35 kentnl or ROT13
20:35 kentnl I know, use ROT13 twice!
20:35 lbrown double security
20:36 diocles joined #pr-challenge
20:36 lbrown y/a-zA-Z/a-zA-Z/
20:36 lbrown or you could do something really clever like y/a-zA-Z/A-Za-z/
20:36 lbrown hackers will never figure that out
20:39 lbrown oh no, it looks like they are using SHA256
20:39 * lbrown sighs with relief
20:39 kentnl \o/
20:39 lbrown unsalted, though :/
20:40 kentnl I'm guessing they have no key lengthening techniques either
20:41 lbrown $password = hash('sha256', $password);
20:41 lbrown that's it
20:42 kentnl *sigh* its not ideal, but at least its not sha1/md5 :(
20:43 lbrown i would make a PR if i knew how to do it properly myself
20:43 kentnl if you wanted something serious, I'd say throw PBKDF2 at it. There's a native implementation in PHP since 5.5
20:43 kentnl http://php.net/hash-pbkdf2
20:44 lbrown thing is, I don't want to touch it
20:44 lbrown because that means writing code to transition old passwords
20:46 lbrown also, they want to support php 5.3+
20:47 kentnl Shouldn't be too hard, as long as the final stored value gets a prefix added to denote its "new" then you could support both old and new passwords simultaneously, and the storage would change when the password did.
20:47 kentnl https://defuse.ca/php-pbkdf2.htm # In that case.
20:47 lbrown yeah but considering that I don't actually know php
20:48 kentnl nobody does.
20:49 lbrown but I still have no idea how variables even work
20:50 * kentnl leans back and squints
20:50 kentnl they're like Perl, except there's no @ and % , and there's no strict or warnings
20:51 * kentnl waits for you to proceed to tell me you don't know how to use a keyboard :p
20:51 lbrown but
20:51 lbrown the arrays
20:51 lbrown the go
20:51 lbrown ahhhh
20:51 kentnl yeh, the arrayashes. Those abominations.
20:52 lbrown that's what we call them?
20:52 kentnl protip:  $x + $y in PHP is apropos of  { %{$y}, %{x} }
20:53 kentnl and array_merge and array_merge_recursive should only be used if your arrayashes are intended to be arrays, not intended to be hashes
20:53 lbrown why
20:53 lbrown just why
20:53 kentnl magical things happen if your keys happen to look enough like numbers ;)
20:54 kentnl php will decide to renumber them for you
20:54 lbrown oh yeah
20:54 lbrown thats why we need sort(), ksort() and asort()
20:54 lbrown and a bunch of hacks I found on various websites
20:55 kentnl comments section for array_merge_recursive: every man, woman, child, and their dogs and cats implementing array_merge_recursive from scratch not to be brain damaged
20:55 lbrown so just acking through some PHP repositories I found so many people looping over directories in what even PHP would consider to be "the wrong way"
20:56 lbrown so if anyone names a file 'false' or '0' the loop will stop ;P
20:56 vroom joined #pr-challenge
20:59 kentnl oh, and of course, there is no recursive version of "+". So you'll find a home brewed alternative somewhere in your code if you look hard enough
21:02 lbrown I'm not sure I want to
21:02 lbrown *the* PHP library for openID is doing this wrong https://github.com/openid/php-openid/pull/122
21:08 lbrown what on earth is libnosefart.la
21:12 jrhunt nintendo music player?
21:15 lbrown quite possibly
21:15 lbrown it came with libxine
21:29 athos joined #pr-challenge
21:31 choroba joined #pr-challenge
21:45 vairav joined #pr-challenge
21:47 vairav joined #pr-challenge
22:06 BRAD joined #pr-challenge
22:27 vroom joined #pr-challenge
22:34 mstr joined #pr-challenge
22:35 nbezzala joined #pr-challenge
22:56 Su-Shee how many people signed up by now?
22:57 neilb 257 people have had an assignment
22:59 sawyer damn...
23:00 Su-Shee neilb: TWO HUNDREDS AND FIFTY?
23:01 Su-Shee neilb: that's almost a yapc::eu!
23:01 BHS_error wow.
23:01 neilb I was hoping for double figures the first time I announced it :-)
23:02 popl neilb: How many people left to assign things to this month?
23:02 sawyer really?
23:02 sawyer oh i misread that nevermind :)
23:03 neilb The signups have finally slowed to a dribble. I’m caught up, so 257 signups, and they’ve all got an assignment. Only 3 re-assignments so far, which is better than I expected as well.
23:03 Su-Shee neilb: for 500 people?! I guessed maybe 30 and if it would turn out well, 50.
23:03 popl neilb: If you've
23:03 popl ugh
23:04 popl neilb: If you've assigned me a project I must've missed it in my email.
23:04 popl Let me double-check.
23:04 neilb I expected 5 to 10 people would signup, and hoped I might hit 15.
23:04 Su-Shee neilb: curious how many are really doing it. my estimate would be 25%.
23:04 neilb popl: check your spam box. But i’ll go look it up
23:04 popl neilb: Did you send assignments via your neil@bowers.com?
23:04 popl *address
23:04 Mithaldu neilb: while you're here, how should people confirm task completion?
23:05 popl goddamnit
23:05 popl neilb: don't bother
23:05 BHS_error I guess eventually there will be like 10 to 15 people really doing the assignment the first month
23:05 popl WTF Google
23:05 popl seriously, I am subscribed to the group and am filtering emails. Why is it in Spam?
23:06 Su-Shee I got mine to my google address
23:06 popl Su-Shee: It wasn't filtered into Spam?
23:06 kentnl gmail, the email service I love to hate, but am stuck with it because everything else I tried and was free I hated more.
23:06 Su-Shee popl: no.
23:06 popl holy crap
23:06 neilb ah, but the group is different from me, so not surprised. A few people have emailed asking WHERE THE FUCK IS MY ASSIGNMENT, but it turned out to be in ther spam box.
23:06 popl my gripes against Gmail keep growing.
23:06 Su-Shee popl: I almost never have false positives
23:07 neilb Hey popl, you got Module-CPANTS-Analyse — nice!
23:07 kentnl lucky popl.
23:07 popl Oh hey, my email from FedEx saying my laptop was going to be late is in here too.
23:08 popl fscking hell
23:08 neilb hi kentnl: I’m afraid it might be a while until I get to your pull request on Module-Path, sorry
23:08 Su-Shee kentnl: what did you get? :)
23:08 kentnl I didn't apply for anything. I'd rather be lazy and poke everyone else with a stick and be sub-useful in that manner.
23:09 neilb kentnl’s just kibbitzing :-)
23:09 Su-Shee haha :)
23:09 Su-Shee well I hate my assignment it's gonna be awesome :)
23:09 popl Why am I lucky?
23:09 neilb Hey kentnl, here’s a project for you: create a module for identifying whether a dist needs a compiler or not.
23:09 Su-Shee popl: e.g. upasana has suddenly to deal with SSL ;)
23:10 neilb Look at the outstanding issues for MetaCPAN and talk to oalders.
23:10 kentnl popl: because you have to find a quality issue of sorts with a module that focuses on quality issues :D
23:10 neilb Could really do with that being done before the end of this month, so I can use for Feb’s assignments ;-)
23:10 popl hah
23:11 kentnl neilb: that sounds like it would be heuristic based and not determinate. I have a bad history with that realm of things *coughcough*certainmodulesofmineiletrot*coughcough*
23:14 kentnl neilb: for instance, I know *I* can tell you this module needs a C Compiler, but I'm not sure how you'd determine it statically: https://metacpan.org/source/FRACTAL/Unicode-Truncate-0.102
23:15 kentnl no xs, no .h, no .c
23:15 neilb See my question in #toolchain
23:15 neilb ah
23:17 Su-Shee neilb: if we have modules which all in all simple haven't been updated for longer because they're basically finished - do you want to know for your list?
23:19 Su-Shee "simply"
23:19 neilb yeah, let me know those, so I can see how my selection heuristic does on them.
23:21 vairav joined #pr-challenge
23:23 Su-Shee awesome. mine has two issues I probably can actually fix :)
23:27 neilb Su-Shee: \o/
23:27 neilb s/probably//
23:28 Su-Shee well I'm careful, one is a speed thingie, a simple change might just move the problem around a bit..
23:31 neilb Devel::NYTProf is your friend
23:31 Su-Shee true..
23:32 Su-Shee well, I will set up a perlbrew with half a dozen perls or so for starters..
23:34 jashank ether++
23:37 Su-Shee challenge for 2016: write a range of standard software in perl.. like.. for blogs. ;)
23:39 sue joined #pr-challenge
23:41 * kentnl has PTSD style flashback-visions of hundreds of people writing html/csv/irc parsers
23:43 Su-Shee kentnl: well at some point somebody has to write contemporary applications in perl.
23:43 BHS_error Su-Shee++
23:43 Su-Shee kentnl: or blogs.perl.org runs on .. PHP some day.. ;)
23:43 Su-Shee ok, probably node.js ;)
23:43 BHS_error ehe
23:43 vroom joined #pr-challenge
23:45 kentnl http://ynea.futureware.at/cgi-bin/infinite_screen.pl # A gift for windows users.
23:46 kentnl if you don't quite get what it does soon after running, simply hold ctrl+shift and move your mouse.
23:46 kentnl all will become apparent :)
23:47 vairav joined #pr-challenge
23:47 kentnl the code is pretty nasty and its implemented in a self-extracting-archive-containing-win32-perl5.8 but its pretty marvelous in terms of what it does

| Channels | #pr-challenge index | Today | | Search | Google Search | Plain-Text | summary