Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2014-02-10

| Channels | #metacpan index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:21 mattp joined #metacpan
02:17 klapperl_ joined #metacpan
02:46 oalders for pod previewing, there's also http://podwebview.odyniec.net/  haven't used it myself, but maybe we should link to it
02:46 dipsy [ POD Web View ]
04:26 jwang joined #metacpan
05:00 kentnl joined #metacpan
06:32 [Sno] joined #metacpan
07:16 ether joined #metacpan
07:45 neilb joined #metacpan
08:17 bowtie_ joined #metacpan
08:21 jibsheet joined #metacpan
08:38 ribasushi still 503's here and there :(
08:38 ribasushi https://travis-ci.org/dbsrgits/dbix-class/jobs/18564362#L148
08:38 dipsy [ Travis CI - Free Hosted Continuous Integration Platform for the Open Source Community ]
09:13 dpetrov_ joined #metacpan
09:24 ribasushi ranguard: ^^
09:24 ribasushi it's been going on for 30 mins+ now :(
09:25 ribasushi ranguard: basically it seems to hit me every time I try to run a travis smoke (and I am not *that* high-profile of a user)
09:25 ribasushi it is sporadic, combined with me doing sporadic dev => hard to pinpoint when it happens (also I do not have shell to the VMs in question to diagnose on the spot)
09:26 ribasushi ranguard: note - travis-ci *may* be at fault too
09:26 neilb joined #metacpan
09:28 ranguard ok, I'll try look at lunch (in a couple of hours) - thanks for letting me know
09:28 ranguard I can see something odd with SSL certs - so have replicated
09:28 ribasushi ranguard: np, I could also file bugs everytime I notice it, but figured IRC is sufficient
09:29 ribasushi it's a start that you have been considering it resolved ;)
09:30 ranguard yea, sorry I thought this was because of the starman service going down - but I think it's unreleated
09:59 ribasushi ranguard: just saw another one https://travis-ci.org/dbsrgits/dbix-class/jobs/18568502#L1229
09:59 dipsy [ Travis CI - Free Hosted Continuous Integration Platform for the Open Source Community ]
10:00 ribasushi ugh that link will disappear sorry didn't think
10:00 ranguard np - I can see it in the fastly error report
10:01 ranguard well, not that specificaly - but getting 503's
10:02 ribasushi aight
10:02 * ribasushi has to run, highlight me if you ever get to a resolution
10:02 ribasushi ranguard++ # debuggering
10:44 kentnl ++ to whoever put the anchor links on the side of the headings, I've been hacking around that with https://addons.mozilla.org/en-US/firefox/addon/show-anchors-2/  . Anchors Ftw
10:44 dipsy [ Show Anchors 2 :: Add-ons for Firefox ]
10:44 dipsy Thanks!
10:46 kentnl ribasushi: travis was giving me invalid-ssl-certs showing certs for completely unrelated websites earlier today, so I suspect DNS failery is responsible for anything ( ether was getting ssl certs for a _different_ domain to me )
10:49 kentnl and re: missing_modules, I *suspect* metacpan may correctly show things as long as you properly map file <-> modulenames via META.json's "provides": feature, but not sure, if somebody can confirm that hunch, it may be nice to document there too
10:53 Mithaldu why are all the code blocks red here? https://metacpan.org/pod/Mail::Folder::Mbox
10:53 dipsy [ Mail::Folder::Mbox - A Unix mbox interface for Mail::Folder. - metacpan.org ]
10:54 Mithaldu oh god all inline code segments are red now, who thought that was a good idea? :(
10:59 kentnl it was better than that pale unreadable blue :)
11:02 Mithaldu ah, the classic school of fixing an issue by creating another issue
11:04 kentnl in this case its preferable, mostly because the style is /also/ a convention I've seen adopted many other places.
11:04 Mithaldu how about *black*
11:05 Mithaldu i know, it's a novel idea and all
11:05 Mithaldu and that means that there's only two clear elements to distinguish inline code
11:07 kentnl alternatively, accept you can't please everyone regardless of what theme you choose, and add theme selection options =)
11:10 kentnl oh. Wait. It _was_ black.
11:10 kentnl http://web.archive.org/web/20130908164723/https://metacpan.org/module/Class::Load
11:10 dipsy urgh. long url. Try http://tinyurl.com/paunyf4
11:10 dipsy [ Class::Load - a working (require "Class::Name") and more - metacpan.org - Perl programming language ]
11:11 Mithaldu kentl i was thinking something along those lines
11:11 Mithaldu kentaro: and yeah, that's kinda my current beef, ton of changes, no discussion ahead of time, no explanation why some changes were made
11:12 kentnl I guess if you really wanted you could use userstyles.js, but thats too complicated for me most of the time ;)
11:12 Mithaldu actually, i will and i already have one for metacpan
11:12 Mithaldu because the whole "make lines span the entire fucking half meter of my screen" thing was entirely useless
11:13 kentnl I "get" changes, and I "get" that most people seem to oppose many changes almost arbitrarily, "change is hard" seems to be the crux of it. I try not to let it get to me =)
11:14 Mithaldu (heh, just checked in FF and the line bit isn't fixed either)
11:14 Mithaldu i don't complain about changes, i complain about changes that have objective impacts
11:14 kentnl And good docs are hard :(, I feel POD in general needs a massive overhaul in how it works, just I have no idea how
11:15 kentnl ( actually, I have /some/ ideas, which is what my MetaPOD project is for, but that will only, maybe, solve some of the issues )
11:16 neilb I find myself just wanting to document my modules with markdown instead of pod
11:16 ribasushi kentnl: your explanation does not account for 503s (as opposed to 404s) and to error entries in ranguard's logs ;)
11:16 neilb With L<>, I guess
11:17 ribasushi is markdown really much better than pod?
11:17 kentnl I don't think markdown solves anything myself. It doesn't really solve the problems POD has with indexability, wide formats, and cross referencing and inheritance-awareness.
11:17 neilb I don't know about better, just this person finds it less frustrating
11:17 ribasushi the stuff I miss in POD are: tables, fottnotes, proper arbotrary link anchors
11:17 ribasushi markdown (nor any other competitor I looked at) takes care of any of these
11:18 ribasushi neilb: "better" as in "technically superior"
11:18 Mithaldu weird that the search box isn't centered anymore
11:18 neilb :-p
11:18 Mithaldu kentnl: can you guess what the objective complaint about not centering the search box is? :P
11:18 kentnl Wikidocs are not much better, they get around the lack of features by having a language for a lack of features which is more or less equating to "write html in another syntax"
11:19 kentnl Mithaldu: screenoverflow?
11:20 Mithaldu kentnl: usability
11:20 kentnl hah, if you make your screen narrow enough, the search box disappears entirely
11:20 Mithaldu assuming the default position of the mouse while browsing is the middle of the window
11:20 Mithaldu accessing the search box before was "move mouse up"
11:20 Mithaldu now it's "move mouse up AND left"
11:20 ribasushi kentnl: so... you are not as much talking about "fix pod", but "fix lack of a sane TeX-like system" ? :)
11:22 kentnl I think, yeh, perhaps, POD is ok-ish for various things, but its really _only_ suited to documenting the facts, but entirely useless at documenting the cross-relationships between facts in a way computers can use.
11:23 ribasushi kentnl: what I am trying to get at is: "Is there a sane pod/markdown/whathaveyou alternative that I missed while researching this a year ago"
11:23 Mithaldu there's some dzil thing that lets you write markdown :v
11:24 ribasushi Mithaldu: which is great except we already established markdown does not solve any of the problems :)
11:25 Mithaldu ribasushi: i was joking, if i see people do that i consider writing strongly worded letters
11:26 Mithaldu oh good, trying to find a commit that removed the dist version menu and lo, the history looks like a christmas tree :(
11:27 ribasushi Mithaldu: why? if I do find a system which is not html and which does what I need from it - I will stop writing pod altogether in-repo, and will have a process lossy-convert stuff to pod before it gets to CPAN
11:27 ribasushi the problem with pod (among other things) is that there is nothing to replace it witrh
11:27 kentnl I don't think any "markup" system really solves this problem. However, various languages have integral tools that unify markup with structure via reflection during documentation generation
11:28 ribasushi Mithaldu: I am now curious what the "strongly worded letter" would contain ;)
11:29 Mithaldu ribasushi: that was code for "i might be unhappy"
11:29 ribasushi Mithaldu: why would you be unhappy if I write my docs in non-pod, but ship pod in the end
11:29 kentnl and because we can't really expect people to ship pre-rendered html, and we don't really have any reliable reflection system that isn't entirely oriented around a given toolkit.... we have fun.
11:29 ribasushi Mithaldu: who are you to care? :)
11:29 Mithaldu ribasushi: i guess it's partly because the only examples of people doing that resulted in docs that were unreadable in the code file
11:30 kentnl hence, MetaPOD is "assume some tool exists to provide reflection data, and store that data in JSON for something to do something useful with one day"
11:30 ribasushi <kentnl> and because we can't really expect people to ship pre-rendered html <--- this isn't entirely true - you can certainly try ;)
11:30 kentnl the problem there is it just doesn't integrate with any of the present CPAN stack
11:30 ribasushi kentnl: and strongarm metacpan into displaying them properly
11:31 ribasushi Mithaldu: shrug - you don't get to rely on docs in-code
11:31 kentnl like, the HTML would be inherently flimsy, the cross links would be dodgy, and you'd be making end users decide how to format data that shouldnt' be their decision
11:31 ribasushi nod
11:31 Mithaldu ribasushi: i consider it anti-social behavior
11:32 kentnl so thats why I MetaPOD, bake the data required to generate structure in places, so that proper tools can do proper things without having to evaluate perl code
11:32 Mithaldu one of the niceties of perl is the fact that docs are readable in code
11:32 Mithaldu and well, as long as it remains readable, i probably don't care
11:32 kentnl ( and the generation of that data is left to either HAND or end users using the right tool for their code )
11:33 ribasushi Mithaldu: but besides "reasable" you seem to imply that pod should happen around the methods it talks about, and be in the same file, etc etc etc
11:33 ribasushi in essence you are putting the onus of a "convention" on the author, when there actully isn't one
11:34 Mithaldu ribasushi: you're talking complicated, i've no idea what you're trying to say
11:35 ribasushi I'll uncomplicate it: "POD was ever only meant to be shipped in a CPAN dist, so it is converted to HTML and read in a browser. How the POD got into the CPAN tarball is not something Mithaldu or anyone else gets a say over"
11:37 Mithaldu i can only imagine you saying that from the point of view of a code author who doesn't need to often read the docs of their own code
11:37 kentnl there's benefits from keeping source <-> release close together.
11:37 ribasushi oh absolutely
11:38 kentnl mostly, "people finding quirsk in shipped code don't spend several hours trying to find out where to patch it"
11:38 ribasushi except "close" is in the eye of the beholder - i.e. what works for the author
11:38 kentnl I was trying to understand some ruby code for travis's backend today and my head just hurt because of so much magic that was impossible to resolve by simply skimming the repo / using grep
11:38 kentnl travis's api*
11:40 kentnl you do have to somewhat anticipate people contributing, its desirable. Its not your primary objective in writing code, but its important if you ever want it to go somewhere without you holding its hand =)
11:40 Mithaldu ribasushi: so would you have a reason to keep your documentation away from your code in your repo?
11:40 Mithaldu or are you arguing on principle?
11:41 kentnl imo, it really depends on the documentation type there too
11:41 kentnl some documentation has _no_ place being beside code
11:41 ribasushi Mithaldu: I am arguing on experience - in DBIC there is a *lot* of documentation like that, and it 1) makes sense in terms of how it got where it is 2) hard to "fix" due to what kentnl said
11:41 kentnl for instance, Moose::Manual::* should of course, not be baked in to moose in the relevant places it applies to.
11:41 Mithaldu ribasushi: examples?
11:41 ribasushi Mithaldu: you should have experienced this - belongs_to etc - they are declared in one place, actually used in another, documented in a third
11:42 ribasushi oh and the attrs to them I *think* are documented in a 4th place
11:42 kentnl ^ thats kindof a problem due to cross-referencing
11:42 Mithaldu eh, i find that's mostly caused by not having the correct pointers to the correct parts of documentation
11:42 Mithaldu kentl: the problem is LACK of cross-referencing
11:43 kentnl a good toolkit would bake that data in at the places that was relevant, regardless of where the documentation was defined
11:43 Mithaldu example:
11:43 kentnl we just don't have that toolkit
11:43 Mithaldu PPI has in every class a class hierarchy list
11:43 Mithaldu which is useless, since that means you have to click through to find the methods
11:44 kentnl yeh, you often need the same documentation on both sides of the relationship, represented as inverses of each other. Doing that by hand is going to be painful for everyone
11:44 ribasushi kentnl: DBIC kinda started on this, with pretty good result, but we never took it further (tuits): https://metacpan.org/source/RIBASUSHI/DBIx-Class-0.08270/maint/gen_pod_inherit#L28 resulting in https://metacpan.org/pod/release/RIBASUSHI/DBIx-Class-0.08270/lib/DBIx/Class/Manual/ResultClass.pod#INHERITED-METHODS
11:44 dipsy [ maint/gen_pod_inherit - metacpan.org ] [ DBIx::Class::Manual::ResultClass - Representing a single result (row) from a DB query - metacpan.org ]
11:45 kentnl imo, the tuits problem there is caused by having to document the relations independent of where they occur
11:46 Mithaldu anyhow, holy shit, getting the version dropdown back is going to be painful
11:46 Mithaldu since it's entirely unobvious how it got lost in the first place
11:46 kentnl if instead each class just said  "{ namespace => "foo", inherits => "bar" } , you could statically extract the relationship model from the places that was relevant
11:47 ribasushi kentnl: I am a tad confused - what do you mean?
11:48 ribasushi by tuits I meant - to expand it beyond just a link and a name, but copy some of the docs over as well
11:48 kentnl I mean, if the relationships change, you have to update -that- file instead of the file the relationship changed in
11:48 kentnl every new relationship has to be added to that file
11:49 kentnl so its double-accounting, and nobody has that sort of time
11:50 ribasushi kentnl: ah that ... hm... good question because we also have https://metacpan.org/source/RIBASUSHI/DBIx-Class-0.08270/lib/DBIx/Class/Relationship/HasMany.pm#L9
11:50 dipsy [ lib/DBIx/Class/Relationship/HasMany.pm - metacpan.org ]
11:50 ribasushi kentnl: I am not sure why SineSwiper did duplicate things
11:50 ribasushi kentnl: well volunteered to dig into this a bit? :)))
11:51 kentnl yeah, having them in the files themselves and having a process to extract them is much more sensible an idea
11:51 ribasushi (by dig I mean figure out why SineSwiper abandoned the inlines, and went for a central repo instead)
11:52 kentnl https://metacpan.org/source/KENTNL/MetaPOD-0.3.6/lib/MetaPOD/Assembler.pm#L249  <-- if you've been observant, you may have noticed these sections hiding in my POD  =)
11:52 dipsy [ lib/MetaPOD/Assembler.pm - metacpan.org ]
11:53 haarg Mithaldu, i agree with you about the C<> formatting
11:53 haarg i'll probably submit a pr to change it to just be monospace
11:53 kentnl its not a complete solution yet, because I keep getting mindblow from anticipating how to improve on things in a way that is sane.
11:53 kentnl haarg: I'd keep the shaded box though, that seems better than _just_ monospace
11:53 haarg i disagree
11:55 kentnl I should toy with a metapod branch on dbix-class just to see what it gives.
11:56 Mithaldu haarg: cheers on that. :)
11:56 Mithaldu i personally don't mind the grey background, or grey box, but i can live without them too
11:56 kentnl one day somebody will make a strictures pragma called 'harden', and I can inject it in my code 'use harden'; && git commit -m "MetaPOD now uses harden. It is super effective'
11:56 Mithaldu it's mostly the red text that's annoying because it parses as "built-in" on metacpan
12:01 kentnl currently roadblocked on trying to use the travis api. The docs are so poor.
12:01 kentnl for instance: has privacy grades .... but doesnt' explain at all how to pass auth tokens to see things from the private aspect.
12:02 kentnl and for instance, it limits the numbers of results from certain requests, but its undocumented how to exceed the default limits, or how to scroll data outside those limits
12:02 ribasushi kentnl: what are you thinking to use it for?
12:03 kentnl I'm wanting to have an automated log mirroring system because fetching, grepping, and diffing them on travis.ci is a nightmare
12:03 ribasushi kentnl: #travis on freenode are rather helpful (when you catch one of the 3 core devs that is)
12:03 ribasushi kentnl++ # I do want that newsletter
12:03 kentnl because my log data tell me "how many deps does X take on a fresh perl", and I'm wanting to compare those lists and their counts
12:04 ribasushi kentnl: as a bonus if you can suck down all ~15,000 reports DBIC produce so far - winning ;)
12:04 kentnl there's a doge meme just waiting to be made there
12:04 ribasushi MUCH DOGE!!!!
12:05 kentnl many references, very lols.
12:06 kentnl but to do what I want to do, I have to enumerate all the repos that currently exist on travis ... but I can only see the most recent 10
12:06 kentnl and I need to enumerate all jobs .... but I can only see the first 25
12:07 kentnl ( I found a way around the latter by sniffing the JSON requests on the website and seeing an undocumented parameter get passed that tells it what number to start at ... but its not very good )
12:08 jibsheet joined #metacpan
12:08 kentnl and of course, you can't actually request the metadata for a single job without fetching the whole log of that job .... inlined into a single field in the json itself.
12:09 kentnl pros: I don't have to do 2 requests to get that data, cons, ... I may have to do this many many times before I get my code right -_-
12:10 haarg joined #metacpan
12:10 rGeoffrey_zzz joined #metacpan
12:11 kentnl not sure why, these days I feel slightly apherensive about talking on freenode.
12:11 kentnl They're just not as cool as perl.org
12:11 ribasushi they are downright raging batshit assholes
12:12 ribasushi but shrug - such is the price for working with a non-dying community
12:13 Mithaldu in my experience freenode is either dead channels or channels with newbies who think they're rockstars
12:13 kentnl can't decide if a festering collection of moulds growing rampantly in a darkly lit corner of a room could be considered a "amazing growth to be proud of"
12:14 ribasushi Mithaldu: fuck off, I'm still a rockstar!!!
12:14 ribasushi (with apologies to Pink)
12:14 kentnl I have concluded, the cool people don't give a fuck if they're cool or not.
12:15 kentnl of course, there's invoking meta self-referncing there, but .... I don't really know anymore
12:15 kentnl If I said I don't care, it could seem I'm trying to be cool, or something like that.
12:16 ribasushi such a coolster
12:17 kentnl coolness = coolness ^ attempting
12:17 kentnl (where ^ is of course, xor )
12:18 Mithaldu ribasushi: do you even have the hair for that?
12:18 kentnl thus, I think if you were to try to be cool, by trying not to be cool, you wouldn't become cool as a result. My head hurts.
12:19 ribasushi Mithaldu: don't touch MY hair
12:19 haarg Mithaldu, kentnl: https://github.com/CPAN-API/metacpan-web/pull/1095
12:19 dipsy [ format <code> blocks as normal monospace text by haarg · Pull Request #1095 · CPAN-API/metacpan-web · GitHub ]
12:21 Mithaldu haarg: cheers :)
12:26 Mithaldu kentnl: the builtin thing isn't cultural
12:27 kentnl Actually, I think my opposition with black-on-white monospaced, is for some reason it renders slightly shorter than normal text for me. Not sure why.
12:27 Mithaldu full code blocks are highlighted with a perl syntax highlighter
12:27 Mithaldu and that marks builtins as red
12:27 Mithaldu if the inline ones are colorful my mind goes "hey, highlighted!"
12:27 kentnl "and experiential" <-- was an example of that.
12:28 Mithaldu kentnl: that's why you force metacpan to text-align: justify :)
12:28 kentnl no, hot horizonal alignment, *vertical size*
12:28 Mithaldu oic)
12:29 kentnl like the text is visually shorter than the surrounding text.
12:29 Mithaldu yeah, currently it's bigger
12:29 haarg kentnl: that's done explicitly
12:29 kentnl Captial C is smaller than a lower case d , which is just weird for me.
12:29 haarg <code> has font-size: 0.8em
12:30 haarg if you make it normal size, it usually ends up looking a lot larger than the surrounding text
12:30 haarg 0.9em may work better
12:31 kentnl Perhaps, its just in conjuction with only being a font difference give me the same effect I get when I ReAd WhEn PeOpLe ThInK tHeYrE sO cLeVeR aNd OrIgInAl bY wRiTiNg LiKe ThIs
12:33 haarg well feel free to comment on the pr
12:33 kentnl I guess the reason I don't get quite so bitten by that with the boxes is the box make up for the vertical space
12:36 kentnl I actually can't find the CSS rule in the browser thats doing it so I can see how it looks without, everything I change also changes the main font size :(
12:37 haarg you should just be able to add a font-size to '.pod code' or 'code'
12:40 kentnl the difference between 1.32em and 1.31 em is _massive_
12:41 haarg i also wonder about making <code> bold
12:42 haarg i can see what you mean about having it just be monospace looking a little odd.
12:42 haarg to my eyes, having it bold addresses that
12:43 kentnl also, browser weirdness,I had text scale on, with text scale on, the difference of 1.31 vs 1.32 em was huge
12:43 kentnl I disable text scale, and now I cant determine a difference between 1.30  and 1.40 em
12:43 kentnl and the text is not that different sizewise -_-
12:45 ribasushi kentnl: such css, so weird, much wtf, wow ?
12:46 kentnl ribasushi++
12:46 kentnl haarg: bold makes it worse to me...
12:46 kentnl but I'm assuming we may have different values of font for 'monospace'
12:46 Mithaldu (kentnl) the difference between 1.32em and 1.31 em is _massive_ <- wat, unless you've got a font size of 100 pixels it shouldn't make a difference
12:46 Mithaldu can you upload your example somewhere?
12:46 haarg kentnl: do you know what font you are getting?
12:47 Mithaldu (or are you looking at IE?)
12:47 kentnl I know right. But it was like, a visible step, for some reason, and now I've reset my zoom level, it no longer happens
12:47 Mithaldu oh, you're in a browser with shitty zoom
12:47 Mithaldu (i.e. not opera)
12:47 kentnl lol
12:47 kentnl haarg: says Andale Mono for me
12:48 haarg ah yeah that's ugly
12:49 haarg Courier is the only thing that looks ok bold actually
12:49 haarg so that's probably out
12:49 kentnl ok, also, weird, I changed a font size in my browser settings, and suddenly its all the much more fucked -_-
12:50 Mithaldu btw, if you suggest using a webfont i'll probably have to find you and stab you
12:50 kentnl Mithaldu: use a webfont
12:51 kentnl </tease>
12:51 Mithaldu kentnl: why did i know you wouldn't see the "probably"
12:51 haarg trying to match font sizes is rather problematic unfortunately, especially with fallbacks etc
12:53 kentnl I'm not sure what "em" means anymore, with my browser set the "default" size for monospace at 32px, '1.32em' is now twice as large as every other font on the page
12:54 haarg em is the font size
12:54 kentnl what the fuck is the point of that setting "let web pages choose their own values" if it aribitrarily fucks the values they set ?!
12:54 Mithaldu kentnl: that means your other fonts aren't defined in em
12:55 Mithaldu the trick with em is you need to have either everything or nothing in it
12:55 Mithaldu also
12:55 Mithaldu link your damn text page
12:55 kentnl text page?
12:55 Mithaldu test
12:55 kentnl I'm just dynamically eding the dom.
12:56 Mithaldu > living on the edge
12:56 Mithaldu > complaining that it cuts
12:57 haarg font-size: 1.32em means font-size: 132%
12:58 kentnl http://jsfiddle.net/PRY3S/  # the monospaced text is 3 times the size of the non-monospaced text, because of browser settings.
12:58 dipsy [ Edit fiddle - JSFiddle ]
12:59 kentnl ah ... so if the font-size at the browser is not constant ....
13:00 Mithaldu kentnl: yep
13:00 Mithaldu if you force a different font size for a certain kind of text, em will just multiply by that
13:01 kentnl I'm just specifying the /default/ size. Not the size pages should have as veto-powers that regulate all sizes -_-
13:02 kentnl oh, so you can fix it, as long as you specify an actual font size at a higher level. k
13:02 Mithaldu possibly
13:03 Mithaldu you're kind of on your own if you tell your browser to change the font size of specific types of text only
13:05 kentnl its just to me that made sense of "well, if they give no value, these should be the sensible default we use instead of rendering as 0pt or 90001pt", not "what ever you specify here will magically be multiplied by arbitary amounts with magical side effects nobody can anticipate"
13:06 ribasushi wow.
13:07 Mithaldu kentnl: em are actually really easy to understand :/
13:07 Mithaldu it's just that you started experimenting without reading the manual :)
13:08 kentnl there is no "manual" for magical fields in the browser that simply specify "here are the default font sizes"
13:09 Mithaldu now i'm confused if you're complaining about em or your browser
13:09 Mithaldu *whether
13:10 kentnl https://support.mozilla.org/en-US/kb/settings-fonts-languages-pop-ups-javascript?redirectlocale=en-US&amp;as=u&amp;redirectslug=Options+window+-+Content+panel&amp;utm_source=inproduct#w_fonts-dialog
13:10 dipsy [ Settings for fonts, languages, pop-ups, images and JavaScript | Firefox Help ]
13:10 kentnl I'm not the one writing the web page. I'm just using my browser. It really should be more apparent that these values are not merely "default sizes"
13:11 kentnl ie: I set the value to what I set it to, because it affects how unformatted plain text renders
13:11 kentnl that is: if somebody serves up a text/plain document
13:12 ribasushi kentnl => just a guy with a browser
13:12 kentnl lol.
13:12 kentnl I know, its not your fault. sorry for raging.
13:13 kentnl Just  I hate debugging things and finding the cause is something else and thats something else etc etc
13:13 ribasushi kentnl: yet... you work with perl... does not compute
13:13 haarg i'm not sure what you expect to happen.  the default font size is X, and someone specifies font-size: 150%;
13:14 haarg what else is it supposed to do except calculate the size as 150% of the default
13:14 kentnl I know. It makes sense when I have all the fact. Just I didn't have those when I made the decision which lead to my unfortunate font size experience  =)
13:15 kentnl I for some reason thought 'em' was a function of dpi, not sure where I have that memory from :/
13:16 ribasushi kentnl: pt is a function of dpi
13:16 haarg also i'm not sure where i got the 0.8 em thing from.  <code> is specified as 12px.
13:17 kentnl I'm guessing I got it glitching because fontscale may have meant <code> was being rendered at 11.4 px which was being roudned to 11px, and so 1.31em vs 1.32 em was enough to make it round the other side, or something.
13:18 Mithaldu since we're talking random bullshit
13:18 Mithaldu asked a dude to link his repo to a module
13:18 Mithaldu and got this answer:
13:18 Mithaldu > I "release often"
13:18 kentnl lol.
13:19 kentnl I filed a bug with incorrect metadata in META.json. It was closed as "not a bug"
13:19 ribasushi kentnl: link?
13:19 metacpan joined #metacpan
13:19 metacpan [metacpan-developer] mjemmeson opened pull request #11: updated vagrant download links (master...mjemmeson/updated-vagrant-links) http://git.io/qVyu3Q
13:19 metacpan left #metacpan
13:19 dipsy [ updated vagrant download links by mjemmeson · Pull Request #11 · CPAN-API/metacpan-developer · GitHub ]
13:19 haarg some people don't like their code repos to be public.  i don't really understand it.
13:20 metacpan joined #metacpan
13:20 metacpan [metacpan-developer] ranguard closed pull request #11: updated vagrant download links (master...mjemmeson/updated-vagrant-links) http://git.io/qVyu3Q
13:20 metacpan left #metacpan
13:20 dipsy [ updated vagrant download links by mjemmeson · Pull Request #11 · CPAN-API/metacpan-developer · GitHub ]
13:20 kentnl https://rt.cpan.org/Public/Bug/Display.html?id=83240
13:20 dipsy [ Bug #83240 for CPAN-ParseDistribution: META.yml lacks ABSTRACT and LICENSE ]
13:21 Mithaldu haarg: in this case the guy just doesn't understand what the value would be, since apparently all commits people give him need fixing either way
13:21 Mithaldu i guess he's unaware of how helpful history review can be in making good commits
13:22 Mithaldu then again, he does use a snowflake indenting style, so i shouldn't expect much
13:22 ribasushi erm no this is a bad analogy
13:22 Mithaldu ribasushi: which is "this"
13:22 Mithaldu ?
13:23 ribasushi *ALL* commits people given are invariably shit, and need fixing up, everyone knows that ;) so "if you open a repo, you will get good patches" is a fallacy
13:23 ribasushi (note - I do like open repos, but for quite different reasons)
13:23 kentnl worse, he then released the dist... and updated the repository data, but not the fields I asked for.
13:24 Mithaldu ribasushi: err, in this case not opening is is a guarantee for shit commit
13:24 haarg ribasushi: that isn't what Mithaldu said at all though
13:24 Mithaldu also
13:24 kentnl I think "writing code" is a guarantee of a shit commit =p
13:24 Mithaldu between shit and perfect there is a continuum
13:24 ribasushi I am saying that no matter what you do you'll get a shit commit, this is how humans are (both contributors and maints ;)
13:24 Mithaldu be less B&W
13:24 kentnl its just the significance of the shittyness that matters.
13:24 ribasushi Mithaldu: blashpemy - it's a binary thing ;)
13:25 haarg of course it's shit
13:25 haarg it's made out of software
13:25 Mithaldu ribasushi: you love to derail things for a living, huh?
13:25 * Mithaldu gives up on the conversation
13:25 kentnl billions of bugs, written by humans, stacked in commits, all over the repository. </channeling some idiot>
13:25 ribasushi Mithaldu: yes I do. also your last patch is shit!
13:26 * ribasushi exits stage left ;)
13:26 kentnl you don't really want to know who I was parodying >_>
13:27 Mithaldu for some reason i'm lately more allergic to channels that have a high occurence of making serious topics into attempts at mid-level jokes
13:28 kentnl For me, humour is how I deal with the reality made of shit.
13:28 ribasushi much reality, such shit, very emo, wow :(
13:28 kentnl Sad. But true. Its the easiest way to counter dumb shit. Turn it in to comedy.
13:29 kentnl Hence, why I was parodying a certain promintent creationist speaker who made me bruise myself with facepalming
13:31 Mithaldu let me guess, you watched the entire nye/ham thing? :P
13:31 kentnl Yeah.
13:31 kentnl Blame Getty, that bastard.
13:31 Mithaldu i got to the point where he mentions the inventor of the MRI
13:31 ribasushi Getty->blame(shit)
13:31 Mithaldu looked it up on wikipedia
13:31 Mithaldu which says
13:31 kentnl yeah, I was also already aware of his discredited claim to fame.
13:32 haarg i really don't get the point of debate things like that.  they aren't going to convince anyone of anything.
13:32 Mithaldu "dude described a machine on paper, claims to have built it. but his design wasn't even similarly used anywhere else and others had built smaller mri machines before, so his was at best the first full-body one."
13:32 Mithaldu or in other words, he didn't get past 5 minutes without lying
13:33 kentnl ^
13:33 Mithaldu at which point there was no reason to watch any more
13:33 kentnl May as well say Leonardo Davinci in fact invented space travel.
13:33 ribasushi but.. he... did
13:33 kentnl And built the first working model and visited the moon
13:36 kentnl Lots of people who didn't grow up around religious nutbags have the luxury of not realising how fucking stupid they can be, and think "well, this seems like a great argument to me, nobody can dispute that".
13:36 kentnl Which completel founders because "motivated reasoning" is a thing.
13:37 Mithaldu kentnl: you know how there's people who're gay or straight and everything inbetween?
13:37 kentnl <sarcasm>No, why would I know anything about gender/sexuality being something that isn't strictly binary?</sarcasm>
13:38 Mithaldu kentnl: you aware that there's a second axis to that, on the far end of which settle people best classified as asexual?
13:38 kentnl Yep.
13:38 Mithaldu the same exists for religion :D
13:39 Mithaldu the entire county i come from is basically areligious
13:39 kentnl Yeah. Its just you have x: [ brains <-> no brains ] y: [ nonvocal <-> hypervocal ] z: [ decriptivisit <-> perscriptivist ]
13:40 Mithaldu it's often really fun to explain to americans when asked what my beliefs are that i don't have any
13:40 kentnl and the nobrains hypervocal perscriptivists of course stick up like dicks, dicks, everywhere.
13:40 Mithaldu hahaha
13:40 Mithaldu i like that
13:40 Mithaldu ps: prescriptivist
13:41 kentnl pre? I always muddle that.
13:41 Mithaldu like prescribing someone a medicine
13:41 kentnl for some reason I read it like "perscription" which is really "prescription" I think./
13:41 kentnl ^zaclky
13:42 Mithaldu it is
13:42 kentnl english and its rules, who has time for that shit .  XD
13:44 kentnl I'm somewhere in the middle of the de/pre line really. I try to get it right. And appreciate anyone who at least shows attempt to get it right.
13:46 kentnl People who don't try at all ( aka: 'u' and 'r' ) or people who go out of their way to be difficult ( i+z z0 @m@z1n9 huw w3 cn rd dis  shit ) make me want to stab them, and hang out their entrails for public display
13:48 Mithaldu i got a simple trick for that
13:49 Mithaldu if i'm in a channel where they're asking for help
13:49 Mithaldu i tell them that english isn't my first language and that i have a hard time understanding them
13:50 haarg bleh CPAN::Changes is such garbage
13:50 kentnl Yeh. If they tell me they're just too lazy or its too hard to spell "you", I'm just going to say "If you dont give enough of a fuck to try make it so I can read it, I'm not going to give enough of a fuck to try reading it"
13:50 Mithaldu kentnl: too aggressive
13:50 Mithaldu i just say i don't understand what they're saying
13:50 Mithaldu end of
13:51 kentnl I have people who literally try justfiy being allowed to write like that, like its some sort of "Life choices" shit and that I should be expected to help them when they have NO INTEREST in helping themselves by helping me understand them
13:52 ribasushi kentnl: Y U NO LET M3 WRAIGHT 1337?!
13:53 kentnl Especially when they write like ribasushi , and its obvious they're trying very hard to make it so I can't read it, because somehow thats "Cool"
13:53 kentnl I experience serious neurological pain even trying to write like that.
13:53 ribasushi whoops, meant to say wraightE, spelling is important, apologies
13:56 haarg ribasushi: taking a shot at fixing the Changes parsing, but ugh
13:56 ribasushi haarg++ # you are awesome
13:56 kentnl whats currently wrong with it?
13:56 haarg the 'spec' is useless
13:57 haarg and the module interface is broken
13:57 kentnl and it has no documentation of how to behave with regards to utf8 ¬_¬
13:58 ribasushi neilb: ^^ more folks are looking at the spec, iirc you had some work done around it already but got sidetracked...?
13:58 * kentnl occasionally has it chew nonbreaking-space characters in the wrapping logic as if they were plain spaces
14:00 * neilb is on the phone
14:00 ribasushi kentnl: honestly - 1st world problems
14:01 ribasushi I would be quite happy if it handled ascii sanely
14:03 kentnl wouldn't it be 3rd world problems?  After all, "ASCII" seems to be more "firstworld", no. =) ?
14:04 kentnl or at very least 2nd world
14:04 ribasushi there ought to be a standard that disallows non-english in any tech-related stuff
14:04 ribasushi this includes changefiles
14:04 kentnl I use → extensively
14:04 neilb I had a discussion offline with omega, and we were both keen to improve things. I'd heard that bricas isn't doing much Perl these days, so emailed him asking if he'd be open to giving co-maint. He'd been away, but I've now heard back and he's given both of us co-maint. Planning on distilling out problems and possible solutions, and was thinking I might work on it at the QA hackathon
14:04 ribasushi kentnl: for which you shall burn in hell
14:05 ribasushi haarg: ^^
14:05 kentnl and unicode nbsp for  (&nbsp;foo&nbsp)  because ( foo ) is pretty, (foo) is ugly, but ( foo ) being wrapped as [ "(" , "foo )"  ] is just horrible as fuck.
14:05 neilb since I figured I could at least have a good argument^Wdiscussion with you about it there :-)
14:07 ribasushi neilb: honestly I do not have *any* interest in the spec, my personal beef with it is that it 1) wastes other developers time by going "hey hey convert to me, I am awesome" and 2) makes changelogs look like crap 2a) makes responsible developers have to decide whether to reject pull requests attempting to do so
14:08 ribasushi neilb: as far as improving/discussing - I am firmly in the camp "changelogs are written for humans by humans, having them machine parseable is utterly wrong"
14:08 ribasushi hence... don't have much to add on the tech. level
14:08 neilb hey, I wasn't trying to start the arguments early!
14:08 kentnl I'm very much in the fan camp. I publish CPAN::Changes data reserialised out =p
14:08 ribasushi neilb: this is the extent of my argument, I am just pointing out there isn't much there
14:09 ranguard ribasushi: I haven't been able to dig into cpan.mc.org much - might be a couple of days before I can more - please open an issue and I'll try track more after that
14:09 kentnl Though. it _can_ be useful for machines. I have machine generated segments that are injected.
14:09 kentnl and sometimes I generate them wrong, and they're easy to fix retroactively after the fact
14:10 kentnl and saying "machine parsability is wrong" is a very bold claim, because machine parsability also infers logical consistency
14:10 kentnl and having dealt with changelogs in the past with inconsistent formatting, I've had my share of issues ;)
14:11 kentnl I'll agree that the content of them should be human written
14:11 kentnl but the structure really aught to be regulated.
14:12 kentnl like, but not the same as commits. Commit _messages_ should be written by humans. But the commits themselves should be structured and ordered by machines. ( Granted, commits _must_ have this structure to work at all, so its not a fair-footed comparison, just an analogy )
14:13 kentnl But I do try to put certain segments in my CPAN::Changes file so that if people decide they dont want to see them, they can easily strip them out
14:14 kentnl or  so that if people decide the data in them is useful to their process, they can automatedly scrape and process that data
14:15 kentnl for instance, "give me all the changes between 1.5 and 2.0 so I can read them" is useful, as is being able to parse the structure so you can present it in a form more to your liking
14:15 michael hi, am trying to follow the metacpan-developer README, is there a problem with running he installdeps bit on the vagrant machine? https://gist.github.com/mjemmeson/8916546 - looks like it's unhappy with the bash eval line
14:15 dipsy [ metcpan-developer - install_modules --installdeps error? ]
14:17 ribasushi kentnl: https://metacpan.org/diff/file?target=RIBASUSHI/DBIx-Class-0.08270/&amp;source=RIBASUSHI/DBIx-Class-0.08210/#Changes
14:17 dipsy [ Search the CPAN - metacpan.org ]
14:17 ribasushi kentnl: you need a spec for this why exactly...?
14:17 kentnl Because when you have a tarball, you don't have 2 tarballs.
14:17 kentnl and if you don't have one of the tarballs ...
14:17 ribasushi ENONPARSEABLE
14:18 ribasushi I still don't get it...
14:18 neilb kentnl: yup. my philosophy is that the minimal spec / convention should be very small and simple, and something it would be hard to argue against. Everything else should be optional progressive refinement. Basically each release should have the 'header line' identifying version and date, and optionally other stuff. And then the body of each release is a chunk of text, which may or may not be formatted according to other
14:18 neilb conventions. This would mean a minimally formatted changelog can still be chunked as per your "gimme all the changes between X and Y example"
14:19 neilb So something like MetaCPAN can present most recent changes for a dist, and a /recent page could show you the extract for that particular release
14:19 ribasushi neilb: I could live with a "the version line must be X, and this line delineates chunks of text", though even this is overboard
14:20 ribasushi you can always look at CPAN/BACKPAN and get the dates, in fact the PAUSE-side-tarball timestamp is the *ONLY* thing that is authoritative
14:20 kentnl ribasushi: and sometimes diffs will show you things you don't want to see, especially if the changelog itself had retroactive changes added, the question is, do you want to see those retroactive changes, or just the ones that have been documented since $LASTREV?
14:20 ribasushi what date someone put in the changelog either manually or by an authoring tool is irrelevant as opposed to when it actually hit the index
14:21 kentnl looking at CPAN/BACKPAN is a lot of extra work, especially if you ever want to automate things
14:21 ribasushi kentnl: this again boils down to "learn how to use a diff tool / select chunks of interest"
14:22 kentnl And again, the diff tool will be _worthless_ if you only have 1 of the two tarballs, or are working on a tool where you are _only_ able to have one tarballs
14:22 ribasushi kentnl: also - how exactly are you seeing a changelog browser that will *not* be hitting cpan/backpan (via metacpan or some other method)
14:22 ribasushi I mean - how do you get a diff if you have "one tarball", there is something you are not mentioning here...
14:22 kentnl you don't get a diff.
14:23 kentnl Because you don't want a diff
14:23 kentnl you only want a list of the documented changes between x and y
14:23 kentnl and those are in the changelog
14:23 kentnl automatedly comparing _diffs_ and divining meaning from them without a human is something I don't consider possible :)
14:23 ribasushi and how does this combat: "if the changelog itself had retroactive changes added" ?
14:24 kentnl Because your solution of "use a diff" would reveal _additonal_ data the user was not interested in.
14:24 ribasushi when I say diffs in this context I mean "./Changelog diffs"
14:25 kentnl for instance, you'll have to mentally filter out any time somebody decided to change line wrapping, or change ordering, or .... any number of things that happen when authors do things without a standard of consistency that prevents them from making stupid changes :)
14:26 ribasushi <ribasushi> kentnl: this again boils down to "learn how to use a diff tool / select chunks of interest"  <--- see last part
14:26 kentnl But when you're wanting to list the output somewhere other than a terminal, say, you want to render it as a web page ...
14:27 ribasushi kentnl: anyway - it is clear you see much more utility to your proposed tool than I do, and believe a spec is the way to achieve that
14:27 kentnl I agree on that much.
14:27 kentnl I get a lot of utility out of some things other people don't touch with a barge pole =)
14:27 ribasushi there's nothing wrong with that, except I am up-front stating "I ain't playing" (unless the spec is relaxed enough that I don't have much to do to "play" )
14:27 kentnl like commitbuilds, and sterile perl tests on travis >:)
14:28 ribasushi shrug - I had sterile tests on travis before you even started playing with it :D
14:28 kentnl yeah, but yourself and myself have that in common
14:28 kentnl but most people don't do that
14:28 kentnl and don't see the point
14:29 ribasushi I see much point, and am actively looking on how to make it easy for more people to do that
14:29 kentnl exactly. =)
14:29 ribasushi it is important to counter the high-speed iteration perl is subjected to lately
14:29 kentnl I'll agree as much that nobody should be told "Make it CPAN::Changes compatible, or else"
14:30 kentnl but all the same, I try to make my data I provide as useful as possible, even if there are 0 people consuming that data or attempting to utilize it.
14:30 kentnl its there if anyone ever finds a need
14:31 ribasushi the thing is - I strongly believe my (non-current-spec compliant) changelogs are *more* useful than many others which are in fact compliant
14:31 kentnl though I'd prefer a "Be this grade of CPAN::Changes and comply with it fully" or a "don't care about it all" status.
14:31 ribasushi hence the pushback against what to me amounts to mindless busywork
14:31 kentnl having lots of intermediate variations on the spec will bring us the nightmare of the inconsistencies of the markdown spec
14:32 ribasushi nod, which is why I was pissed at bricas for pimping a clearly unfinished spec
14:33 * ribasushi likes when he loudly agrees with kentnl
14:33 kentnl maybe, I could be argued to see value in a CPAN::Changes::Alternative system, where you put some hints in the top of your changes file to say "Im not spec comformant, and here is how nonconformant I am" so that a transformer can arbitrarily parse your changefile and emit CPAN::Changes
14:33 ribasushi kentnl: x_changes_spec => {} ;)
14:33 kentnl I remember the good old days where we never agreed on anything. What happened to those =P
14:34 neilb I started thinking about that, then decided to see if I could do something with CPAN::Changes (first)...
14:34 BinGOs We stopped listening to RATM
14:34 ribasushi kentnl: indeed
14:37 kentnl x_changes_spec => { format => 'sqlite' ... }
14:37 * kentnl kids
14:39 kentnl Maybe we could get a "I have my changes emitted from `git log` because I hate people who read my changes file" flavoured transformer :p
14:40 neilb I've seen some of those!
14:43 kentnl I just idly had a thought: "Yak shaving: The primary method of yak reproduction. Shed wool sprouts its own yaks"
14:43 Mithaldu kentnl, ribasushi: https://gist.github.com/wchristian/ae61deb94e66dcc6fc54
14:43 dipsy [ gist:ae61deb94e66dcc6fc54 ]
14:45 kentnl but but but, at least the people would see each others attempts and not duplicate them, and somebody might spot what they all did wrong.
14:45 Mithaldu yep
14:46 kentnl the solution here, is go "Fuck it", and use git-cpan-patch or whatever, reify a git repo from the CPAN releases, and pretend you're somebody who cares
14:46 Mithaldu the main problem is
14:46 Mithaldu the releases don't have enough granularity to do proper sleuthing
14:47 kentnl are there patches in rt?
14:47 Mithaldu anyhow, i'll just make a repo from the releases and work off that
14:47 Mithaldu i didn't come here to argue with him on teamwork, but to fix broken shit
14:47 kentnl At very least, $^O exists, and you could probably have used that value to prevent it breaking both places ...
14:48 Mithaldu he does that
14:48 Mithaldu it's why it breaks
14:48 Mithaldu let me show you the code
14:48 kentnl *furrow*
14:49 Mithaldu https://metacpan.org/source/MARKOV/Mail-Box-2.111/lib/Mail/Message/Body/File.pm#L136
14:49 dipsy [ lib/Mail/Message/Body/File.pm - metacpan.org ]
14:49 Mithaldu vs
14:49 Mithaldu https://metacpan.org/source/MARKOV/Mail-Box-2.111/lib/Mail/Message/Body/File.pm#L153
14:49 dipsy [ lib/Mail/Message/Body/File.pm - metacpan.org ]
14:50 Mithaldu oh god, i just noticed the "local *IN;" stuff
14:51 haarg does -s not return the real file size on windows?  or is that whole thing just garbage?
14:52 Mithaldu -s returns byte size
14:52 Mithaldu i assume size is meant to return the *message size*
14:53 Mithaldu https://metacpan.org/pod/Mail::Message::Body#obj-size
14:53 dipsy [ Mail::Message::Body - metacpan.org ]
14:53 Mithaldu whatever that means
14:54 Mithaldu oh, he also says:
14:54 Mithaldu "For many purposes, for instance to calculate the Content-Length, I need the real number of bytes not the platform native number."
14:55 kentnl acutally, I recall reading something that recommended localising *IN, might have been in Capture::Tiny
14:55 kentnl well, at least, STDIN
14:55 Mithaldu i'm p sure he does it because it's really old code and was meant to work in perl versions without lexical filehandles
14:55 kentnl no idea what *IN is
14:56 kentnl yeah. Ugh. I don't want to remember why that exists.
14:56 kentnl but but but
14:56 kentnl it has "my "
14:56 kentnl in the same scope
14:57 haarg my predates lexical fine handles
14:57 haarg file
14:57 kentnl oh, right. My bad, conflating the issues.
15:00 kentnl I think there's a level people get to in IT sometimes where things about IT that we take for granted get cut away due to being considered non-critical to their goals
15:00 kentnl like for instance, most people would be aghast that somebody had no repo, _and_ no email
15:00 kentnl but try filing a bug with Knuth, I recall it takes surface mail
15:01 kentnl maybe a public SCM is one of those things, maybe his SCM is not publicisable and the inertia required to change is too much for them
15:01 Mithaldu i can't get over this: "When you install the module under windows, it clearly warns that some tests break."
15:02 kentnl yah, "warns that tests break". Wut.
15:03 kentnl oh, I see
15:03 kentnl * On *Windows* you will get a large number of failing tests, but they
15:03 kentnl * are usually harmless.  Please help me to get rit of them.
15:04 kentnl http://www.cpantesters.org/cpan/report/3a0877d6-75f5-1014-b920-79d03567e6ce
15:04 dipsy [ CPAN Testers Reports: Report for Mail-Box-2.111 ]
15:04 Mithaldu kentnl: yeah, wow, that message scrolls by so fast it's amazing
15:04 Mithaldu i'm afraid to tell him though
15:04 Mithaldu he might put a <> or a sleep there
15:05 kentnl I'd go one better
15:05 kentnl I'd say you cannot rely on a warning in installation to be seen, at all.
15:05 kentnl because installation may not even be performed by a human
15:05 Mithaldu he'll argue i was clearly installing it as a human
15:05 kentnl In this case, yes.
15:05 kentnl But not in all cases.
15:06 kentnl Many cases will just run batch installs, and a human will triage the failures later, sometimes
15:06 kentnl chances are you'll see the warning in triage
15:07 kentnl ... assuming it is in the triage log
15:07 kentnl I have been known to read my own test reports via the internet XD
15:08 Mithaldu :P
15:09 kentnl because ... I've had my log files set on fire about 30 seconds after submitting them to cpantesters at the end of my install run XD
15:10 kentnl it was a bug on my behalf, but still happened :)
16:05 haarg well i have a Changes parser thing that should roughly handle what i would consider common practice for change logs
16:06 haarg not really any sensible way to fit it into the CPAN::Changes api though
16:13 oalders we should really figure out the whole Changes thing
16:14 oalders there's a discussion here https://github.com/bricas/cpan-changes/pull/14
16:14 dipsy [ Add support for DBIx::Class "nested" Changes files by omega · Pull Request #14 · bricas/cpan-changes · GitHub ]
16:14 oalders but we're waiting for bricas to chime in
16:14 oalders haarg: neilb is going to be at the QA hackathon and he has some interest in this as well
16:14 oalders maybe we can sort it there at the latest
16:15 oalders maybe we just need a MetaCPAN::Changes
16:15 neilb oalders: forgot to tell you: bricas gave omega and me co-maint
16:15 haarg that's probably what i'm going to do next.  just so i can play with it some more.
16:15 oalders neilb: nice
16:16 haarg the problem is that a fix will require an api break in CPAN::Changes
16:16 neilb so been thinking that's one for the QAH as well
16:16 haarg of course, it might make sense for CPAN::Changes to be a 'spec following' module, and MetaCPAN::Changes meant to adapt to what is found in the wild
16:48 neilb haarg, oalders: added to the project list for the QAH
16:49 oalders neilb++
17:17 ranguard joined #metacpan
17:35 [Sno] joined #metacpan
17:54 neilb joined #metacpan
18:58 neilb joined #metacpan
19:45 grantm Many thanks for the new bookmarkable headings feature - haarg++
19:46 haarg it's something i've been wanting for a while now, finally got around to implementing it
19:47 jayallen joined #metacpan
19:59 timbunce_ joined #metacpan
20:00 mudler joined #metacpan
20:04 neilb joined #metacpan
20:31 rurban joined #metacpan
20:31 rurban look: http://metacpan.org/release/JIPIPAYO/0.02 sigh
20:31 dipsy [ 0.02 - Simple utility to find perl modules dependencies recursively in your project. - metacpan.org ]
20:39 cooper joined #metacpan
20:40 rwstauner http://cpan.metacpan.org/authors/id/J/JI/JIPIPAYO/
20:40 dipsy [ Index of /authors/id/J/JI/JIPIPAYO/ ]
20:41 rwstauner nice
20:41 rurban I sent him an email to fix it
20:41 rwstauner :-)
20:43 haarg RFC: https://github.com/haarg/metacpan-web/commit/3c1af1a8fffc7b92588697d7027787bed699d14a
20:43 dipsy [ improve issue links · 3c1af1a · haarg/metacpan-web · GitHub ]
20:45 haarg trying to improve the issue linking.  tries to determine if issues are referencing rt.perl.org/rt.cpan.org/github on a per link basis.
20:46 neilb joined #metacpan
20:53 jibsheet thank you for not trying to handle pre-rt.perl.org issue numbers
21:11 timbunce joined #metacpan
21:33 mattp joined #metacpan
21:44 rurban left #metacpan
22:10 ether ribasushi: side note.. I think "debugging" and "debuggering" are *slightly* different... ;)
22:10 ether not sure exactly what debuggering involves; perhaps some sort of horrible sexual orientation reconditioning therapy?
22:12 haarg thoroughly unbuggered
22:12 * ether recently learned that the father of one of my peers at the church I grew up in was gay, and was undergoing said theraphy for a long long time. :/
22:13 ether -h
22:13 rwstauner maybe it means "removing the use of the word 'bugger'"
23:31 ether_ joined #metacpan
23:53 neilb joined #metacpan

| Channels | #metacpan index | Today | | Search | Google Search | Plain-Text | summary