Perl 6 - the future is here, just unevenly distributed

IRC log for #rosettacode, 2011-02-10

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

All times shown according to UTC.

Time Nick Message
00:15 Polynomial joined #rosettacode
00:43 Mathnerd314 left #rosettacode
04:39 kpreid 79.169.23.202 seems to be fond of minimizing character count. I'm reverting most of his formatting changes, because AFAIK most people don't use that brace style, and one-space indent isn't a good idea.
04:39 kpreid And minifying CSS just isn't right.
04:40 kpreid He (er, or she) also moves variable decls to the top of blocks, which was reasonable for JS but he did it for C as well.
04:48 mwn3d_phone I've seen a couple anons doing style edits. Sometimes the noise is annoying. Depends on my mood.
04:51 kpreid this guy systematically went through all C/Java/JS/CSS, it looks like
04:51 kpreid mwn3d_phone: I'd appreciate at least one other person looking at what I'm doing and telling me whether it's good
04:52 kpreid I'm of the opinion that the style they edited them into is horrible and so should be reverted, but ...
04:54 kpreid they definitely understand and are working towards *correctness*, it's just the formatting
04:54 mwn3d_phone There were a few changes that were real code changes that didn't change the workings of the programs
04:55 mwn3d_phone There was at least one spat of ternary-operator-ifying
04:55 mwn3d_phone Technically correct but in general the ternary operator looks confusing
04:56 kpreid I agree with those, actually :)
04:56 kpreid I'm asking specifically about whether you agree or disagree that the code *formatting* changes (indent/braces) are for the worse.
04:56 mwn3d_phone Braces on new lines I don't agree with
04:57 mwn3d_phone Single space indents I don't agree with (at least three is good enough)
04:58 kpreid well, I've done a bunch
04:58 mwn3d_phone I'll check on it when I'm at a real computer. Mine is off for the night.
04:58 kpreid I was just doing RC change reviewing on 'couldn't sleep' grounds, so I think I'll see if ther are flames on my talk page tomorrow.
04:58 kpreid ...before doing the rest of them.
04:58 mwn3d_phone Heh good call
04:59 mwn3d_phone Id hate to have a style war
04:59 mwn3d_phone That'd be pretty unproductive
05:00 kpreid I always figured that it should be like WP's approach to US/UK spellings: Don't gratuitously change them; be consistent within an article.
05:01 mwn3d_phone Seems like an ok idea to me
05:02 mwn3d_phone Except coding style trancends articles. The J community even went so far as to have a house style page set up.
05:02 mwn3d_phone I think the ruling back then was to let the language user communities decide.
05:03 mwn3d_phone With anons coming in randomly that could get difficult...especially for the more popular languages.
06:32 Coderjoe C (unless using a relaxed-decl extension or C99) requires variable declaration at the top of a block
11:27 FireFly joined #rosettacode
11:46 BenBE left #rosettacode
11:51 BenBE joined #rosettacode
12:53 kpreid Coderjoe: C99 is not nonstandard
12:54 kpreid in fact, I think we should be using more C99 and letting the compilers catch up
12:54 kpreid http://kpreid.livejournal.com/34848.html
12:54 fedaykin "kpreid: Catching up to 1999"
13:08 MigoMipo joined #rosettacode
14:14 shortcircuit kpreid, mwn3d_phone: Given that there are a large number of style-specific edits of late, perhaps it would make sense to simply standardize on a style per language-in-question, and have something to point to.
14:15 shortcircuit Otherwise, every now and again there will be someone with the energy to do mass edits, but with a different style than present.
14:15 shortcircuit It might make sense to run languages through whatever 'lint' tool is available, as was done with pylint and the Python code a year or two ago.
14:39 kpreid shortcircuit: Having a standard style is good. But in the absence of one, I think the rule should be to not do mass style edits, unless there is community consensus.
14:39 kpreid Does that seem reasonable to you?
14:39 kpreid (I mean strictly formatting. If someone wants to go through and improve correctness, even in some not-100%-uncontroversial way, that wouldn't be prohibited.)
14:44 shortcircuit I don't like trying to prohibit any class of non-destructive edit; I don't think it's preemptively enforceable short of having some kind of wiki extension that forces an editor cool-down period.
14:44 kpreid Not saying it should be technically enforced, just that that's what the rules should say.
14:44 shortcircuit Without preemptive enforcement, it becomes a "you broke some rule you didn't know about, shame on you."
14:45 kpreid Um, no. There's such a thing as having rules written down.
14:45 shortcircuit I'm not saying that's the way it would be intended to be delivered, just how it would be interpreted.
14:45 kpreid But as a separate issue: do *you* agree that one-space-and-lonely-braces is a bad style and it's worth reverting those edits?
14:45 shortcircuit Point me to one of the revisions in question?
14:46 kpreid Look at any of the edits http://rosettacode.org/wiki/Special:Contributions/79.169.23.202 made on Feb 4.
14:46 fedaykin "User contributions - Rosetta Code" http://rldn.net/16oH8
14:47 kpreid I have reverted some of them; most the rest are waiting in my feed queue until I finish getting input on what to do from you-all
14:48 shortcircuit Well, taking [[Collatz_conjecture]] as an example, I like the change to the return code. I don't like the usage of the ternary operator in that case.
14:48 fedaykin http://rosettacode.org/wiki/Collatz_conjecture
14:48 kpreid I'm not talking about the code structure.
14:48 kpreid I'm talking about the formatting *only*. The one-space indent, and moving braces to their own lines.
14:49 shortcircuit k
14:50 shortcircuit Speaking from a practical standpoint, I prefer spaces to tabs in indentation on RC; tabs are difficult to enter in manually, as the browser catches them and uses them for on-page navigation.
14:50 shortcircuit Though I'd prefer two spaces to one. I've not seen one-space indentation before.
14:51 shortcircuit As far as the placement of the braces, I hold no preference of same-line to separate-line. I've coded on projects with stylistic mandates going both ways.
14:51 shortcircuit In the case of braces, if that's all that's changing, then that doesn't make sense.
14:52 shortcircuit Ditto indentation, though hmm...
14:52 shortcircuit If there's going to be mass-edits intended for stylistic changes, then I'd have to agree those should be pre-agreed.
14:52 kpreid they also moved variable decls to the beginning of blocks
14:52 kpreid making big long lists of variable names
14:52 shortcircuit That's kind of a tricky area.
14:53 kpreid and compressed the CSS (removing all unnecessary whitespace and semicolons and replacing "black" with "#000" because it's shorter)
14:53 kpreid (Though it is, of course, fallacious to reject *all* their contributions because *some* of them are obviously wrong)
14:53 shortcircuit We're not specifying which variant of C we're working with, and I believe most C coders out there are more familiar with C89 as a fallback standard than trust C99, but I could be wrong.
14:54 shortcircuit Regarding C, it would be helpful if we could standardize on a primary variant, and then note alternates with workswith.
14:54 kpreid <soapbox>
14:54 kpreid C99 is not a "variant". C99 is THE NEW VERSION OF THE STANDARD.
14:54 shortcircuit (I saw your note on C89/C99 earlier)
14:54 kpreid it was standardized TWELVE YEARS AGO.
14:54 kpreid MOVE ON FROM C89.
14:54 kpreid </soapbox>
14:55 slavik1 c99 is a standard, so is c89, except c89 is an older standard, but I don't think every compiler in the world fully implements c99
14:55 kpreid gcc does
14:55 slavik1 although c99 is awesome since it allows variable definition in beginning of ANY block :)
14:55 shortcircuit Believe me, I know. As a coder who regularly touches C, I wish we could drop C89.
14:55 kpreid (er, sorry, misread your sentence)
14:55 slavik1 kpreid: afaik, not ever gcc implements _all_ of c99
14:56 slavik1 c99 encompasses c89
14:56 slavik1 shortcircuit: how about a disclaimer that code must compile when the c99 flag is specified with gcc?
14:56 kpreid slavik1: it implements enough that they retired their "what parts of c99 have we implemented yet" page
14:56 slavik1 since microsoft's compiler isn't standard anyway
14:56 kpreid (as I found out when I researched that c99 blog post)
14:56 slavik1 kpreid: in that case, I retract my argument ^^
14:57 slavik1 c99 or bust!
14:57 slavik1 </soapbox>
14:57 mwn3d_phone The disclaimer idea would get noisy if placed somewhere that anons would look for it (ie the task page)
14:57 shortcircuit slavik1: Then I have to specify that the code must compile, which isn't possible for the solutions to many of the tasks. Not because they're syntactically invalid, but because they contain only the small subset of a program that the task requires.
14:57 shortcircuit mwn3d_phone: The best place for something like that is the edit page.
14:58 slavik1 shortcircuit: oh ... nvm in that case
14:59 shortcircuit Anyway, it isn't RC's place to force standards compliance or say "you have to be this tall to ride the language wave, and not have any current medical conditions"
14:59 shortcircuit Where there's a question of compliance, it should be possible to sufficiently specify the language variant in question using a note or other template.
15:01 shortcircuit While C may seem a clean enough example case, there are a large number of languages on RC that have only de facto standards rather than de jure ones, and there are even more without either de facto or de jure standards.
15:02 shortcircuit Trying to work around that is much of the impetus behind RC's "if that's how it's done in a particular language, then that's how it's done" with respect to libraries, modules, external code, etc.
15:03 slavik1 shortcircuit: then have that as a field for submissions? or have a guideline to include that?
15:03 shortcircuit kpreid: As far as C goes; I don't mind standardizing on C99 implicitly, and having code that's nonstandard C99--but compliant with some other C standard or implementation--noted with a template flag of some sort.
15:03 shortcircuit slavik1: Having a submission field flag is dependent on getting the Semantic MediaWiki stuff straightened out, which is beginning to feel like a pipe dream.
15:04 shortcircuit As far as a guidelines, that's relatively easy and fine.
15:05 slavik1 shortcircuit: yeah, realized that the wiki is being used after pressing enter
15:06 shortcircuit As far as code formatting goes, I don't have a problem noting in edit pages that mass-formatting edits should be discussed first. I'd also like to see what lint-type tools are available for various languages, and having the users of those languages formulate an agreed-upon style set and produce the appropriate tool configurations for reproducible results.
15:07 shortcircuit mwn3d_phone: I'd have to disagree with you, though, on the number of spaces for indentation. I'm a fan of two-space; it ties in well with old typing habits. :P
15:08 shortcircuit That's more tongue-in-cheek, though. I generally wind up working to other folks' style rules, so my only concerns are practical wrt wiki and site behavior.
15:09 * shortcircuit muses that it'd be nice to have linting in-line with the syntax highlighting.
15:11 shortcircuit kpreid: Regarding one-space-and-lonely braces...I don't care, but it's not something I'd find solely worth an edit or counteredit. If you want to start putting together lint rules on the wiki for folks to work out and come up with something of an agreement on, that'd be great.
15:11 kpreid I have no interest in that, sorry.
15:11 * shortcircuit is off for a bit. Needs breakfast, and needs to get to work.
15:12 kpreid I *do* have interest in fixing this mess.
15:12 kpreid Anyway, nobody has told me *I'm* wrong, so I will keep applying my own judgement (eventually).
15:15 mwn3d_phone kpreid: btw the edits I've seen you do to java code look fine to me so far.
15:15 mwn3d_phone Keep on keepin on
15:53 Coderjoe C99 allows lazy declaration, but C89 DOES allow declaration at the top of any block.
15:54 parsleyfirefly1 joined #rosettacode
15:54 parsleyfirefly1 left #rosettacode
15:55 Coderjoe wrt MS and standards: VS2010's C/C++ compiler is closer to C99 than other versions. hey, they finally added stdint.h! (one of the ones I noticed. I'm not sure what else happened under the hood)
16:00 parsleyfirefly1 joined #rosettacode
16:00 parsleyfirefly1 left #rosettacode
16:02 parsleyfirefly1 joined #rosettacode
16:02 parsleyfirefly1 left #rosettacode
16:06 parsleyfirefly1 joined #rosettacode
16:08 parsleyfirefly1 left #rosettacode
16:12 parsleyfirefly left #rosettacode
16:14 parsleyfirefly joined #rosettacode
17:45 slavik1 left #rosettacode
17:52 slavik1 joined #rosettacode
18:42 mwn3d_phone1 joined #rosettacode
18:43 mwn3d_phone left #rosettacode
18:43 mwn3d_phone1 is now known as mwn3d_phone
19:05 BenBE left #rosettacode
19:10 mwn3d_phone left #rosettacode
19:11 BenBE joined #rosettacode
19:22 BenBE left #rosettacode
19:28 BenBE joined #rosettacode
20:03 * shortcircuit grumbles
20:03 shortcircuit Asterisk apparently doesn't support IPv6.
20:25 opticron yeah, they're working on it
20:25 opticron I think 1.8 has some support for it
20:26 opticron what version are you on?
20:43 shortcircuit 1.6.2, looks like.
20:43 shortcircuit That's on Debian squeeze/sid
20:48 mwn3d_phone joined #rosettacode
20:50 opticron oh...you're installing from a package
20:51 opticron I've never installed asterisk from a package :)
20:58 shortcircuit I'm amazed, but the Gentoo ebuild is even older--1.2.37
21:00 shortcircuit opticron: What do you run it on? I'm competent with debian-derived distroes and with Gentoo, not so much anything else.
21:00 opticron mostly debian
21:01 opticron a little centos here and there
21:01 shortcircuit 'distroes'? My spelling is getting worse.
21:02 opticron the only problem with installing it from source is if your system has EVERYTHING installed from packages and it's not a dev system
21:02 opticron dev tools and headers and whatever else can be huge on the system
21:53 Mathnerd314 joined #rosettacode
22:04 MigoMipo left #rosettacode
22:34 parsleyfirefly1 joined #rosettacode
22:34 parsleyfirefly1 left #rosettacode
22:34 parsleyfirefly1 joined #rosettacode
23:10 Mathnerd314 left #rosettacode
23:25 Mathnerd314 joined #rosettacode
23:29 mwn3d_phone left #rosettacode
23:32 mwn3d_phone joined #rosettacode
23:35 FireFly left #rosettacode

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