| zrafa | again | 00:33 |
|---|---|---|
| zrafa | hi | 00:33 |
| wolfspraul | hi | 00:50 |
| wolfspraul | Made in Optimistan | 10:41 |
| wolfspraul | Ben NanoNote and Milkymist One are Made in Optimistan | 10:41 |
| wolfspraul | http://optimistswithoutborders.org/ideas | 10:41 |
| whitequark | so... pink | 11:09 |
| whitequark | *brain melts* | 11:10 |
| wpwrak | whitequark: so that's what happens when barbie goes to university, takes philosophy classes, and finds an HTML editor someone left open in a classroom PC | 12:08 |
| wolfspraul | I see - optimism unleashes its irresistible power :-) | 12:16 |
| wolfspraul | wpwrak: what space/tab convention do you prefer personally? | 12:16 |
| wolfspraul | in c sources | 12:16 |
| wpwrak | linux kernel style. one tab per indentation level. four spaces for wrapped lines. void *foo(int bar)\n{\tvars;\n\tbody;\n} and \tif (cond) {\n\t\t...\n\t} | 12:20 |
| blogic | wpwrak: \o/ | 12:20 |
| wpwrak | max 80 characters per line. if you have to wrap anyway, make it < 80. | 12:21 |
| blogic | wpwrak: penalties for using spaces as indentation need to be imposed | 12:21 |
| wpwrak | blogic: oh, lots of editors auto-handle this. what's important is that it's 8 :) | 12:21 |
| wpwrak | i.e., there's a penalty in indenting. keeps you from writing ten levels deep nested madness - simply because there's no room for it :) | 12:22 |
| Ayla | what editor are you using? | 12:22 |
| wpwrak | vim | 12:22 |
| Ayla | hmm | 12:22 |
| wpwrak | (but any vi will do) | 12:22 |
| Ayla | here I get two spaces after a IF without brackets, instead of a real tab | 12:23 |
| Ayla | do you know how to change that? | 12:23 |
| wpwrak | dunno. i don't use auto-indentation :) | 12:24 |
| wpwrak | after a while, your fingers learn to do this on their own :) | 12:24 |
| Ayla | well, then there's litle point to use VIM instead of gedit for instance | 12:31 |
| wolfspraul | wpwrak: 8 or 4? (first you say "4 spaces for wrapped lines" then "important is that it's 8" | 12:32 |
| wpwrak | don't know gedit. i like the compact and logical UI of vi | 12:32 |
| wpwrak | wolfspraul: 8 per indentation level. +4 for wrapping. e.g., \tfoo = very_long_identifier(\n\t argument, argument,\n\t still_more); | 12:33 |
| wolfspraul | ok, got it | 12:34 |
| wpwrak | let's see what else ... spaces around operators except for unary and basic arithmetic (+, -, *, /) | 12:42 |
| wpwrak | sizeof(foo) looks like a function. return foo; doesn't. switch and case at the same level (that's a bit of a compromise - otherwise, switch gets too messy to use) | 12:43 |
| wpwrak | space between cast and thing being cast. no pascal parentheses in boolean expressions. (i.e., a == b && c == d and not (a == b) && (c == d) ) | 12:44 |
| wpwrak | yeah, i think that's most of it. many more small details, of course ) | 12:46 |
| wpwrak | ah, also if .. else follows kernel style. so if (foo) bar; else blah; or if (foo) { bar; ...; } else { blah; ...; } or even if (foo) bar; else { blah; ...; } but not if (foo) { bar; ...; } else blah; instead, use if (foo) {bar; ...; } else { blah; } | 12:47 |
| Ayla | I don't agree with switch/case | 12:48 |
| Ayla | and the kernel guidelines forbid if (foo) bar; else { blah; ...; } | 12:49 |
| Ayla | it should always be if (foo) { bar; } else { blah; ...; } | 12:49 |
| Ayla | when there's a "else", that is | 12:49 |
| Ayla | and even if "bar" is a single line of code | 12:49 |
| wpwrak | Ayla: this was discussed sometime ago on lkml and the verdict was that if (cond) one; else { one; two; } is okay | 13:07 |
| wpwrak | the example in CodingStyle is is (cond) { one; two; } else { one; } | 13:07 |
| wpwrak | s/is is/is if/ | 13:07 |
| wpwrak | but that was a long long time ago :) | 13:08 |
| wpwrak | anyway, gotta run. dentist is waiting | 13:08 |
| Ayla | ok | 13:13 |
| DocScrutinizer51 | wpwrak: which direction are you running? | 13:13 |
| DocScrutinizer51 | toward or away from dentist? | 13:13 |
| lars_ | wpwrak: just wanted to ask, when "sometime ago" was. CodingStyle is very explicit about that if one part has brackets the other part should have brackets two. | 13:44 |
| wpwrak | DocScrutinizer: toward :) it's the fitting of a crown. the nasty bits are already done. | 15:28 |
| wpwrak | lars_: hmm, maybe ~6 years ago. lemme see if there are still any traces around | 15:29 |
| wpwrak | hmm, some old wisdom i found while searching: http://www.cs.helsinki.fi/linux/linux-kernel/2001-03/0260.html | 15:50 |
| wpwrak | found the thread. was only a bit more than two years ago. and i mis-remembered: if (foo) x; else { y; z; } is still considered objectionable. just if (foo) x; else y; isn't. | 16:01 |
| wpwrak | here's the thread; http://openmoko-public-mailinglists.1958.n2.nabble.com/PATCH-ASoC-Clean-up-coding-style-issues-in-GTA02-td2580184.html | 16:01 |
| DocScrutinizer | EEEEW! (i.e., a == b && c == d and not (a == b) && (c == d) ) | 19:01 |
| wpwrak | unlearn your pascal :) | 19:02 |
| DocScrutinizer | nevaaar | 19:03 |
| wpwrak | all those redundant parentheses just make the code hard to read | 19:03 |
| DocScrutinizer | haha | 19:03 |
| wpwrak | C has a very thoughtfully designed precedence system. there's no need to pretend it's the mess wirth has created in pascal. | 19:04 |
| DocScrutinizer | a == b && c == d || a == b & c == d | a = b & c == d && a |= b & c == d | 19:04 |
| DocScrutinizer | indeed, every thoughtful | 19:05 |
| DocScrutinizer | ;-P | 19:05 |
| wpwrak | that's why the great gnu has created compiler warnings :) | 19:06 |
| qi-bot | [commit] Jiri Brozovsky: Compiler from (limited) C to language used by HP48 calculators (master) http://qi-hw.com/p/openwrt-packages/5edbd6c | 19:09 |
| whitequark | I think that C's operator system can be improved | 19:18 |
| whitequark | greatly | 19:18 |
| whitequark | just borrow from perl | 19:18 |
| DocScrutinizer | yeah, they tried to fix C compilers that way. The first C 'compilers' were actually bloated macro assemblers for a simple reason of hw not allowing anything better. So C isn't designed for coders but for optimal RAM footprint and CPU usage of the compiler itself, and no compiler warnings level will turn C into a decent coding language. Actually the fewer the syntactic alternatives to code one semantic chunk, the better for readability | 19:19 |
| wpwrak | DocScrutinizer: please go on, while millions of happy C programmers productively ignore your ranting ;-) | 19:22 |
| DocScrutinizer | which ranting? I'm just stating age old known facts | 19:23 |
| DocScrutinizer | and I'm happy with assembler even | 19:23 |
| DocScrutinizer | if you want to to start ranting, ask me about my notion regarding redundancy of ms-basic constructs: while 'condition' do bla blub enddo. vs repeat bla blub until not 'condition'. | 19:26 |
| DocScrutinizer | s/to to/me to/ | 19:26 |
| wpwrak | rants against ms-basic are about as relevant as placing dinosaur traps ;-) | 19:29 |
| blogic | is ranting about gw-basic ok ? :D | 19:32 |
| DocScrutinizer | whatever, on a human logic level your arguments about readability due to precedence system conflict with your recourse to compiler warnings which for sure weren't invented because nobody ever is messing up a term like a == b && c == d by leaving out one char | 19:34 |
| DocScrutinizer | or put simpler: I don't agree on a == b && c == d being ANY readable | 19:36 |
| DocScrutinizer | a && b << 3 | c == d | 19:38 |
| DocScrutinizer | a & b < 3 || c = d | 19:38 |
| DocScrutinizer | a && b < 3 || c == d | 19:39 |
| DocScrutinizer | a && b < 3 | c == d | 19:40 |
| viric | gcc managed to compile C without any optimal ram footprint | 19:41 |
| viric | does that turn C more decent? | 19:41 |
| DocScrutinizer | neither you nor gcc existed when the compilers were built according to my above rationale | 19:42 |
| DocScrutinizer | C however existed already | 19:42 |
| DocScrutinizer | C is an extremely powerful macroassembler, nothing beyond. A lot of experienced coders will agree on that | 19:43 |
| viric | and what new exactly does your idea bring? | 19:44 |
| DocScrutinizer | eh? | 19:45 |
| viric | :) | 19:45 |
| DocScrutinizer | and what new does your question introduce to the topic? | 19:45 |
| viric | :) | 19:45 |
| blogic | i thinkt the answer is "the above rational" or not ? | 19:45 |
| blogic | +e | 19:45 |
| DocScrutinizer | I think you lost me | 19:46 |
| blogic | s/rational/rationale/g | 19:46 |
| blogic | where is the sed bot ß | 19:46 |
| DocScrutinizer | and the answer is 42, always been | 19:46 |
| viric | Maybe it is me who got lost :) | 19:47 |
| viric | I find C ok for its job | 19:48 |
| DocScrutinizer | another undeniable fact is that *read*ability increases with redundancy, until it crosses a certain threshold. It maybe *write*ability that's better without "redundant" parentheses etc | 19:49 |
| viric | do you like lisp? :) | 19:50 |
| viric | C is quite a good balance between the complexity to write/read and the complexity of writing a compiler for it. | 19:51 |
| viric | giving enough granularity to the programmer to achieve code that runs fast | 19:52 |
| DocScrutinizer | nonsense, nowadays no high level procedural language lacks the opportunity to allow coders to write fast code | 20:15 |
| whitequark | blogic: sed bot is too stupid to understand perl syntax. just s///, nothing more. | 20:24 |
| whitequark | btw, this rant is funny | 20:25 |
| whitequark | (note that I'm not saying any of the above is not true. just funny a bit.) | 20:26 |
| blogic | s/rational/rationale/ | 20:34 |
| blogic | i thinkt the answer is "the above rational" or not ? | 20:34 |
| blogic | s/rational/rationale/ | 20:34 |
| blogic | *magic* | 20:34 |
| DocScrutinizer | xxxxxxxxx | 20:57 |
| DocScrutinizer | s/x/ABC/g | 20:57 |
| wolfspraul | people surely love that feature ;-) | 20:57 |
| DocScrutinizer | yes | 20:57 |
| wolfspraul | we should get some of those interactive fiction bots into the channel... | 20:57 |
| DocScrutinizer | luckily buffer size is limited | 20:57 |
| DocScrutinizer | what annoys way more is you can't fix 2 typos with sedbot | 20:59 |
| DocScrutinizer | unless they are identical | 20:59 |
| DocScrutinizer | tis sucks somtimes | 21:00 |
| DocScrutinizer | s/tis/this/;s/somt/somet/ | 21:00 |
| DocScrutinizer | duh | 21:00 |
| DocScrutinizer | :-D | 21:00 |
| DocScrutinizer | actually it's probably not exactly the buffersize that's limited, rather the max length of a single post as defined by freenode | 21:17 |
| DocScrutinizer | xxxxx#xxxxx#xxxxx#xxxxx#xxxxx#xxxxx#xxxxx#xxxxx#xxxxx#xxxxx# | 21:20 |
| DocScrutinizer | s/x/0123456789/g | 21:20 |
| DocScrutinizer | hah | 21:20 |
| DocScrutinizer | xxxxx# | 21:20 |
| DocScrutinizer | s/x/0123456789/g | 21:21 |
| DocScrutinizer | whitequark: sedbot it just not editing anything but the least post - blogic's last post was >>[2012-01-04 20:45:57] <blogic> +e<< | 21:33 |
| DocScrutinizer | and even /g parameter isn't meant to change that | 21:33 |
| DocScrutinizer | a funny idea though: s/x/y/G edits *all* posts ever posted by the user since last /join ;-D | 21:35 |
| DocScrutinizer | wolfspraul: what do you think, will users love *that*? ;-) | 21:35 |
| wolfspraul | ;-) | 21:49 |
| whitequark | hmm | 21:53 |
| whitequark | s,hmm,hm, | 21:53 |
| whitequark | :/ | 21:53 |
| DocScrutinizer | sillii idea to think sedbot would act on all lines starting with "s" ;-D | 21:59 |
| DocScrutinizer | s/illii/illy/;some illegal command | 22:01 |
| DocScrutinizer | hahaha | 22:01 |
| DocScrutinizer | ok | 22:01 |
| DocScrutinizer | s/ok/OK/;!seen DocScrutinizer51 | 22:02 |
| DocScrutinizer | in fact, I think: not even that will work | 22:04 |
| DocScrutinizer | s/,//;s/:// | 22:04 |
| DocScrutinizer | ooh | 22:04 |
| DocScrutinizer | s/o/(/ | 22:05 |
| DocScrutinizer | MEH!! | 22:05 |
| DocScrutinizer | thete's not been any /g | 22:05 |
| DocScrutinizer | there* | 22:05 |
| DocScrutinizer | I I wonder | 22:06 |
| DocScrutinizer | s/I // | 22:06 |
| DocScrutinizer | now THAT is *BAD* | 22:07 |
| DocScrutinizer | ibot/infobot/apt obeys /g | 22:08 |
| DocScrutinizer | qi-bot has a bug | 22:08 |
| DocScrutinizer | am I right in thinking qi.bot has a fine bug in this regex-sustition function? | 22:11 |
| DocScrutinizer | s/in/with/ | 22:11 |
| DocScrutinizer | :-/ | 22:12 |
| DocScrutinizer | who's qi-bot's master? | 22:15 |
| jow_laptop | yay, Bitcoin spam via twitter relayed to IRC | 22:59 |
| wolfspraul | yay well | 23:00 |
| wolfspraul | I actually clicked on the link to see whether there is open hardware there, but couldn't find it right away | 23:00 |
| jow_laptop | ;) | 23:00 |
| wolfspraul | jow_laptop: what do you think about bitcoins? | 23:08 |
| wolfspraul | seems at some point a technology like that begs for dedicated secure hardware | 23:08 |
| wolfspraul | which should be open to be secure imho | 23:08 |
| roh | wolfspraul: bc arent the answer to the question for digital money. they are an experiment stirring stuff up to see whats possible. | 23:18 |
| wolfspraul | yes, roughly my thinking too. but as a transaction medium they work today, I would sell nanos and milkys for bitcoins, no problem. | 23:19 |
| wolfspraul | also the combination of cryptography and networking is so interesting, maybe it can be used for other applications | 23:20 |
| wolfspraul | from a hardware side, I think this stuff cannot work unless there is secure hardware | 23:20 |
| roh | in the end it boils down to a technical issue. how to make sure its an exchange currency and not a currency imploding as the real ones right now due to speculation | 23:21 |
| DocScrutinizer | bitcoin, a chain letter philosophy concept | 23:31 |
| wolfspraul | well I'm a small business man, and I am struggling with payment systems every day | 23:36 |
| wolfspraul | every time I struggle, I think of bitcoin now and there is *hope* :-) | 23:36 |
| wolfspraul | thank you bitcoin! | 23:36 |
| wolfspraul | so I definitely sell hardware for bitcoins, no problem | 23:37 |
| wolfspraul | DocScrutinizer: you send me some bitcoins, I send you a Milkymist One ;-) | 23:37 |
| wolfspraul | I do not believe in the revolutionary part of bitcoins though, they will not replace other currencies etc. | 23:38 |
| wolfspraul | but maybe a small little payment system, for geeks first? right now it seems to work... | 23:38 |
| wolfspraul | unless someone breaks through the crypto part right away and makes an infinite number of them, I think they will continue to stay around | 23:39 |
| DocScrutinizer | I found nobody who was able to convince 9 others and me about bitcoins being a thing worth looking into | 23:40 |
| DocScrutinizer | and tbh I'm not looking for somebody | 23:40 |
| erikkugel | everything has it's merits, but bitcoin's main advantage currently is hiding illegal activity. Once it had OTHER advantages, popularity might grow. | 23:41 |
| --- Thu Jan 5 2012 | 00:00 | |
Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!