#eulora Logs for 24 Feb 2017



February 24th, 2017 by Diana Coman
danielpbarron: great [13:20]
danielpbarron: did the server go down for a sec or was it on my end? [13:21]
mircea_popescu: im still crafting [13:21]
danielpbarron: i didn't lose connection to my irc machine [13:22]
danielpbarron: day+ long craft was most of the way done [13:22]
mircea_popescu: sux [13:22]
danielpbarron: wait are you sure you're crafting? [13:24]
danielpbarron: i only see foxy online [13:24]
mircea_popescu: nop, it popped. [13:24]
mircea_popescu: hm. [13:24]
mircea_popescu: well that sucks, i was doing a multi-day craft. [13:24]
diana_coman: uhm, that's weird: the server WAS on all the time; but foxy DID get disconnected at some point too [13:56]
diana_coman: I thought it was only on my side [13:57]
mircea_popescu: i guess it must have lost connection [13:57]
Birdman: its more normal for me to lose connection at some point overnight than not [16:02]
Birdman: rarely do i wake up and find my guy still barehanding [16:03]
mircea_popescu: sucks. [16:06]
Birdman: ya this machine is all messed up [16:07]
diana_coman: so serious and so gogetta [17:14]
mircea_popescu: i read georgeta [17:19]
diana_coman: same mary with another hat as it were [17:20]
diana_coman: with another voil georgette might be [17:20]
mircea_popescu: jej [17:23]
Birdman: can i pick that stuff up in like 20 min danielpbarron ? [19:49]
Birdman: !~calc 9999*1.19*200 [20:20]
jhvh1: Birdman: 9999*1.19*200 = 2379762 [20:20]
Birdman: seems i found two dm tinies over a days worth of barehanding, not bad, and it may be worth the most to sell the claim and enumeration itself than to build with my terrible bits [20:25]
mircea_popescu: worth trying out anyway [20:38]
DicePower: howdy! [22:00]
DicePower: So I just came up with a game changing idea to revolutionize coding! [22:01]
DicePower: (or perhaps it already exists, but I've never seen it) [22:01]
ben_vulpes: DicePower: oyea? [22:03]
DicePower: I tried to hype a couple other programming channels up for my big idea, but no one showed any interest in hearing it, so I figured, why not share it with the awesome people of #eulora!? [22:04]
DicePower: So what do you do when you have 100 errors in your computer program? [22:06]
DicePower: You take the first one, paste it into Google or some debugging Web site, and sometimes sift through a number of solutions before finding the correct one for your problem. [22:07]
DicePower: At least that's how I was taught to do it, and have always done things. One error message at a time, go through solutions until something works, see how many errors remain, and repeat with the first error message in the new batch. [22:08]
DicePower: The first error message is often thought to be the most important one. [22:10]
DicePower: Or at least the most important one for whatever problem it turns out to be related to. [22:10]
DicePower: Or at the very least, "as good a place as any to start." [22:11]
DicePower: But why only take one error message at a time? That seems inefficient. Of course, you can only fix one problem at a time, but there is probably very often lots of other information in those 99 error messages that could narrow down the associated problem. [22:12]
DicePower: Why not make a Web site where you paste your entire error output from an attempted compilation? It could search the Web for solutions to the first error message like Google, but then use the other error messages to provide a more precise diagnosis of your problem. [22:14]
DicePower: It could even output the remaining solutions to one page, so the programmer could compare them at a glance. [22:15]
DicePower: In fact, it could attempt to do this for multiple problems in the code at once, perhaps. [22:16]
DicePower: So at best, if it was able to map the error messages presentations to code problems with perfect precision, it could find the correct fixes to each of your problems, or at least narrow the possibilities dramatically compared to searching one error message at a time. [22:18]
DicePower: Which could increase productivity. [22:19]
mircea_popescu: this would work if a) google weren't an utter pile of crap dedicated entirely to keeping dumbericans dumb and b) the text it has to work with weren't produced by the sort of imbeciles who write wikipedia and experts sex change. [22:26]
DicePower: I mean, aren't those limitations of current programming anyhow? [22:28]
mircea_popescu: mno. [22:28]
mircea_popescu: most programming worth the name is not done by esltards. [22:28]
DicePower: I see [22:28]
mircea_popescu: also most programming worth the name isn't done by the sort of cubical drone whose mental metabolysm consists of google checklists. [22:29]
DicePower: Maybe that's one's just me :D [22:29]
mircea_popescu: i've seen plenty of "tech companies", "projects", etcetera. [22:29]
DicePower: Is the process really all that different with good programmers though, or do they just have more experience and thus more developed instinctual skill? [22:31]
mircea_popescu: instinctual skill ? [22:31]
mircea_popescu: what's that, like meat bread ? [22:31]
mircea_popescu: good programmers understand the negative space of their problem. everyone else is undistinguishable from a trained monkey in any practical sense. [22:32]
DicePower: I mean, they've gone through the Google checklist process enough times to have experience with many types of problems. [22:32]
lobbes: there's quick duct-tape fixes, and then there is understanding your problems at a more fundamental level, which a google search checklist cannot provide [22:34]
lobbes: can only be achieved by actually putting in the work to learn [22:34]
mircea_popescu: you can't pull yourself up into orbit by your britches ; you can't become a don juan through masturbating to enough tumblr porn. [22:34]
mircea_popescu: you don't become a programmer through doing dumb shit like that. [22:35]
DicePower: Interesting analogies. [22:35]
DicePower: What sorts of resources would these good programmers then use to solve problems they don't already know how to solve? [22:39]
ben_vulpes: the space of "problems they don't already know how to solve" is far too broad to be a useful category [22:46]
ben_vulpes: the most difficult shit i've ever had to deal with is the mess of other people's code. side-effect soup in the browser. [22:48]
ben_vulpes: the resource i always fall back on is thinking really hard about the problem until the muses bless me with a coherent model and path forward. a path which typically reduces to applying the flamethrower to the all knots of stupidity in sight. [22:49]
ben_vulpes: bbl [22:51]

Comments feed: RSS 2.0

Leave a Reply