--- Log opened Thu Oct 25 00:00:00 2012 |
00:01 | <@gnolam> | You've already abandoned Vinland Trail? Or is this a parallel projecT? |
00:01 | <@gnolam> | s/T/t |
00:02 | <@AnnoDomini> | VT has reached its design goals. I may tweak it further later, but at the moment, I want to make this. |
00:03 | <~Vornicus> | How are you generating terrain and rivers? |
00:04 | <@AnnoDomini> | First I designate 95% of the space to be land, and the rest to be water. |
00:04 | <@AnnoDomini> | Then I randomly expand the splotches of water, so they make bigger pools. |
00:04 | <~Vornicus> | I see. So no plasma fractal or anything doing the stuff. |
00:05 | <@AnnoDomini> | Then I randomly assign terrain types according to stuff appropriate to climate bands. |
00:05 | <@AnnoDomini> | No, I wouldn't know where to begin with fractals. |
00:06 | <~Vornicus> | basic method afaict is to use a plasma fractal (easy to look up) to assign altitudes, then use the altitudes to assign terrains. |
00:06 | <~Vornicus> | You can then do things like "rain piles up on the windward side of mountains", "rivers go from cell to cell in the steepest direction", and some other stuff like that. |
00:07 | <@AnnoDomini> | I'll look it into it later. Right now I want something that resembles a proper map, so I can get on with the rest of the critical components. |
00:08 | | gnolam [lenin@B19C04.C4B928.9A56F0.CF2CAC] has quit [Client closed the connection] |
00:08 | | gnolam [lenin@Nightstar-ccbf4b44.cust.bredbandsbolaget.se] has joined #code |
00:08 | | * AnnoDomini copies a bit of log into the coments in the mapgen file. |
00:10 | | himi [fow035@D741F1.243F35.CADC30.81D435] has joined #code |
00:10 | | mode/#code [+o himi] by ChanServ |
00:14 | <~Vornicus> | As far as your rivers go |
00:15 | <~Vornicus> | After each step of your random walk, add all 8 tiles around the previous step to the visited list. |
00:17 | <@AnnoDomini> | Intredasting. |
00:18 | <~Vornicus> | This will prevent your river from looping. |
00:18 | <~Vornicus> | Successful termination occurs when the river reaches a sea or other river tile; failure occurs when you run out of tiles to go to that are not on the visited list. |
00:22 | <&McMartin> | Argh |
00:22 | | * McMartin breaks everything on a test branch with an experiment, spends like three hours explaining to people that the test branch experimental change is not The Way The Spec Always Was |
00:22 | <~Vornicus> | what |
00:23 | <&McMartin> | We have a somewhat fuzzy spec |
00:23 | <&McMartin> | I did some experimentation on its claimed requirements to see if they were really requirements. |
00:23 | <&McMartin> | It looks like they are |
00:23 | <&McMartin> | Cue humongous I THOUGHT WE COULD RELY ON X across three departments |
00:23 | < gnolam> | AnnoDomini: In case this wasn't posted during peer's malevolent disconnection of me, http://www-cs-students.stanford.edu/~amitp/game-programming/polygon-map-generati on/ might be of interest. |
00:24 | < gnolam> | (Even if you go another route for the basic generation, some of the stuff in there can still be useful) |
00:24 | <~Vornicus> | I wanted to adapt that somewhat to spherical worlds |
00:26 | <@AnnoDomini> | Cool. |
00:35 | | Derakon[AFK] [Derakon@Nightstar-a3b183ae.ca.comcast.net] has quit [Ping timeout: 121 seconds] |
00:37 | | cpux [cpux@Nightstar-c5874a39.dyn.optonline.net] has joined #code |
00:48 | | Derakon [Derakon@Nightstar-a3b183ae.ca.comcast.net] has joined #code |
00:48 | | mode/#code [+ao Derakon Derakon] by ChanServ |
00:52 | | Attilla_ [Obsolete@Nightstar-be2d2d0f.as43234.net] has joined #code |
00:53 | | Attilla [Obsolete@Nightstar-699a2f89.as43234.net] has quit [Ping timeout: 121 seconds] |
00:58 | | Attilla_ [Obsolete@Nightstar-be2d2d0f.as43234.net] has quit [[NS] Quit: ] |
01:02 | | celticminstrel [celticminst@Nightstar-05d23b97.cable.rogers.com] has joined #code |
01:04 | <@AnnoDomini> | Vornicus: Don't you mean all nine spaces? Because it otherwise could double back. |
01:05 | < Syka> | man |
01:05 | < Syka> | there was this game a while ago |
01:05 | < Syka> | that was basically a mindfuck |
01:10 | <~Vornicus> | Anno: could but only on the second walk step; it's probably easier to do all nine though because it's less hassle. Make sure your visited pile is a set though. |
01:11 | <@AnnoDomini> | A set? |
01:11 | < Syka> | like you walked down one hallway, turned around, then turned around again, and there was a door there suddenly |
01:12 | <~Vornicus> | Anno: as opposed to a list or something. |
01:13 | <~Vornicus> | A set is good because it has fast existence checks and often has good differencing tools. |
01:13 | <@AnnoDomini> | I'm using another dimension of the map to store visited status. |
01:13 | <@AnnoDomini> | Syka: http://www.youtube.com/watch?v=13YlEPwOfmk |
01:15 | < Syka> | ahah what |
01:26 | <@AnnoDomini> | Mindfuck, that's what. |
01:27 | < gnolam> | Hah |
01:49 | < gnolam> | Oh hey, looks like we can make due with the regulator we've got after all. Input voltage of 14-16 V. |
01:57 | < gnolam> | Next up: to check if the computer still boots at 14 V. |
01:58 | < gnolam> | (The regulator had an unexpectedly high dropout voltage; it only reaches saturation at ~14.5 V) |
02:00 | | * AnnoDomini partially succeeds with rivers. |
02:00 | <@AnnoDomini> | They don't look fucked up anymore, but thanks to me being unable to do the random walk correctly, they all go in the same direction if they can. |
02:08 | | Derakon is now known as Derakon[AFK] |
02:17 | <@AnnoDomini> | This seems like a very simple problem, but I can't solve it. |
02:17 | <@AnnoDomini> | I want to choose between eight options, but without knowing beforehand which options are valid. |
02:18 | <@AnnoDomini> | Trying to do "keep trying until you find a valid option" goes into a semi-infinite loop. |
02:18 | | mode/#code [+ao ToxicFrog ToxicFrog] by ChanServ |
02:19 | <@AnnoDomini> | I have no idea how to mark options already considered as out of the pool of options. |
02:21 | <@Reiv> | Anno: Iterate through choosing valid options |
02:21 | <@Reiv> | Then pick randomly from the List Of Good. |
02:22 | <@AnnoDomini> | This I do not know how to do. |
02:23 | <&ToxicFrog> | AnnoDomini: either actually remove each option from the pool after considering it, or use an ordered pool, shuffle it first, and then iterate in order |
02:39 | <@AnnoDomini> | Hmm. Runtime error. Probably trying to write to an out-of-bounds location. |
02:48 | | mac [NSwebIRC@Nightstar-b902fba7.chi.megapath.net] has joined #code |
02:49 | < mac> | Hey any of you guys know how to protect a .htaccess file on a server? |
02:50 | | * AnnoDomini boggles. Floating point exception? |
02:51 | < mac> | heh? |
02:53 | <@AnnoDomini> | No idea how to help you. |
02:54 | | * AnnoDomini finds that for some reason, a counter is isn't incremented. But the only reason the counter wouldn't be incremented is that there aren't any valid options, in which case the damn thing should break; |
02:54 | < mac> | Pastebin it |
02:55 | <@rms> | mac: Apache confs ship with that done already |
02:55 | < mac> | well we are having problems with russians changing things and adding php files :>C |
02:55 | < mac> | we have problems rms |
02:56 | <@AnnoDomini> | I'm not sure pastebinning would help. Not unless I pasted half my project. |
02:56 | < mac> | just paste the counter |
02:56 | < mac> | anno |
02:56 | <@rms> | But basically you just pattern match for .*/\.ht.* then make the allow order: deny,allow, then deny all. |
02:57 | <@rms> | But if you need security help I've got experience getting a LAMP stack (PHP) pci/dss compliant. |
02:57 | <@rms> | (Among other things) |
02:58 | <@AnnoDomini> | mac: If you insist. http://pastie.org/5112407 |
03:01 | < mac> | anno i couldn't find a problem |
03:02 | < mac> | rms , What if im using network solutions? |
03:02 | <@rms> | Is that a company or a product? |
03:02 | < mac> | the company for hosting |
03:03 | < mac> | and i guess their product by using their web interface if thats what you meant by product |
03:03 | <@rms> | Is that a shared host? |
03:04 | <@rms> | ... or a VPS? |
03:04 | <@rms> | (If it's the former you're kind of at their mercy for security changes) |
03:07 | <@rms> | If you can remove the write permisions on directories that don't need them, do so. |
03:07 | <@rms> | For the ones that do, make sure php scripts can't be run from there. |
03:10 | < mac> | ok |
03:10 | < mac> | its a shared host |
03:10 | | * AnnoDomini reverts to rivers going in one direction only, because fuck that shit. |
03:11 | <@AnnoDomini> | I'll do this later. |
03:11 | < mac> | were can i host a server thats not a shared host |
03:11 | < mac> | ? |
03:11 | < mac> | i dont want to host my own. :/ |
03:13 | <@rms> | If you know how manage one I suggest getting a VPS |
03:13 | <&ToxicFrog> | You rent a dedicated server and prepare to pay $$$ for it. |
03:14 | < mac> | how much are we talking |
03:14 | <@rms> | VPSes tend to be around 10$/mo in the entry-level area. |
03:14 | <@rms> | You can get cheaper. |
03:15 | <@rms> | Amazon might still be offering free ones. |
03:15 | <&ToxicFrog> | ...where? IME dumb shared hosting is $10/mo at the low end |
03:15 | <@Namegduf> | Linode do 512MB VPSes for $20. |
03:16 | <@Namegduf> | Providers which go below 512MB can get to $10 or less. |
03:16 | < Syka> | doing anything in Australia is $150/mo |
03:16 | < Syka> | :D |
03:16 | <@rms> | http://www.cheapvps.co.uk/ |
03:16 | <@Namegduf> | Linode's price is pretty typical. |
03:17 | <@Namegduf> | There's a lot of ripoff providers that rely on a lack of research, though. |
03:17 | | * rms is with VPSVille |
03:17 | < Syka> | i've got a fibervolt VPS |
03:17 | < Syka> | i haven't done too much with it, got to set it up |
03:17 | <@rms> | I'd go with the bigger names if you can, like Linode, Amazon or such |
03:17 | < Syka> | except i downloaded a 500MB test file in ten seconds |
03:17 | < Syka> | so uh |
03:18 | < Syka> | i just sat there stunned as that's better than my *LAN* here |
03:18 | <@rms> | Only issue with a VPS is you have to set it up yourself. |
03:18 | <@Namegduf> | All that said, doing web hosting below 512MB is difficult. |
03:19 | <@rms> | You /can/ get "managed" VPSes, but I'm weary of those. |
03:19 | < Syka> | my fibervolt has 128MB of ram lol |
03:19 | <@Namegduf> | You need a bunch of careful tweaking for PHP and MySQL. |
03:19 | <@Namegduf> | (And probably can't use Apache) |
03:19 | < Syka> | ...except it /was/ $25 for a whole year, so I don't mind |
03:19 | <@Namegduf> | So I'd take the baseline Linode at $20 as fairly close to the best you'll find. |
03:19 | | * rms ran PHP and MySQL on 64MB of RAM |
03:20 | | * rms thinks Namegduf is spouting BS |
03:20 | < mac> | currently we dont the person who is having this problem does not use php. |
03:20 | < Syka> | rms: 1 req/sec? :P |
03:20 | <@Namegduf> | rms: I think you missed the part where I said you needed careful tweaking, dumbass. |
03:20 | <@rms> | I didn't do anything special. |
03:20 | < Syka> | imo i don't think php and mysql need 512mb of ram |
03:20 | <@Namegduf> | Yes you did. |
03:20 | <@rms> | Nope. |
03:20 | < Syka> | i've ran postgres on 512mb |
03:20 | <@Namegduf> | MySQL eats 100MB on startup for InnoDB on the default config. |
03:20 | <@Namegduf> | As a flat cost. |
03:21 | < Syka> | wait 256mb, cos it was on my raspi |
03:21 | <@Namegduf> | There is no way that would even start without disabling InnoDB. |
03:21 | < Syka> | wtf does innodb need 100MB of RAM for |
03:21 | <@rms> | Might have been after the ->128MB upgrade |
03:22 | <@rms> | Eitherway, the mobo had two ram slots, and I only had 64MB sticks for it. |
03:22 | < mac> | so they also added a crap ton of other files to the server, so obviously they were some how able to ftp in or something |
03:22 | <@rms> | Syka: probably does memmap'ing for speed. |
03:22 | < mac> | is it possible to inject a file into a website if the js isnt secure or something? |
03:22 | <@rms> | mac: Not obviously |
03:22 | < mac> | there is no sql set up. |
03:22 | <@rms> | They just need an upload script, that's it. |
03:23 | < Syka> | mac: you can upload files with a suitably dangerous php script |
03:23 | < mac> | how would some one go about denying it? |
03:23 | <@rms> | "js isn't secure" <-- if the JS is client-side only you don't need to worry about it's security IMO |
03:23 | < Syka> | the only js security worry is XSS stuff |
03:24 | < Syka> | but even then that's clientside, and nobody cares about clients |
03:24 | <@rms> | That's server-side stuff though |
03:24 | <@rms> | <rms> If you can remove the write permisions on directories that don't need them, do so. |
03:24 | <@rms> | <rms> For the ones that do, make sure php scripts can't be run from there. |
03:24 | < Syka> | ...fuck |
03:24 | <@rms> | Syka: ? |
03:24 | < mac> | well shit |
03:25 | <@Namegduf> | rms: My suspicion is that you somehow got a config with InnoDB disabled, because it's a real bloated PoS. Also you already weren't using Apache, I'd guess, or at least not any kind of modern version of it. |
03:25 | <@Namegduf> | Well, I say real bloated PoS, it's just... a lot heavier. |
03:25 | <@rms> | *Shrugs* |
03:26 | <@rms> | All I know is that computer was running MySQL and serving PHP pages while having at most 128MB RAM. |
03:26 | < mac> | rms so can i do that from ssh-ing in?? |
03:26 | <@rms> | Yes |
03:26 | < mac> | yes? |
03:27 | <@rms> | What are you hosting anyways? |
03:29 | < mac> | a bunch of sites. |
03:29 | < mac> | have 2 go be back later |
03:29 | | mac [NSwebIRC@Nightstar-b902fba7.chi.megapath.net] has quit [[NS] Quit: Page closed] |
03:30 | <&ToxicFrog> | |
03:31 | < Syka> | rms: one of my UPSs just freaked the fuck out |
03:31 | <@rms> | Ah |
03:31 | <@rms> | Love it when that happens |
03:31 | < Syka> | was like beeeeepebebebebebebebeeeeeeeeeee- beee |
03:31 | < Syka> | it was a little desktop one |
03:32 | < Syka> | i have it on my workbench |
03:34 | <&ToxicFrog> | The cats managed to turn mine off at one point. |
03:35 | <@Namegduf> | rms: Another option is that your default config was better tuned than typical distributions offer. |
03:35 | <@Namegduf> | http://www.lowendbox.com/blog/reducing-mysql-memory-usage-for-low-end-boxes/ <- This goes over some of the key things |
03:36 | <@Namegduf> | Or at least mentions them in passing while offering a config with low memory usage. |
03:36 | <@rms> | Could be. |
03:36 | <@rms> | On reflection, I was probably using lighttpd on that box |
03:36 | <@rms> | Apache was in use on another box around the same time, so I may have gotten confused in that regard. |
03:36 | <@Namegduf> | I think that 25MB is before InnoDB or after some tweaking; I know the 100MB-on-startup thing because I've run MySQL and had "fun" with that. |
03:37 | <@rms> | Heh |
03:37 | <@Namegduf> | Probably, yeah. Apache's rough. |
03:38 | <@rms> | So, Arch Linux is pissing me off as of late, but I love pacman (3.x; 4.x is shit) and hate apt and debian(-likes)'s rampant idiocies. Suggestions for distros for me to try? |
03:39 | <@rms> | I've been trying slackware, it's a bit lacking in the packages dept. |
03:40 | <@Namegduf> | I use Debian everywhere, so no real advice here. |
03:46 | <&ToxicFrog> | SUSE and Debianoids. Sorry. |
03:57 | | mac [NSwebIRC@Nightstar-fe8a1f12.il.comcast.net] has joined #code |
03:57 | < Syka> | debian/derivs everywhere for me |
03:57 | < Syka> | apt = <3 |
04:02 | | Kindamoody[zZz] is now known as Kindamoody |
04:23 | < celticminstrel> | apt makes no sense to me. |
04:24 | <&McMartin> | rms: Your only remaining option is Fedora, AFAICT. |
04:24 | <&McMartin> | Or gentoo~ |
04:24 | <&McMartin> | If you go fedora get the LXDE spin because fuck GNOME |
04:26 | | Derakon[AFK] is now known as Derakon |
04:51 | | celticminstrel [celticminst@Nightstar-05d23b97.cable.rogers.com] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] |
04:51 | < Syka> | i wrote my own package manager once |
04:51 | < Syka> | i made a teeny tiny little distro |
04:51 | < Syka> | under 30mb with coreutils and bash and some stuff |
04:52 | < Syka> | wrote my own package manager in C++ |
04:52 | < Syka> | it installed packages fine |
04:52 | < Syka> | and could even grab from a network |
04:52 | < Syka> | but then I fucked up in manually writing one package manifest and deleted root |
04:53 | <@Tamber> | *snrk* |
04:53 | < Syka> | fun times |
04:53 | <@Tamber> | Package mongler. :D |
04:53 | < Syka> | so yeah |
04:53 | < Syka> | basically it was an install-only package manager |
04:55 | < Syka> | it arguably has limited uses |
04:56 | < Syka> | but since the distro was ram-boot only, it would pretty much start up, then install bash and a few other things |
06:21 | | Derakon is now known as Derakon[AFK] |
06:46 | | mac [NSwebIRC@Nightstar-fe8a1f12.il.comcast.net] has left #code [""] |
07:20 | | Kindamoody is now known as Kindamoody|breakfast |
07:52 | | himi [fow035@D741F1.243F35.CADC30.81D435] has quit [Ping timeout: 121 seconds] |
08:11 | | Attilla [Obsolete@Nightstar-0c734f9a.as43234.net] has joined #code |
08:29 | | Kindamoody|breakfast is now known as Kindamoody |
08:46 | | Kindamoody is now known as Kindamoody|out |
08:54 | | Attilla_ [Obsolete@Nightstar-0c734f9a.as43234.net] has joined #code |
08:56 | | Attilla [Obsolete@Nightstar-0c734f9a.as43234.net] has quit [Ping timeout: 121 seconds] |
09:10 | | Courage [Moltare@583787.FF2A18.190FE2.4D81A1] has quit [Ping timeout: 121 seconds] |
09:14 | | Moltare [Moltare@583787.FF2A18.190FE2.4D81A1] has joined #code |
09:39 | <@froztbyte> | rms: mind elaborating on these "rampant idiocies"? |
09:40 | <@froztbyte> | though I think we went through this last time as well, and it just boiled down to you not liking things |
10:09 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has joined #code |
10:09 | | mode/#code [+o himi] by ChanServ |
10:56 | | RichyB [richardb@Nightstar-3b2c2db2.bethere.co.uk] has joined #code |
11:59 | <@Tarinaky> | http://dilbert.com/dyn/str_strip/000000000/00000000/0000000/100000/60000/7000/80 0/167830/167830.strip.zoom.gif |
12:22 | | Attilla_ [Obsolete@Nightstar-0c734f9a.as43234.net] has quit [Ping timeout: 121 seconds] |
12:28 | | Attilla [Obsolete@Nightstar-b9c27e79.as43234.net] has joined #code |
12:34 | <@rms> | froztbyte: Example packages: AIDE (which has the default configuration file changed to /please/specify/-c/to/point/to/a/config/file without changing the man page) and postfix which launches a full curses UI on install which fucks with automated deployment. |
12:34 | <@rms> | Then they split each package into a billion little packages, and provide no way (that I've found) to automatically install certain subsets of the packages installed (IE: always install -dev subpackages) |
12:34 | <@rms> | Then there's thier package manager, which is also split into 5 (IIRC) programs, so I'm always having to `man` to figgure out which sub-command goes with which program. |
12:34 | <@rms> | Finally there's 17 steps to making a package in the rare cases that I have to roll my own, which gives me more deployment headaches. |
12:35 | <@froztbyte> | sounds like you're mostly doing it wrong, tbh |
12:36 | <@rms> | Oh, I'm not sure if this one is just an Ubuntu-ism, but Debian might do it too: packages pointlessly requiring services, then when services get installed they automatically start fucking up your beautiful Principle of Least setup. |
12:36 | <@froztbyte> | set your debconf priority and the question screens go away || preload the settings through one of the debconf hooks |
12:36 | <@froztbyte> | rms: ubuntu is a really really really terrible thing to use as a base for anything |
12:37 | <@froztbyte> | which is unfortunate |
12:37 | <@rms> | Yeah, no shit |
12:37 | <@froztbyte> | it seems like anything has to become horrible for it to be popular |
12:37 | | * rms admined Ubuntu Servers for 2 years and hated it. |
12:39 | <@rms> | So there's a way to automatically install all the -dev packages without me needing to explicitly do something? |
12:40 | <@froztbyte> | installing the build-deps should pretty much cover all of that |
12:40 | <@rms> | IE: I install SDL freshly and it just grabs SDL-dev just from 'apt-get install sdl' |
12:40 | <@froztbyte> | which is one more step, but still little enough |
12:41 | <@froztbyte> | though why do you need -dev that often? |
12:41 | <@froztbyte> | (as a matter of interest, more than anything) |
12:41 | <@rms> | Because I don't want to keep track of which workstation has which dev-libs? |
12:42 | <@froztbyte> | no, that's not what I mean |
12:42 | <@froztbyte> | what leads to you requiring dev libs that often? |
12:42 | <@rms> | If I know a program has SDL and it runs, I assume that workstation can compile SDL programs. |
12:42 | <@froztbyte> | like, what sort of work do you do that requires it all the time? |
12:42 | <@rms> | I try new software frequently |
12:43 | <@froztbyte> | hmm |
12:43 | <@froztbyte> | irunno |
12:43 | <@froztbyte> | I can see where your hate stems |
12:43 | <@froztbyte> | but I'm not sure if it's justified |
12:44 | | * Azash never understood the idea behind ubuntu server |
12:44 | <@rms> | Maybe, I've never actually used debian proper. I /do/ know that the postfix and AIDE brain-damage is directly attributable to the debian packages though since I made the assumption at one point that debian would do things more sanely than ubuntu. But nope, exact same packages. |
12:44 | <@froztbyte> | Azash: "server OS for people who are scared of new things" |
12:45 | <@rms> | CTO's reason "because I wanted to make sure it'd run without hardware issues and I'm used to Debian" |
12:45 | < Syka> | Azash: i think the reason it exists is to sell Landscape licenses or something |
12:45 | <@rms> | Yes, the CTO was a WTF-generator. |
12:45 | | * Azash advocates installing gentoo for everything |
12:45 | <@froztbyte> | bad idea |
12:46 | <@froztbyte> | Syka: that was the long-term play |
12:46 | < Syka> | why did i try and highlight that |
12:46 | <@froztbyte> | gentoo is terribad for server maint |
12:46 | <@rms> | Azash: you can setup gentoo to use distcc for compilations right? |
12:46 | <@froztbyte> | yes |
12:46 | < Syka> | i swear tvtropes has conditioned me |
12:46 | <@Azash> | Yeah, but my remark was made in jest |
12:46 | <@froztbyte> | and even to just pull binpackages |
12:46 | < Syka> | to highlight all the black blocked text |
12:46 | <@Azash> | I hoped the spoiler tags would make that obvious |
12:47 | <@Azash> | Syka: It is a spoiler tag |
12:47 | <@rms> | Ah, cause I know another sysad that just runs gentoo boxes |
12:47 | <@froztbyte> | the really nice thing is the about gentoo is the build system |
12:47 | < Syka> | it... is? |
12:47 | < Syka> | my irssi just shows a big box |
12:47 | <@Azash> | Copy it and paste somewhere |
12:47 | <@froztbyte> | it's by far the nicest buildsystem I've come across |
12:47 | <@Azash> | (I usually paste into the send buffer) |
12:47 | < Syka> | oh |
12:47 | < Syka> | well i feel like an idiot |
12:47 | <@froztbyte> | but beyond that....meh |
12:47 | < Syka> | 'why does this bit of text look like it's a spoiler section?' 'because it is' |
12:48 | | * Azash pats syka |
12:48 | < Syka> | I have never installed Gentoo before |
12:48 | <@Azash> | I set the fore- and background colour to black so hilighting won't do anything in irssi |
12:48 | <@Azash> | People make it seem terrible but it's pretty neat, as long as everything goes according to plan |
12:48 | < Syka> | all my hardware is either a) underpowered or b) uses Nvidia |
12:48 | <@rms> | There's a fork of it now |
12:49 | < Syka> | and I honestly do not want to go through compiling drivers with nv support |
12:49 | < Syka> | i did it in like 2009 and it was like forcing your face through an electric cheese grinder made of /why/ |
12:49 | <@Azash> | WHY |
12:50 | | * Azash looks back fondly on having to go to sleep after he had the bright idea to emerge kde-meta |
12:50 | < gnolam> | Colors. Really? |
12:50 | < Syka> | once i compiled libre office |
12:50 | < Syka> | ...on a netbook |
12:50 | <@Azash> | Well, I figured syka was going for the same thing so I thought I'd put it there |
12:51 | < Syka> | and then it /errored/ halfway through |
12:51 | <@Azash> | Sorry if it bothered |
12:51 | < Syka> | i dont know shit about irc colours :P |
12:51 | < Syka> | 90% of the rooms I go to have it turned off |
12:51 | < Syka> | after the first few times someone discovered gay.pl |
12:51 | <@Azash> | Oh dear |
12:51 | | * AnnoDomini patpats Azash. I thought your meme was well-executed. |
12:51 | < Syka> | imagine giant ascii dicks |
12:52 | < Syka> | as far as the eye can see |
12:52 | < Syka> | and all the moderators are asleep |
12:52 | <@Azash> | AnnoDomini: The Kornheiser meme is my favourite so I got excited~ |
12:52 | < Syka> | there was a kornheiser of a microsoft advertisement i saw the other day |
12:52 | < Syka> | it was some office guy grinning with 'maximum speed' below it |
12:53 | < Syka> | it was hilarious for absolutely no reason |
12:53 | <@Azash> | It seems like one of those old 80's computer ads, but worse |
13:15 | <@Azash> | https://crypto.stanford.edu/~dabo/pubs/abstracts/ssl-client-bugs.html |
13:18 | < RichyB> | Azash: sounds like a good idea to me, actually. |
13:18 | < RichyB> | "install Gentoo to solve your problems" isn't necessarily the best idea in the world. |
13:19 | < RichyB> | "If anyone bugs you about a problem, suggest: \"install Gentoo to solve your problems\"." is going to be pretty effective |
13:19 | < RichyB> | Effective in that the complainer will soon find themselves stuck on a half-broken Gentoo box, unable to get their wifi connection working in order to come bug you again for more help on IRC. ;) |
13:19 | <@Azash> | Haha |
13:21 | < Syka> | problems that gentoo legitimately fixes: |
13:21 | < Syka> | - my computer has too much /sound/! |
13:21 | <@Azash> | That's probably not gentoo's fault |
13:21 | < Syka> | - my computer's working graphics scare me :( |
13:22 | < Syka> | - I'm a Windows convert and software installing in under 45 minutes is new and terrifying for me |
13:22 | < Syka> | :P |
13:22 | <@Azash> | Probably giving gentoo a bit of a bad rap there though, except for the installation part |
13:23 | < Syka> | yeah, Gentoo isn't That Bad |
13:23 | < Syka> | I'm just being facetious :P |
13:25 | < RichyB> | I've met a guy who was running a PaaS service that did really well by using Gentoo. |
13:26 | < RichyB> | Suited them really well because it was actually worth the price (in sysadmin-hours) to spend a while twiddling with the base Gentoo image that they were deploying on their guests. |
13:27 | < RichyB> | (since they were deploying really large numbers of guests) |
13:27 | | * Azash snerks at trolls. http://steamcommunity.com/app/221410 |
13:39 | <@rms> | http://trenchescomic.com/tales/post/but-what-if-they-stab-me |
13:51 | < gnolam> | rms: ... there's just no way anyone could take that as anything other than sarcasm. |
13:52 | <@rms> | Yeah |
13:57 | | celticminstrel [celticminst@Nightstar-05d23b97.cable.rogers.com] has joined #code |
14:10 | | Kindamoody|out is now known as Kindamoody |
14:46 | | Kindamoody is now known as Kindamoody|out |
15:28 | <@Azash> | EvilDarkLord: Out of sheer curiosity, do you know anyone named Ylikotila? |
15:33 | <@EvilDarkLord> | Azash: Doesn't ring a bell. |
15:36 | | * Azash nods * |
15:50 | < celticminstrel> | Why is there a * on the end of that action? |
15:51 | < gnolam> | It's a pointer to nodding. |
15:51 | < gnolam> | Obviously. |
15:51 | < celticminstrel> | XD |
15:52 | <@Tamber> | cnods? |
15:53 | <@Azash> | celticminstrel: To close the emote |
15:53 | <@Azash> | I forgot the bold though :( |
15:53 | < celticminstrel> | But it doesn't. |
15:53 | | * Azash frowns * |
15:53 | < celticminstrel> | [Oct 25@10:36:46am] o Azash nods * |
15:54 | < celticminstrel> | At least, not universally. |
15:57 | | celticminstrel [celticminst@Nightstar-05d23b97.cable.rogers.com] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] |
16:39 | | Syka is now known as syksleep |
17:17 | | Kindamoody|out is now known as Kindamoody |
17:24 | | gnolam is now known as AbominableGnolam |
18:12 | <@Tarinaky> | I have a couple of days before I have to start on any assignments again. |
18:12 | <@Tarinaky> | So I've decided to download and install UDK. |
18:12 | <@Tarinaky> | Suggestions? |
18:12 | <&McMartin> | UDK = ? |
18:13 | <@Tarinaky> | Unreal Development Kit. |
18:13 | <@Tarinaky> | From what I gather it's like a free version of the Unreal Engine. |
18:13 | <&McMartin> | Ah, OK, I guessed wrong~ |
18:16 | <@Tarinaky> | I have no idea what I'm supposed to do with this though. |
18:17 | <&McMartin> | "Play with what I thought it was - the Unity devkit - instead?" |
18:30 | <@Tarinaky> | I don't really see how that solves the issue of ignorance. |
18:30 | <@Tarinaky> | It just displaces it. |
18:30 | <&McMartin> | Unity has getting started guides and lots of free content on their site to mash up. |
18:49 | <@Tarinaky> | This seems to have more gun though. |
18:51 | <@Azash> | Doing unity first leaves you some stepping stones though |
18:51 | <@Azash> | Just as starting philosophy with Nietzsche, doable but perhaps disproportionately challenging |
18:56 | <@Azash> | 20:52 < Crank> if("foo" == TRUE && "foo" == 0 && TRUE != 0) echo "yay php"; |
19:02 | <@iospace> | heh |
19:03 | <@iospace> | that would work in C too I believe |
19:06 | <@Azash> | Could "foo" equal 0? |
19:08 | <@Azash> | From elsewhere: "Credible defense of PHP means using it a non-trivial time, which means you are not fully sane, which makes a credible defense impossible. QED." |
19:11 | <@iospace> | oh, sorry, i was looking at TRUE != 0 XD |
19:13 | <@Tarinaky> | There're lots of tutorials for level making in UDK. Harder to find help on what to do to get started with scripting. |
19:32 | < RichyB> | Yeah, that could work in C. |
19:32 | < RichyB> | You'd have to have both of the string constants ("foo" and "foo") allocated at different addresses |
19:33 | < RichyB> | by luck, one of them would be allocated at 0, and the other at some address which is equal to whatever TRUE contains |
19:33 | < RichyB> | obviously that's very silly |
19:34 | < RichyB> | but it would not be against the rules for a C compiler to output microcode where (assume that echo statement gets changed to a printf call) "yay php\n" gets printed. |
19:34 | <&ToxicFrog> | Doesn't the C standard guarantee that (a) NULL always compares equal to 0 and (b) no other pointer does? |
19:35 | < RichyB> | I'm almost certain that the C standard does not guarantee that no other pointer compares equal to 0. |
19:35 | | AbominableGnolam is now known as gnolam |
19:37 | < RichyB> | e.g. consider mmap(0, 4096, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_FIXED | MAP_ANONYMOUS, -1, 0) |
19:37 | < RichyB> | or some microprocessor or DSP where your address space is a few hundred bytes starting from 0. |
19:39 | <&ToxicFrog> | Point. |
20:18 | | Derakon [chriswei@Nightstar-a3b183ae.ca.comcast.net] has joined #code |
20:20 | <@iospace> | i'm having a lot of derps today |
20:20 | | * iospace head desks |
20:21 | | * Azash lends iospace some antiderps |
20:22 | | Vash [Vash@Nightstar-3ba4108e.wlfrct.sbcglobal.net] has joined #code |
20:22 | | mode/#code [+o Vash] by ChanServ |
20:25 | | * Derakon kneels on a big rubber ball rather than sit in a chair. |
20:25 | < Derakon> | It may look silly, but my hip doesn't hurt any more! |
20:25 | < Derakon> | Plus I can bounce along to my music~ |
20:29 | <@Azash> | ~ |
20:30 | < Derakon> | Ending lines with a tilde is a verbal tic I picked up in #fleet, where it's used to indicate sarcasm, levity, or other "not really serious" lines. |
20:30 | < Derakon> | Works better than peppering your lines with smileys, let alone the dreaded "lol". |
20:34 | <@Azash> | Lol is pretty stigmatized, unfortunately |
20:34 | < Derakon> | s/un// |
20:34 | <@Azash> | How else am I meant to express amusement? |
20:34 | < Derakon> | I find a simple "Heh" does the trick. |
20:34 | < Derakon> | If you're actually laughing, then more descriptive emotes can be used. |
20:35 | <@Azash> | On a sidenote, from what I figure from hanging out around /g/ channels, the tilde is from the wannabe-japanese culture espoused there |
20:35 | < Derakon> | The problem with "lol" is the massive disjoint between the degree of emotion felt and the degree of emotion stated to be felt. |
20:35 | <@Azash> | Do you know the webcomic Gunnerkrigg Court? |
20:35 | < Derakon> | I do. |
20:35 | <@Azash> | Have you seen the writer's reaction to 'lol' ? |
20:36 | < Derakon> | Don't think so. |
20:36 | <@Azash> | He dislikes internetspeak, so if you type 'lol' on the GKC forums, it's replaced by 'LAUGHING ON LINE' |
20:45 | <&McMartin> | "lol is not a punctuation mark" |
20:46 | <&McMartin> | I find ~ is best interpreted as "ha ha only serious" |
20:46 | | * Tarinaky speaks in a deep monotone... "I am literally laughing the hardest I have ever laughed." |
20:46 | | * Tarinaky does not smile as he says this. |
20:46 | <@Tarinaky> | "My sides. They are splitting." |
20:46 | | Kindamoody is now known as Kindamoody[zZz] |
20:46 | <&McMartin> | WITH AXE |
20:47 | < Derakon> | URRGZOB IS TEN COMEDIANS |
20:47 | <&McMartin> | Hooray |
20:51 | <&ToxicFrog> | URRGZOB IS TEN RISC PROCESSOR CORES |
20:51 | | * ToxicFrog carves his way through another section of the Related Work |
20:51 | | * iospace smacks ToxicFrog with a dopefish |
20:51 | <@iospace> | silence wench! |
20:55 | | * ToxicFrog shift-registers iospace into the frame memory grid |
20:59 | <@iospace> | :< |
21:02 | | * Derakon eyes the wxPython documentation. |
21:02 | < Derakon> | Known subclassese of App: App, App, App, App, App, App, InspectableApp, PySimpleApp, PyWidgetTester |
21:02 | <~Vornicus> | well then |
21:12 | <&McMartin> | App, App, WonderfulApp, GloriousApp, BakedBeansAndApp |
21:20 | < gnolam> | Azash: a forum I hang out on used to replace "lol"s with "I'm dumb!". |
21:21 | < gnolam> | Seeing it meant that it was true. |
21:23 | | celticminstrel [celticminst@Nightstar-05d23b97.cable.rogers.com] has joined #code |
21:39 | | * Derakon watches Windows hang for ~15s while trying to copy a 1kB file off of a USB thumb drive. |
21:41 | < Derakon> | Bleh, the small demo program doesn't actually reproduce the bug I'm trying to fix in the 10k-line program. |
21:41 | < Derakon> | (That being, when the program closes, it doesn't actually close, leaving the console window lying around and holding onto any handles it had seized during operation) |
22:13 | | * Derakon tracks it down to three things: two windows and a network connection. |
22:13 | < Derakon> | The network connection's easy enough to clean up. |
22:13 | < Derakon> | The windows, though, I've no idea why they'd be problematic. |
22:14 | < Derakon> | One of 'em's the camera view window and the other's the window for the mosaic, so they have OpenGL in common, but a third window uses OpenGL plenty and it doesn't make closing the program hang. |
22:20 | <&McMartin> | Is that third window marked kill-process-on-close? |
22:20 | < Derakon> | No; only the main window is, and it's not any of those three. |
22:21 | < Derakon> | Hm, okay, the camera view window is problematic because it creates a thread that's not going away. |
22:24 | < Derakon> | Which is probably because I'm not using timeouts on my "wait for something to happen" bits. |
22:27 | < Derakon> | ...nope, adding timeouts didn't help any. |
22:28 | < Derakon> | Of course, I'm not doing anything different; just looping, with the timeout, until I get something to work with. |
22:30 | < Derakon> | Okay, WTF. |
22:30 | < Derakon> | My main module basically consists of "Make the app, then start its main loop". |
22:30 | < Derakon> | I just put a print statement after that last line, and I see it printed when I try to close the app. |
22:31 | < Derakon> | But the console window still doesn't close. |
22:31 | < Derakon> | So, despite the main thread of execution being done, the program doesn't end! |
22:32 | < Derakon> | ...ah, daemon threads. |
22:35 | < Derakon> | ...no, modifying the thread to be non-daemon didn't fix it either. |
22:36 | | RichyB [richardb@Nightstar-3b2c2db2.bethere.co.uk] has quit [[NS] Quit: Leaving] |
22:37 | < Derakon> | Got the boolean backwards. You want daemon to be True, not False. |
22:37 | < Derakon> | To be fair, Python's docs on this are badly-phrased. |
22:37 | < Derakon> | "The entire Python program exits when no alive non-daemon threads are left." |
22:38 | < Derakon> | Hooray, closing the main window now cleanly exits the program! |
22:51 | <@iospace> | :D |
23:11 | <@Reiv> | Derakon: That's... a bit double-negative there. |
23:12 | | Vash [Vash@Nightstar-3ba4108e.wlfrct.sbcglobal.net] has quit [[NS] Quit: I lovecraft Vorn!] |
23:15 | | * AnnoDomini cannibalizes an earlier project for an easy saving mechanism. |
23:15 | <@AnnoDomini> | I sure love struct dumping. |
23:18 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has quit [Ping timeout: 121 seconds] |
23:19 | < gnolam> | As in... fwriting an entire struct? :P |
23:19 | <&McMartin> | I hope that's python import struct |
23:19 | < gnolam> | If so, you're a bad person and should be ashamed of yourself. |
23:19 | <&McMartin> | fwriting structs stopped being viable once 64-bit architectures became paramount. |
23:19 | < gnolam> | It was never viable to begin with. |
23:19 | <&McMartin> | (Also IFF 4 lyfe) |
23:20 | <&McMartin> | gnolam: 20 of the past 30 years of personal computer history give this the lie. |
23:21 | <&McMartin> | Anyway, struct dumping in python also specifies endianness and word size |
23:21 | < gnolam> | Practical experience says otherwise. |
23:21 | < gnolam> | (No, not my code.) |
23:22 | <&McMartin> | Yes, it fails because you can't transport it. |
23:22 | <@AnnoDomini> | I recall someone saying that struct dumping can be compatible across 32/64 bit architectures if one keeps array sizes divisible by 8 or something like that. |
23:22 | < gnolam> | So I know for a fact that MSVC and GCC at least used to pad differently. |
23:22 | <&McMartin> | AnnoDomini: No, that's guessing and hoping the compiler doesn't put in extra buffers. |
23:23 | <&McMartin> | gnolam: Sure, and that was irrelevant from between about 1982-1995 on the DOS platform. |
23:23 | < Derakon> | My complaint with custom binary formats is that they aren't compellingly better than text files. |
23:23 | < Derakon> | Or, if you're that worried about disk space, compressed text files. |
23:23 | <&McMartin> | Derakon: Storing IEEE floats exactly. |
23:23 | <&McMartin> | That's about it. |
23:23 | < Derakon> | ...mm, point. |
23:23 | <&McMartin> | And then you can use hex-encoded ones, yes. |
23:23 | < gnolam> | Derakon: They're a heck of a lot faster. |
23:23 | <&McMartin> | Oh, and yeah. |
23:24 | <&McMartin> | The best thing to do is to process text once on install/compile time and then do accelerated binary load later. |
23:24 | < Derakon> | gnolam: most programs do very little saving/loading. |
23:24 | <&McMartin> | Saves on marshal/unmarshal costs |
23:24 | <&McMartin> | ... though that's more important for comm protocols |
23:25 | < Derakon> | Anyway, the main point is that custom binary formats are a chump's game. Let someone else deal with the hassle of writing bug-free loaders/unloaders. |
23:25 | < gnolam> | I'm pretty sure Anno's context is game programming. |
23:25 | < Derakon> | Even then! |
23:25 | <&McMartin> | Yeah, I'd say "nethack's level files are the way to go there"~ |
23:25 | < gnolam> | In which case, you have a _lot_ of data to load and speed matters. |
23:26 | <@AnnoDomini> | Are there, like, any libraries for C that allow easy struct-to-file and back again transfers? |
23:26 | < Derakon> | In game programming, your stuff to load is going to be one of a) game assets, which are already binary and have well-defined loaders for you to use, or b) your game structure formats, which for most games are going to be very small. |
23:27 | < Derakon> | When I was working on Jetblade, the amount of time I spent on loading sprite image data (PNG format) vastly dwarfed the amount of time I spent loading level definition data (text files). Granted that Jetblade had procedurally-generated levels, but even when loading a pre-set level it didn't take noticeable time to parse the text. |
23:28 | < Derakon> | Is there a place for binary data formats? Absolutely. Should you put off transitioning to binary formats as long as possible? Also absolutely. |
23:41 | | Derakon [chriswei@Nightstar-a3b183ae.ca.comcast.net] has quit [[NS] Quit: leaving] |
23:59 | <~Vornicus> | most of the time you don't need a properly binary format. |
--- Log closed Fri Oct 26 00:00:15 2012 |