| --- Log opened Fri Aug 03 00:00:07 2012 |
| 00:05 | | Vash [Vash@Nightstar-e8057de2.wlfrct.sbcglobal.net] has quit [[NS] Quit: I lovecraft Vorn!] |
| 00:20 | <&jerith> | http://i.imgur.com/Qxkj4.png |
| 00:21 | <&jerith> | ToxicFrog: I use irsii/screen/debian and then iTerm/ssh/OSX to connect to it. |
| 00:21 | <&jerith> | Or connectbot/android. |
| 00:21 | <&jerith> | Or, until my netbook broke and was then stolen, xterm/ssh/ubuntu. |
| 00:21 | | * TheWatcher eyes that message |
| 00:22 | | Nemu [NeophoxProd@Nightstar-cf9cbbe2.asahi-net.or.jp] has quit [Ping timeout: 121 seconds] |
| 00:22 | <&jerith> | I also use Adium for work stuff. It's a pretty crap IRC client, but it's a great IM client and I'd rather not have /another/ app that I'll forget to start. |
| 00:24 | <&ToxicFrog> | McMartin: it's not, no >.M |
| 00:24 | <&ToxicFrog> | I just installed it to /opt |
| 00:28 | | You're now known as TheWatcher[t-2] |
| 00:55 | | iospace is now known as io\gone |
| 00:57 | | You're now known as TheWatcher[zZzZ] |
| 01:02 | | Derakon[AFK] is now known as Derakon |
| 02:01 | | * McMartin may end up installing it to ~/bin |
| 02:11 | | * Derakon ponders Pyrel, tries to decide how to handle the next bit. |
| 02:11 | <&Derakon> | "The next bit" being I want to write the code that allows the player to examine the game map. I.e. the look command. |
| 02:16 | | io\gone is now known as iospace |
| 02:23 | | Syloq_Home [Syloq@NetworkAdministrator.Nightstar.Net] has quit [Client closed the connection] |
| 02:23 | | Attilla [Obsolete@Nightstar-40716bea.as43234.net] has quit [Ping timeout: 121 seconds] |
| 02:48 | | Syloq_Home [Syloq@NetworkAdministrator.Nightstar.Net] has joined #code |
| 02:59 | <~Vornicus> | What kind of data do you get off of Angband when you do that? |
| 03:10 | <&Derakon> | The contents of the tile under inspection. |
| 03:10 | | Kindamoody[zZz] is now known as Kindamoody |
| 03:10 | <&Derakon> | But the command has multiple modes of operation. By default, it skips over boring tiles -- so when you press a direction, it picks the nearest interesting tile vaguely in that direction. |
| 03:11 | <&Derakon> | Or you can toggle it into exact mode, in which case you just move a cursor about the map and it'll show you what's under the cursor. |
| 03:11 | <&Derakon> | (It doesn't tell you anything that's hidden, e.g. secret doors) |
| 03:18 | <~Vornicus> | So the big deal here is, "skip over boring tiles" |
| 03:19 | <&Derakon> | Well, that and how to structure things internally. |
| 03:19 | <&Derakon> | My current "prompt the user for input" system is in hindsight rather flawed. |
| 03:19 | <~Vornicus> | Things have flags: "boring" - you can use this to mark different types of floor, if you so desire, so you can have cave floor and smooth floor, or gribblies - and "hidden", which is used to hide secret doors. |
| 03:19 | <&Derakon> | Yeah, I have that already. |
| 03:19 | | Vash [Vash@Nightstar-e8057de2.wlfrct.sbcglobal.net] has joined #code |
| 03:19 | | mode/#code [+o Vash] by ChanServ |
| 03:19 | <&Derakon> | I figure I'll just project out 90? cones in the direction the user wants to look in and iterate over non-boring-marked stuff. |
| 03:51 | | iospace is now known as iospacedout |
| 04:08 | | Rhamphoryncus [rhamph@Nightstar-5697f7e2.abhsia.telus.net] has joined #code |
| 04:39 | <&ToxicFrog> | McMartin: I mean, I installed 'clojure' to ~/bin/ and all of the actual jars and stuff to /opt/clojure |
| 04:51 | <&McMartin> | 7j |
| 04:52 | <&McMartin> | Right |
| 04:52 | <&McMartin> | I just stuffed the jars into ~/bin too because what the Hell |
| 05:14 | | * Derakon ponders his Prompt system, tries to decide if he wants to try to subclass Prompt from Thing and store them in the GameMap's normal database, or create a special entry for them. |
| 05:14 | <&Derakon> | To expound, GameMap has a mapping of keys to sets of Things, so you can ask it for e.g. all of the Things that accept keyboard input, or all of the Things that are Items. |
| 05:14 | <&Derakon> | But since these are sets, they are unordered, and Prompts have an inherent order because they can be layered |
| 05:15 | <&Derakon> | (A Prompt is any request from the user for extra input, along with associated metadata, e.g. "which item do you want to throw" with a list of throwable in-inventory items) |
| 05:15 | <&Derakon> | Prompts also need to intercept input, which would ordinarily go to all Listener Things. |
| 05:16 | <&Derakon> | It's a bit galling to have to store Prompts in a special container in the GameMap...but alternatives that would preserve their heirarchy seem even worse. |
| 05:17 | <&Derakon> | (E.g. assigning a priority to each Prompt, thus when you want to find the top Prompt you have to iterate over them all...and heaven help you if you get two Prompts with the same priority) |
| 05:17 | <&Derakon> | I suppose the GameMap could just store the top Prompt, and let it hold all Prompts beneath it in a recursive datastructure. |
| 05:21 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has quit [Ping timeout: 121 seconds] |
| 05:34 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has joined #code |
| 05:34 | | mode/#code [+o himi] by ChanServ |
| 06:48 | | Derakon is now known as Derakon[AFK] |
| 07:04 | | Vash [Vash@Nightstar-e8057de2.wlfrct.sbcglobal.net] has quit [[NS] Quit: I lovecraft Vorn!] |
| 07:16 | <&jerith> | You have Things and Items? |
| 07:16 | <&jerith> | <3 |
| 07:16 | <&jerith> | Also </3 |
| 07:17 | <&jerith> | (I have a project with that kind of hierarchy in it.) |
| 07:32 | < Rhamphoryncus> | Oh hey, no bugs |
| 07:32 | < Rhamphoryncus> | After 20 minutes of manually replacing one set of index numbers with another set, in a couple different places, followed by reordering elements to match the new indexes.. that's rather surprising |
| 08:04 | | * Rhamphoryncus looks at pentakis dodecahedron |
| 08:06 | < Rhamphoryncus> | A dodecahedron is my base shape, but since I'm also handling the centroid of each pentagon and adjusting it outwards, I could instead think of it as a pentakis dodecahedron |
| 08:09 | < froztbyte> | haha |
| 08:09 | < froztbyte> | that's always fun when it works |
| 08:09 | < froztbyte> | (the replace thing) |
| 08:13 | < Rhamphoryncus> | Yeah, I expected to spend an hour going through different parts to fix the ones I did wrong |
| 08:13 | | EvilDarkLord [jjlehto3@Nightstar-a5db08cc.org.aalto.fi] has quit [Operation timed out] |
| 08:14 | < froztbyte> | have you seen that Brett Whatsisface talk of Inventing On Principle? |
| 08:14 | < Rhamphoryncus> | nope |
| 08:14 | < froztbyte> | hmm |
| 08:14 | < froztbyte> | will get link now |
| 08:14 | < froztbyte> | but one of the key points he goes on about is tighter feedback loops |
| 08:14 | < froztbyte> | so that you get better visibility of the innards of your app |
| 08:15 | < Rhamphoryncus> | huh |
| 08:15 | < Rhamphoryncus> | I have no idea what that means :D |
| 08:15 | < froztbyte> | which reminded me of my way of writing things in high school (sprinkle debug prints /everywhere/, trace each bit of data as it changes) |
| 08:15 | < froztbyte> | Rhamphoryncus: ah, you might enjoy this then |
| 08:15 | < froztbyte> | sec |
| 08:15 | < froztbyte> | http://vimeo.com/36579366 |
| 08:17 | < Rhamphoryncus> | no flash, won't load |
| 08:18 | < Rhamphoryncus> | ahh youtube version |
| 08:19 | | * Rhamphoryncus activates his collection of bodges and gets it playing ;) |
| 08:20 | < froztbyte> | haha |
| 08:20 | < froztbyte> | youtube-dl ftw |
| 08:20 | < froztbyte> | that thing downloads /so many/ sites |
| 08:20 | < froztbyte> | s/ds /ds from / |
| 08:22 | < Rhamphoryncus> | I have the hd suite greasemonkey script. I've used others over the years, they either don't work or break after a while. hd suite is my current survivor |
| 08:23 | < Rhamphoryncus> | It doesn't make the videos play in the browser (google has that turned on half the time), but let's me download. However, I don't want to wait for the download to finish, so I browse to /tmp and open it up |
| 08:23 | < Rhamphoryncus> | Lamely, this works much better than streaming a file directly in totem, as it tries to be all 'leet with streaming and buffer and shit, while mostly falling on its face |
| 08:23 | < froztbyte> | youtube-dl is nice because the guy appears to keep it up to date pretty well, and you can just roll a checked-out version |
| 08:24 | < froztbyte> | where an update is just a git pull away! |
| 08:24 | < froztbyte> | (it also has a -U self-update flag) |
| 08:24 | < froztbyte> | Rhamphoryncus: fuck totem. |
| 08:24 | < froztbyte> | totem is one of the things that led the fight in the GTK flagship retardation |
| 08:24 | < Rhamphoryncus> | totem works better than vlc in the general case. For instance, the zoom mostly works |
| 08:25 | < froztbyte> | how doesn't that work on vlc? |
| 08:25 | < froztbyte> | you're not running gentoo with weird useflags or something, are you? :P |
| 08:26 | < Rhamphoryncus> | Nope, boring ubuntu |
| 08:26 | < froztbyte> | ah, that's nearly as broken |
| 08:27 | < Rhamphoryncus> | I can't recall the specifics, but it chops off parts of the image in a nonsensical way |
| 08:27 | < froztbyte> | (mostly because they don't have the manpower to deal with things) |
| 08:27 | < Rhamphoryncus> | Oh, aspect ratio is broken too, closely related |
| 08:27 | < froztbyte> | Rhamphoryncus: hmm |
| 08:27 | < froztbyte> | lemme know when you run into it again? |
| 08:28 | < froztbyte> | I've used vlc for quite a bunch of stuff, and can't remember really ever having any major unsolvable issues |
| 08:28 | < Rhamphoryncus> | Huh. Looks to be working on that video, heh |
| 08:29 | < Rhamphoryncus> | I do remember trying to chop off letterboxing |
| 08:31 | < Rhamphoryncus> | Aha, if I open a dog whisperer video the crop only makes the window wider/narrower while changing the aspect ratio to match |
| 08:32 | < Rhamphoryncus> | With the talk the ratio remains unchanged, it just cuts off the right side |
| 08:32 | < froztbyte> | hmm |
| 08:32 | < froztbyte> | and what screensize is this? |
| 08:34 | < Rhamphoryncus> | windowed, but maximized or fullscreen on my 1680x1050 are just as bad |
| 08:39 | <&McMartin> | [programming] vaporware says, "There are 28 places in the Inform source where keyword sets are switched in or out." |
| 08:39 | <&McMartin> | :gonk: |
| 08:41 | < Rhamphoryncus> | heh |
| 08:44 | | * Rhamphoryncus fixes his vlc UI to include time >.> |
| 09:26 | | Attilla [Obsolete@Nightstar-ac56e080.as43234.net] has joined #code |
| 09:31 | < Rhamphoryncus> | froztbyte: hrm, interesting talk. I see 3 parts to it: the idea of following a principle, his principle of code having immediate feedback, and his methods of providing that feedback |
| 09:32 | < froztbyte> | yar |
| 09:32 | < froztbyte> | the feedback part is /very/ hard to get right |
| 09:32 | < froztbyte> | because nearly no tools currently do that for anything |
| 09:32 | < Rhamphoryncus> | Yup. It's the weakest in his whole talk |
| 09:32 | < Rhamphoryncus> | Those tools are great.. for a graphic artist using programming as their medium |
| 09:32 | < froztbyte> | I think the closest is "break out to repl" |
| 09:32 | < Rhamphoryncus> | They're useless for programming |
| 09:33 | < froztbyte> | not necessarily |
| 09:33 | <&McMartin> | Repls? Nah, they're pretty great for prototyping |
| 09:33 | < froztbyte> | think of the case where you have data inputs and lots of munging and whatnot |
| 09:33 | < Rhamphoryncus> | But it's NOT like it hasn't been tried. Many have tried to do graphical programming in some fashion. That's the problem though: programming is abstract, it's pure, so there's no intrinsic visual form to tie it back to |
| 09:33 | < froztbyte> | being able to step through that flow easily, and breaking in at the middle |
| 09:33 | < froztbyte> | that'd be pretty cool |
| 09:34 | < Rhamphoryncus> | It would, but it's hard to get right |
| 09:34 | <&McMartin> | Rhamphoryncus: Mindstorms~ |
| 09:34 | < Rhamphoryncus> | hmm? |
| 09:34 | <&McMartin> | The biggest problem with graphical programming formats is meaningfully serializing them. |
| 09:34 | <&McMartin> | LEGO Mindstorms are a graphical programming language. |
| 09:34 | < froztbyte> | hell, a good example would be from my day job: when I need to tweak a $something (caches are the most common target), lots of "nail, meet BIG HAMMER" going on |
| 09:35 | < froztbyte> | (we build caches that work at ISP level) |
| 09:35 | <&McMartin> | YoYo Games's Game Maker system also uses a graphical programming language that could be but *isn't* 1-1 translatable with (a subset of) its scripting language |
| 09:35 | < Rhamphoryncus> | How is lego mindstorm applicable to writing a compiler? They're not. They, just like graphics, are using programming as the medium |
| 09:35 | <&McMartin> | Um |
| 09:36 | <&McMartin> | Are you just swapping "compiler" for "operating system" in the long-discredited, 30+-year-old rants about how certain languages aren't Real? |
| 09:37 | < Rhamphoryncus> | no |
| 09:37 | <&McMartin> | Alt: It's got while loops and assignable variables, so I'm pretty sure writing a compiler with Mindstorms is theoretically possible |
| 09:38 | <&McMartin> | LEGO Mindstorms is a graphical programming language that compiles to instructions to an ARM processor that animates LEGO Robots. |
| 09:38 | < Rhamphoryncus> | To get value out of a visual or physical medium you'd need to invent the format as you go, not rely on existing animation tools |
| 09:38 | <&McMartin> | And by "animates" I mean "controls motors and actuators". |
| 09:39 | < Rhamphoryncus> | Lego mindstorm is one problem domain. Programming is a set of many problem domains |
| 09:39 | <&McMartin> | Aaaaaaand, now you're back in the 30+ year old rant. |
| 09:39 | < Rhamphoryncus> | Not sure which rant you refer to |
| 09:39 | <&McMartin> | You can't write an OS with this graphical programming language, so it is not a real programming language. |
| 09:39 | <&McMartin> | s/graphical/garbage-collected/ |
| 09:39 | <&McMartin> | s/graphical/interpreted/ |
| 09:40 | <&McMartin> | s/graphical/dynamically-typed/ |
| 09:40 | <&McMartin> | ad nauseam |
| 09:41 | < Rhamphoryncus> | If you're picking out 1 problem domain and saying it's mandatory, then yes, it's a lame argument. There's many more where your new methods will pay off |
| 09:41 | <&McMartin> | You appear to be doing that with compilers |
| 09:41 | < Rhamphoryncus> | I'm not, it's the other way around. There's only 1 problem domain where your new methods work, in every other it does not |
| 09:41 | < Rhamphoryncus> | That was just an example |
| 09:41 | <&McMartin> | Yeah, uh |
| 09:42 | <&McMartin> | REPL-based programming is nearly as old as programming at all, you can actually get a shitload of mileage out of it in the compiler space |
| 09:42 | < Rhamphoryncus> | The talk itself gave several distinct methods for distinct problems |
| 09:43 | < Rhamphoryncus> | REPL isn't a good method by the principle in that talk |
| 09:43 | < Rhamphoryncus> | And I've got nothing against REPL. I use it in python all the time |
| 09:44 | < ShellNinja> | http://www.codinghorror.com/blog/2012/08/today-is-goof-off-at-work-day.html |
| 09:45 | < Rhamphoryncus> | I think what the talk is leading to is the need for a visualization toolkit, so you can have a diagnostic script paired with your program that tells the environment what to track for the time dimension (or multiple such dimensions), how to visualize it, that sort of thing |
| 09:48 | < Rhamphoryncus> | That's the point I'm trying to make. The talk only gives the problem, it doesn't provide the solution. The solution is something we need to discover. |
| 10:04 | | You're now known as TheWatcher |
| 10:17 | | * Rhamphoryncus tries to figure out the angles of a dodecahedron |
| 10:18 | < Rhamphoryncus> | or of a pentakis dodecahedron, but I can bodge that from a normal dodecahedron |
| 10:32 | | * TheWatcher eyes emacs 24 |
| 10:33 | <@TheWatcher> | wtf have they broken in hs-minor-mode |
| 10:47 | | RichyB [richardb@Nightstar-3b2c2db2.bethere.co.uk] has joined #code |
| 11:39 | | RichardB [richardb@Nightstar-3b2c2db2.bethere.co.uk] has joined #code |
| 11:41 | | RichyB [richardb@Nightstar-3b2c2db2.bethere.co.uk] has quit [Ping timeout: 121 seconds] |
| 12:11 | | RichardB [richardb@Nightstar-3b2c2db2.bethere.co.uk] has quit [[NS] Quit: Leaving] |
| 12:12 | | RichyB [richardb@Nightstar-3b2c2db2.bethere.co.uk] has joined #code |
| 12:34 | < Rhamphoryncus> | Scotland just formed GB :O |
| 12:34 | < Rhamphoryncus> | err wrong channel heh |
| 13:03 | | * TheWatcher vaguely stabs virtualbox's clipboard integration code for being as reliable as an amnesiac goldfish |
| 13:33 | | RichardBarrell_m [richardbarr@3EB374.80E115.4FC5A1.49A56F] has joined #code |
| 13:33 | | RichardBarrell_m [richardbarr@3EB374.80E115.4FC5A1.49A56F] has quit [[NS] Quit: Bye] |
| 13:36 | | iospacedout is now known as iospace |
| 14:37 | < froztbyte> | haha |
| 15:11 | | Rhamphoryncus [rhamph@Nightstar-5697f7e2.abhsia.telus.net] has quit [Client exited] |
| 15:22 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has quit [Ping timeout: 121 seconds] |
| 15:27 | | Kindamoody is now known as Kindamoody|out |
| 16:31 | <~Vornicus> | Rham: what angles precisely are you trying to f--- he's not here |
| 17:54 | | Kindamoody|out is now known as Kindamoody |
| 18:18 | | Kindamoody is now known as Kindamoody|gaming |
| 18:24 | | RichyB [richardb@Nightstar-3b2c2db2.bethere.co.uk] has quit [[NS] Quit: Leaving] |
| 18:36 | | Vash [Vash@Nightstar-e8057de2.wlfrct.sbcglobal.net] has joined #code |
| 18:36 | | mode/#code [+o Vash] by ChanServ |
| 19:39 | | Kindamoody|gaming is now known as Kindamoody[zZz] |
| 19:45 | | sshine [simon@Nightstar-883ecc1d.brahmaserver.dk] has quit [Ping timeout: 121 seconds] |
| 19:45 | | sshine [simon@Nightstar-883ecc1d.brahmaserver.dk] has joined #code |
| 20:19 | | cpux|2 [cpux@Nightstar-c5874a39.dyn.optonline.net] has joined #code |
| 20:21 | | cpux [cpux@Nightstar-c5874a39.dyn.optonline.net] has quit [Ping timeout: 121 seconds] |
| 20:42 | | Moltare [Moltare@583787.FF2A18.190FE2.4D81A1] has quit [Ping timeout: 121 seconds] |
| 20:45 | | Moltare [Moltare@583787.FF2A18.190FE2.4D81A1] has joined #code |
| 20:51 | | Derakon [Derakon@Nightstar-a3b183ae.ca.comcast.net] has joined #code |
| 20:51 | | mode/#code [+ao Derakon Derakon] by ChanServ |
| 20:52 | | Derakon[AFK] [Derakon@Nightstar-a3b183ae.ca.comcast.net] has quit [Operation timed out] |
| 21:33 | | cpux [cpux@Nightstar-c5874a39.dyn.optonline.net] has joined #code |
| 21:35 | | cpux|2 [cpux@Nightstar-c5874a39.dyn.optonline.net] has quit [Ping timeout: 121 seconds] |
| 22:25 | <&McMartin> | Also, woo, 30-year anniversary of the C64 |
| 22:26 | <@Tamber> | \o/ |
| 22:45 | | Vash is now known as Vash[CoX] |
| 22:45 | | Vash[CoX] is now known as Vash[CoH] |
| 23:06 | | * McMartin crackles with science |
| 23:07 | | * McMartin commits a varied patch with the explanation "make previous five commits stop being awful" |
| 23:07 | <~Vornicus> | Now I'm curious as to what the last five commits were about |
| 23:07 | <&McMartin> | First implementation of a feature |
| 23:08 | <&McMartin> | However, there were a bunch of silly typos in the parts that had to deal with the systemic parts so despite working fine internally they don't work in the system |
| 23:08 | <&McMartin> | Also I got two tests backwards in a way that made the use case not blocked by a different bug work but all other cases fail, and those just came to light. |
| 23:09 | <&McMartin> | (The others involved looking in the wrong place for external, real data) |
| 23:36 | | Kindamoody[zZz] is now known as Kindamoody |
| --- Log closed Sat Aug 04 00:00:21 2012 |