| --- Log opened Sat Nov 17 00:00:21 2012 |
| 00:17 | | Derakon[AFK] is now known as Derakon |
| 00:55 | <@RichyB> | http://ro-che.info/ccc/18.html |
| 00:55 | <@RichyB> | Heh. |
| 01:22 | | ToxicFrog [ToxicFrog@ServerAdministrator.Nightstar.Net] has quit [Operation timed out] |
| 01:29 | | ToxicFrog [ToxicFrog@ServerAdministrator.Nightstar.Net] has joined #code |
| 01:45 | | Derakon is now known as Derakon[AFK] |
| 01:51 | <@RichyB> | iospace: while I'm busy annoying the shit out of you with trivial linked-list sorting functions, I made faster, constant-space sorting things. Revision 1f50414f7f of https://github.com/RichardBarrell/snippets/blob/master/bubble.c added a nicer algorithm and 143de46243 made the bubblesort() 2x faster anyway. :) |
| 02:04 | | * McMartin wonders what kind of developer thinks that having no beta testers is something to brag about |
| 02:22 | <@RichyB> | Someone who tests very thoroughly themselves? |
| 02:22 | <@RichyB> | I've heard of companies who kind of make that work. |
| 02:23 | <@RichyB> | They have a development team that downs tools before the QA dept goes in. |
| 02:23 | <@RichyB> | If QA finds any nontrivial bugs at all, the whole release is rejected. |
| 02:23 | <@RichyB> | The approach is that the separate QA dept is actually assuring the dev team's ability to keep its bug count down to zero, rather than the code that's produced by itself. |
| 02:24 | <@RichyB> | So, uh, the boast there is that they're doing automated testing, manual testing themselves and correctness proofs to a sufficient standard that they only need acceptance-testing and skip external testing mid-process. |
| 02:24 | <@RichyB> | No idea if the approach is sane. |
| 02:28 | <~Vornicus> | If you're not building a user interface then yes. |
| 02:28 | | Attilla [Obsolete@Nightstar-732a9fe2.range109-156.btcentralplus.com] has quit [Ping timeout: 121 seconds] |
| 02:28 | <~Vornicus> | But the moment humans have to touch something you need people outside the dev team -- and preferably completely unfamiliar with your program -- to touch it. |
| 02:29 | < ToxicFrog> | HCI goes well beyond beta testing, IME |
| 02:30 | <@Azash> | Interaction should probably be part of the design from square one |
| 02:31 | <~Vornicus> | Yes, but even the best damn interaction designer can make groaners. |
| 02:34 | < ToxicFrog> | Right. My point is that you should be testing the UI well before beta, with actual people. |
| 02:34 | <~Vornicus> | true |
| 02:43 | <&McMartin> | No, this was an IFComp Entry that was, of course, entirely unusable shit |
| 02:46 | <~Vornicus> | of course. |
| 02:48 | | RichyB [richardb@Nightstar-3b2c2db2.bethere.co.uk] has quit [[NS] Quit: Leaving] |
| 03:12 | <@Alek> | "How many Prolog programmers do you need to change a lightbulb?" "Yes." |
| 03:22 | <&McMartin> | hooray |
| 03:33 | | Kindamoody[zZz] is now known as Kindamoody |
| 03:44 | | Vash [Vash@Nightstar-b43e074a.wlfrct.sbcglobal.net] has quit [[NS] Quit: I lovecraft Vorn!] |
| 03:44 | | Vash [Vash@Nightstar-b43e074a.wlfrct.sbcglobal.net] has joined #code |
| 03:44 | | mode/#code [+o Vash] by ChanServ |
| 03:50 | | syksleep is now known as Syk |
| 04:35 | <@Alek> | http://qdb.us/308510 |
| 05:01 | <~Vornicus> | jerith: upon examination I have determined that the tyrian ship graphics are too subtle. |
| 05:11 | < celticminstrel> | The underline makes that look like gdb rather than qdb. |
| 05:42 | <@Syk> | hmmm. |
| 05:45 | <@Syk> | i think I need to actually set up dropbox properly |
| 05:58 | | Kindamoody is now known as Kindamoody|breakfast |
| 06:01 | <@froztbyte> | ....how non-properly can it be? |
| 06:01 | <@froztbyte> | Alek: someone bought me a fridge magnet once, it read |
| 06:01 | <@froztbyte> | "Programming is an art form that fights back" |
| 06:02 | <@Syk> | froztbyte: i have to manually start it |
| 06:02 | <@froztbyte> | that's not necessarily a problem |
| 06:02 | <@froztbyte> | I have it set up that way too |
| 06:13 | | ErikMesoy|sleep is now known as ErikMesoy |
| 06:14 | | celticminstrel is now known as celmin|sleep |
| 06:34 | | Kindamoody|breakfast is now known as Kindamoody |
| 06:40 | <&jerith> | 04:23 <@RichyB> They have a development team that downs tools before the QA dept goes in. |
| 06:40 | <&jerith> | That's very Waterfall, and thus doomed. |
| 06:57 | | Kindamoody is now known as Kindamoody|out |
| 07:09 | | celmin|sleep [celticminst@Nightstar-05d23b97.cable.rogers.com] has quit [Ping timeout: 121 seconds] |
| 07:23 | | Derakon[AFK] is now known as Derakon |
| 07:26 | <@Syk> | jerith: dooomed! doooooooooomed!</futurama> |
| 07:35 | | Derakon is now known as Derakon[AFK] |
| 08:21 | | cpux [cpux@Nightstar-98762b0f.dyn.optonline.net] has quit [Connection reset by peer] |
| 08:24 | | cpux [cpux@Nightstar-98762b0f.dyn.optonline.net] has joined #code |
| 08:33 | | Vash [Vash@Nightstar-b43e074a.wlfrct.sbcglobal.net] has quit [[NS] Quit: I lovecraft Vorn!] |
| 08:39 | | Zemyla [zemyla@Nightstar-8fbb7981.compute-1.amazonaws.com] has joined #code |
| 08:43 | < Zemyla> | Hey everyone. |
| 08:46 | <@ErikMesoy> | Hello, person. |
| 08:46 | <@Syk> | hai |
| 09:24 | | You're now known as TheWatcher |
| 09:59 | | cpux [cpux@Nightstar-98762b0f.dyn.optonline.net] has quit [[NS] Quit: Well, most things get better when I kick them!] |
| 09:59 | | cpux [cpux@Nightstar-98762b0f.dyn.optonline.net] has joined #code |
| 10:15 | | cpux [cpux@Nightstar-98762b0f.dyn.optonline.net] has quit [[NS] Quit: Well, most things get better when I kick them!] |
| 10:23 | <@Syk> | so many lols were had |
| 10:23 | <@Syk> | fff wrong window |
| 10:24 | < auREAX> | it's okay |
| 10:24 | < auREAX> | you can have lols here as well |
| 10:24 | | * Syk engages in lols |
| 10:24 | <@Syk> | know what isn't lols? |
| 10:24 | <@Syk> | gnome-shell.css |
| 10:25 | < auREAX> | you know what else isn't lols? |
| 10:25 | < auREAX> | gnome |
| 10:25 | < auREAX> | :( |
| 10:25 | <@Syk> | i like gnome 3.6 |
| 10:25 | <@Syk> | ...with the correct adddons |
| 10:25 | < auREAX> | it's more the developers |
| 10:25 | <@Syk> | thats true |
| 10:35 | <@Syk> | :O #666666 is a base colour in the default gnome theme! |
| 10:35 | <@Syk> | GNOME: SHELL OF THE DEVIL |
| 10:35 | <@Syk> | there's also a Red Hat copyright notice at the top |
| 10:36 | <@Syk> | so not only is gnome the shell of the devil, he's wearing a damn fine fedora while he's using it |
| 10:39 | <@AnnoDomini> | Why so transparent? |
| 10:49 | <@Syk> | ? |
| 10:52 | <@AnnoDomini> | That colour has transparency. If it were fully opaque, it would be #666666FF. I think. |
| 10:55 | <@Syk> | this is CSS |
| 10:55 | <@Syk> | #RRGGBBOO |
| 10:56 | <@Syk> | if OO isn't specified it is automatically opaque |
| 11:04 | <@AnnoDomini> | I seeeeeee. |
| 11:56 | <@Syk> | the amount of work I have done today: 0! :D |
| 11:56 | <@Syk> | that's right, I have done /so much work/ i can only express it in a factorial |
| 12:24 | | celticminstrel [celticminst@Nightstar-05d23b97.cable.rogers.com] has joined #code |
| 12:27 | | VirusJTG [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has joined #code |
| 12:32 | < auREAX> | dayum |
| 12:37 | | fztm [FZTM@B83485.1FBDF7.86221C.2082D1] has joined #code |
| 12:38 | | fztm [FZTM@B83485.1FBDF7.86221C.2082D1] has quit [[NS] Quit: ] |
| 12:44 | | VirusJTG_ [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has joined #code |
| 12:45 | | VirusJTG [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has quit [Ping timeout: 121 seconds] |
| 12:48 | | Courage [Moltare@583787.FF2A18.190FE2.4D81A1] has joined #code |
| 12:48 | | Moltare [Moltare@583787.FF2A18.190FE2.4D81A1] has quit [Ping timeout: 121 seconds] |
| 12:49 | | VirusJTG_ [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has quit [[NS] Quit: Program Shutting down] |
| 12:50 | | VirusJTG [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has joined #code |
| 12:50 | < VirusJTG> | any one played around with windows phone 8? |
| 12:52 | <@Syk> | nah, i'd probably throw the device against a wall |
| 12:53 | < VirusJTG> | yeah, thats kind of why ia am asking |
| 12:53 | < VirusJTG> | it it as annoying as it looks? |
| 12:54 | < VirusJTG> | my poor smart phone is starting to have electrical issues |
| 12:54 | < VirusJTG> | time for a new one, just having trouble deciding what to get |
| 12:55 | < VirusJTG> | well, its had electrial and booting issue for the past 4 months trueth be told... |
| 12:59 | < VirusJTG> | and according to samsungupgrade.com my current phone is worth $15... |
| 12:59 | <@Syk> | VirusJTG: it's uh |
| 12:59 | <@Syk> | well, let me put it this way |
| 12:59 | <@Syk> | Windows Phone has sub-1% market share for a reason |
| 13:00 | < VirusJTG> | you make an excelent point Syk |
| 13:00 | < VirusJTG> | I'll stick with andriod |
| 13:00 | < VirusJTG> | now what duel core phone is the most rootable? |
| 13:00 | <@Syk> | Samsung Galaxy Nexus |
| 13:00 | <@Syk> | ~$335 i think |
| 13:01 | < VirusJTG> | that was a very quick responce, I'm impressed |
| 13:01 | <@Syk> | well uh |
| 13:01 | <@Syk> | considering the Samsung GNex comes with an unlocked bootloader~ |
| 13:01 | <@Syk> | it's an Android dev phone, essentially |
| 13:01 | | Attilla [Obsolete@Nightstar-b6de7b68.range86-171.btcentralplus.com] has joined #code |
| 13:01 | <@Syk> | you can flash any ROM you like |
| 13:02 | <@Syk> | VirusJTG: basically all you need to do is unlock it with fastboot and then install a custom recovery img |
| 13:02 | <@Syk> | which, if you use windows, there's automated tools for that |
| 13:02 | <@Syk> | it's easy on Linux as well, just need the Android SDK, which is free |
| 13:02 | < auREAX> | http://en.sfml-dev.org/forums/index.php?topic=9304.0 |
| 13:02 | < auREAX> | I love threads like these |
| 13:03 | < VirusJTG> | yeah I use windows because 1 I like to game 2 i have to support it at work so i need to know it well, and 3 its memory injection protection is laughable.... |
| 13:04 | < VirusJTG> | well shoot my carrier doesn't seem to have the galaxy nexus |
| 13:05 | <@Syk> | VirusJTG: who's your carrier |
| 13:05 | < VirusJTG> | t-mobile |
| 13:05 | <@Syk> | they're GSM aren't they? |
| 13:05 | < VirusJTG> | they are cheep... |
| 13:05 | <@Syk> | VirusJTG: the Galaxy Nexus was sold by Google unlocked |
| 13:05 | <@Syk> | lemme see |
| 13:05 | < VirusJTG> | to the googles! |
| 13:06 | <@Syk> | hmm dunno where you can get it in the USA |
| 13:06 | < VirusJTG> | the nexus 4 looks intersting |
| 13:06 | <@Syk> | VirusJTG: the Galaxy Nexus was mainly bought unlocked |
| 13:06 | <@Syk> | the Nexus 4 is good too |
| 13:06 | <@Syk> | it's a quadcore |
| 13:06 | < VirusJTG> | and sold unlocked |
| 13:06 | <@Syk> | also it's fssubdfsing sold out everywhere |
| 13:07 | < VirusJTG> | more CPU power is always welcomed |
| 13:07 | <@Syk> | VirusJTG: every Nexus will be unlocked |
| 13:07 | <@Syk> | the only problem with the nexus 4 is LG |
| 13:07 | <@Syk> | and the glass back, I suppose |
| 13:07 | <@Syk> | I'm not too sold on the LG bit |
| 13:07 | < VirusJTG> | it will bounce.... once..... |
| 13:07 | <@Syk> | but then again, I'd not touch a Samsung phone with a 10ft pole |
| 13:07 | <@Syk> | yet I own a Galaxy Nexus |
| 13:08 | <@Syk> | if it's a Nexus, it's pretty much guaranteed to be good. |
| 13:08 | < VirusJTG> | if i pay that much for hardware I want to own it and have full control over it, ya know? |
| 13:08 | <@Syk> | VirusJTG: if you can get your hands on one, and if you don't have anything against LG, the Nexus 4 will probably be better |
| 13:08 | < VirusJTG> | thats why I want a rootabke phone, it gives me that control |
| 13:08 | <@Syk> | well most phones are 'rootable' |
| 13:08 | <@Syk> | officially or not :P |
| 13:08 | | * VirusJTG thinks Syk speeks with wisdom |
| 13:09 | < VirusJTG> | this is true..... |
| 13:09 | <@Syk> | anything new and Samsung will be not supported well |
| 13:09 | <@Syk> | as theyre usin some new Exynos thing which nobody supports |
| 13:09 | < VirusJTG> | htc seems to be reasonable rootable |
| 13:09 | <@Syk> | except HTC are... ehhh. |
| 13:09 | <@Syk> | I dunno |
| 13:09 | < VirusJTG> | and they take abuse, my cirrent one if from them and has been dropped many times |
| 13:10 | <@Syk> | I've owned two HTC phones, and it's a bit of a pain. |
| 13:10 | < VirusJTG> | yeah, the root process takes 2 hours |
| 13:10 | <@Syk> | if you want the most well supported ROMs and such, the Nexuses are the way to go |
| 13:10 | < VirusJTG> | I guess I just have to wait until they are in stock then |
| 13:11 | <@Syk> | yeah |
| 13:11 | <@Syk> | either the Galaxy Nexus or the Nexus 4 |
| 13:11 | <@Syk> | the difference is not huge |
| 13:11 | < VirusJTG> | I think i can coax my poor for for a bit longer |
| 13:11 | <@Syk> | the GNex has support now, and you can probably get your hands on it easier |
| 13:11 | < VirusJTG> | looking at it, the nexus 4 is superior hardware to even the Gs3 |
| 13:11 | <@Syk> | while the Nexus 4's GSM stack hasn't been pushed to AOSP, so ROM support might be a little while |
| 13:12 | <@Syk> | VirusJTG: considering the GS3 is a six month old phone, is more or less to be expected |
| 13:12 | < VirusJTG> | true |
| 13:12 | <@Syk> | I think it runs very close to the GS3 |
| 13:12 | < VirusJTG> | but 6 month is better than 3 year... |
| 13:12 | <@Syk> | the GS3 does a little better i think though |
| 13:13 | | Attilla [Obsolete@Nightstar-b6de7b68.range86-171.btcentralplus.com] has quit [[NS] Quit: ] |
| 13:13 | <@Syk> | because it's been fairly optimised by now |
| 13:13 | <@Syk> | VirusJTG: the GNex is about a year old, is plenty powerful, has current ROM support and has 4.2 right now |
| 13:13 | <@Syk> | in realistic terms, it's just as good as the Nexus 4 |
| 13:13 | | * VirusJTG pokes at googles website |
| 13:14 | <@Syk> | you may have to get it from a third party |
| 13:14 | < VirusJTG> | wow chrome books are cheep |
| 13:15 | <@Syk> | yeah but theyre chromebooks |
| 13:15 | <@Syk> | VirusJTG: google 'Galaxy nexus' and select Shopping |
| 13:15 | <@Syk> | it'll have a few there |
| 13:16 | < VirusJTG> | not displaying for me... |
| 13:16 | | Attilla [Obsolete@Nightstar-b6de7b68.range86-171.btcentralplus.com] has joined #code |
| 13:16 | < VirusJTG> | all i see are 4s, 7s, and 10, |
| 13:16 | <@Syk> | i mean in google search itself |
| 13:18 | < VirusJTG> | yup found it that way, thanks Syk |
| 13:21 | <@Syk> | np |
| 13:23 | <@Syk> | yay 4.2 update for my nexus 7 |
| 13:23 | <@Syk> | gonna backup my phone and put the stock 4.2 on my gnex too... |
| 13:23 | < VirusJTG> | another half hour and i am off to home depot for some supplies |
| 13:24 | < VirusJTG> | time to winterise the house |
| 13:24 | <@Syk> | hurry up tablet |
| 13:24 | <@Syk> | i want to play some need for speed |
| 13:25 | | * VirusJTG gets out adn pushes :-P |
| 15:18 | <@Azash> | How I imagine the node.js vision: https://images.4chan.org/wsg/src/1352686167547.gif |
| 15:18 | <&jerith> | So far C# has managed to be substantially nicer than Java. |
| 15:20 | <@Syk> | Azash: ahaa |
| 15:20 | <@Syk> | oh man, gimme a minute |
| 15:20 | <@Syk> | i think i have a few of those gifs |
| 15:25 | <@Syk> | Azash: http://reddrgn.net/infomercial3.gif |
| 15:25 | <@Syk> | Azash: this is more like node.js |
| 15:25 | <@Syk> | "oh god how do i hold all of these requests" |
| 15:26 | <@Syk> | (there's also -1.gif and -2.gif :P) |
| 15:42 | | Attilla [Obsolete@Nightstar-b6de7b68.range86-171.btcentralplus.com] has quit [Ping timeout: 121 seconds] |
| 15:43 | | Attilla [Obsolete@Nightstar-b6de7b68.range86-171.btcentralplus.com] has joined #code |
| 16:03 | | Syk is now known as syksleep |
| 18:31 | | Kindamoody|out is now known as Kindamoody |
| 18:32 | | VirusJTG [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has quit [Client closed the connection] |
| 18:33 | | VirusJTG [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has joined #code |
| 18:35 | | cpux [cpux@Nightstar-98762b0f.dyn.optonline.net] has joined #code |
| 18:35 | | PAndemic [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has quit [Client closed the connection] |
| 18:42 | | Pandemic [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has joined #code |
| 18:59 | < auREAX> | Azash: "i maek wwebsite no blocking" -- Ryan Dahl, solver of the halting problem |
| 19:17 | <&jerith> | node.js has a bunch of cool ideas in it, but as far as I can tell the implementation is lacking. |
| 19:17 | <&jerith> | And there are some silly ideas in there as well. |
| 19:18 | <&jerith> | For starters, JS. |
| 19:20 | <~Vornicus> | what's node.js do? |
| 19:20 | <@Tamber> | Everything except that one little tiny thing you need to finish your project. |
| 19:22 | | Kindamoody is now known as Kindamoody[zZz] |
| 19:25 | <&jerith> | Vornicus: It's an event-driven JS app server. |
| 19:41 | <~Vornicus> | ...well then. |
| 19:46 | <@AnnoDomini> | So. It has come to this. |
| 19:46 | <&jerith> | It has. |
| 19:48 | <@ErikMesoy> | That's what you said about Steam too. :p "So it has come to this. A package manager for Windows." |
| 19:55 | < ToxicFrog> | It desperately needs one~ |
| 19:55 | <@AnnoDomini> | s/needs/needed/ :P |
| 19:56 | <&jerith> | s/needed/needs/~ |
| 19:58 | <@AnnoDomini> | As Erik mentioned, there's Steam. And the Windows Installer thingy in the control panel. |
| 19:59 | <@AnnoDomini> | Does it let you download the stuff you want, rather that requiring a CD, these days? |
| 19:59 | <&jerith> | Appstore != package manager. |
| 20:00 | <@AnnoDomini> | What's the distinction? |
| 20:01 | < ToxicFrog> | The windows installer thingy in the control panel does not automatically download anything, AFAIK. |
| 20:01 | < ToxicFrog> | Steam is very much games-focused and is not a substitute for a general purpose package manager. |
| 20:02 | < auREAX> | hence steam for linux |
| 20:02 | < ToxicFrog> | (and comes with a lot of excess baggage like community integration that makes sense for a gaming client but less so for a package manager) |
| 20:02 | <@Azash> | "very much" as in "package manager for a subset of products in a subset of product type" :p |
| 20:03 | <@Azash> | I wonder if it would be doable to make a package manager for Windows |
| 20:03 | <&jerith> | Package managers manage packages. |
| 20:03 | <@Azash> | Okay |
| 20:03 | <@Azash> | Er, application edge-bleediness manager? |
| 20:03 | <&jerith> | Steam and friends only operate on discrete, standalone apps. |
| 20:05 | <&jerith> | Also, *proper* package managers give you reasonably easy access to the source. |
| 20:05 | <@Azash> | auREAX: kveri |
| 20:05 | < auREAX> | who now |
| 20:05 | < auREAX> | oh |
| 20:05 | < auREAX> | I was waiting for you to finish |
| 20:41 | | Derakon[AFK] is now known as Derakon |
| 20:51 | < Zemyla> | You know, I came here because you guys were discussing Spacechem earlier in your logs. |
| 20:58 | <&jerith> | That happens from time to time. |
| 21:11 | < ToxicFrog> | Yeah, it's as much code as gaming. |
| 21:37 | <@AnnoDomini> | How do I copy directories with cp? |
| 21:37 | <&jerith> | AnnoDomini: cp -r |
| 21:37 | <@AnnoDomini> | Thank you. |
| 21:38 | <@AnnoDomini> | Will that recursively copy the contents? |
| 21:38 | <@AnnoDomini> | Even if they are directories? |
| 21:38 | <&jerith> | The "-r" means "recursive". |
| 21:39 | | * TheWatcher generally suggests -a over -r as it preserves links |
| 21:41 | <&jerith> | Good point. |
| 21:42 | <@AnnoDomini> | Okay. So if I wanted to copy irclogs from home to public_html, I'd go "cp -r ~/irclogs ~/public_html"? |
| 21:42 | <&jerith> | Yup. |
| 21:43 | <@AnnoDomini> | This... does nothing. For some reason. |
| 21:46 | <@AnnoDomini> | Or rather, it does do something, but the changes don't show up in apache's autogenerated directory browsing. |
| 21:46 | <@TheWatcher> | Permissions are probably wrong |
| 21:46 | <@TheWatcher> | can apache read the directory? |
| 21:47 | <@TheWatcher> | (ie: you might need to do a `chmod -R o+rX ~/public_html/irclogs`) |
| 21:47 | <@AnnoDomini> | chown did it! |
| 21:47 | <@AnnoDomini> | Thank you. |
| 21:57 | <&McMartin> | lol wtf chownt |
| 22:06 | | * TheWatcher stabs module authors who don't document things properly, arghs |
| 22:07 | <@TheWatcher> | I should not have to look at the source to figure this shit out, damnit |
| 22:09 | <@ErikMesoy> | What does it indicate when a Python class calls self(x) without having a self() function? I'm guessing it's shorthand for a call to __init__() . |
| 22:09 | <&Derakon> | Uh, I think it means that the class has implemented the magic function that lets it be invoked as a function itself. |
| 22:09 | <&Derakon> | I think it's __call__. |
| 22:10 | <@ErikMesoy> | mhm |
| 22:12 | <@rms> | I know in Java it means it's calling the constructor. |
| 22:12 | <@rms> | (If that helps any) |
| 22:12 | <@rms> | s/the/a/ |
| 22:12 | <&McMartin> | self is an instance. If you're calling that you're calling __call__ |
| 22:12 | <&McMartin> | You name the *class*, not the instance, to get a constructor call, in Java as in Python |
| 22:13 | <&Derakon> | Yep. |
| 22:13 | <&McMartin> | Well |
| 22:13 | <&McMartin> | In C++ as in Pythin. |
| 22:13 | <&Derakon> | class Foo: def __call__(self): print "Hello, world!" |
| 22:13 | <&Derakon> | Foo()() |
| 22:13 | <&McMartin> | Java *requires* the new operator. |
| 22:13 | <&Derakon> | Hello, world! |
| 22:14 | | * McMartin eyes this Markov chain. |
| 22:14 | <@rms> | Nope, in Java it's this() to call a constructor of the current class, super() to call the parent's constructor (has to be the first call of the constructor though) |
| 22:14 | <&McMartin> | "Each episode is a reboot stamping on a human face in a corner" |
| 22:14 | <@AnnoDomini> | Markov the Vampire? |
| 22:14 | <@rms> | (When you want a constructor to call another constructor) |
| 22:14 | | mode/#code [+ao ToxicFrog ToxicFrog] by ChanServ |
| 22:14 | <&McMartin> | rms: "self" isn't a keyword in python. |
| 22:14 | <@rms> | Ah right |
| 22:15 | <&McMartin> | The thing you describe is a parser hack to get around Java's lack of initializer lists, more or less. |
| 22:15 | <&McMartin> | (Since, IIRC, if you don't put that call to super() in, then if there isn't a default constructor it won't compile.) |
| 22:16 | <&ToxicFrog> | AnnoDomini: incidentally, if you are copying large amounts of stuff, consider using rsync -aPh rather than cp -a |
| 22:16 | <&ToxicFrog> | The former gives you progress indicators and is resumably. |
| 22:16 | <&ToxicFrog> | *resumable. |
| 22:17 | | * McMartin eyes his OpenGL implementation, which claims to support 8192x8192 textures. |
| 22:18 | <&McMartin> | Well, OK, I guess that's reasonable |
| 22:18 | <&Derakon> | "Support" and "can display at all quickly" are not necessarily the same thing. |
| 22:18 | <&McMartin> | Yeah |
| 22:18 | <&Derakon> | Remember when I was mucking around with camera displays and determined it was faster to do 4x 512x512 textures instead of 1x 1024x1024? |
| 22:18 | <@AnnoDomini> | ToxicFrog: Okay. Not in this case, though. It's just a couple MB of text. |
| 22:18 | <&McMartin> | But the last time I did OpenGL programming I didn't get to assume anything better than 256x256/ |
| 22:18 | <&McMartin> | *256x256 |
| 22:18 | <&Derakon> | Heh. |
| 22:19 | <&McMartin> | As in, "would not actually work, at all" |
| 22:22 | <&ToxicFrog> | AnnoDomini: yeah. Just saying for future reference, next time you find yourself copying multiple gigabytes or whatever. |
| 22:25 | | * McMartin starts moving logic out of his CPU and into his GPU. |
| 22:29 | <&McMartin> | Right |
| 22:29 | | * McMartin gives two fingers to Progress, refers to fixed functionality uniforms and attributes *everywhere* |
| 22:30 | | * McMartin is hidebound, and still uses such atrocities of the Old World like a "Projection Matrix" and a "Modelview Matrix" |
| 22:30 | <@TheWatcher> | *gasp* |
| 22:30 | <@TheWatcher> | Teh Horrorz! |
| 22:32 | | * McMartin continues to remain skeptical of the fanatic campaign to erase these things from history |
| 22:32 | | * TheWatcher likewise |
| 22:32 | <&Derakon> | AIUI it's basically a campaign towards performance at all costs. |
| 22:32 | <&McMartin> | I mean, yes, once you have UBOs you can manage this data in better and more generic ways. |
| 22:32 | <&McMartin> | But UBOs are only guaranteed to exist on DX11 cards. -_- |
| 22:32 | <&McMartin> | (Most DX10 cards have them anyway) |
| 22:33 | <&McMartin> | Derakon: This one, I think, is actually "stop having globals". |
| 22:33 | <&McMartin> | Immediate Mode has been deprecated since like 1.5, and its replacements are so basic that you have to do tricks with GLEW to even bother checking for them. |
| 22:34 | <&McMartin> | The justification for wiping all these out is "well, now that we actually have a way to modularly specify blocks of constants, we don't need this shitload of global variables anymore" |
| 22:35 | <&McMartin> | Which, you know, OK, but could you maybe also publish some basic vertex and fragment shaders that replicate fixed functionality, with comments and without "bypass shader" instructions |
| 22:35 | <&McMartin> | Also, since I'm targeting DX9 cards here, I don't get that way of modularly specifying blocks of constants. |
| 22:36 | <@TheWatcher> | Egads, man, targetting old cards? What kind of OpenGL programmer are you?!~ |
| 22:37 | <&McMartin> | s/old/current/ |
| 22:37 | <&jerith> | TheWatcher: A grumpy one, apparently.~ |
| 22:37 | | * McMartin also stabs Canonical. INTEL HAS GIVEN YOU THE UBO CODE ALREADY, PLEASE TO USE IT |
| 22:38 | <@TheWatcher> | (Can't say I blame him, the state of opengl right irritates the hell out of me) |
| 22:38 | <@TheWatcher> | *right now |
| 22:38 | <&McMartin> | A lot of the Good Stuff seems like it should be backportable to 2.x |
| 22:38 | <&McMartin> | By which I mean VAOs and UBOs, yes |
| 22:38 | | * jerith has come to see a certain amount of grumpiness as a sign of maturity in programmers. |
| 22:39 | <&McMartin> | I actually wouldn't *mind* targeting 3.2 core here |
| 22:39 | <@Tamber> | jerith, it's a sign of when they reach the point of realisation that *everything is broken, and nobody can/will do anything about it*?:p |
| 22:39 | <&McMartin> | I just don't get to because it will only run on one of my four machines. |
| 22:39 | < Courage> | This takes approximately three days for runescript programmers~ |
| 22:40 | <&McMartin> | If I were an engine writer, I'd be really wanting the things they added. |
| 22:40 | <&McMartin> | Hell |
| 22:40 | <&McMartin> | I've looked at some of my own code in 1.x |
| 22:40 | <&McMartin> | And noticed all this stuff I wouldn't need if I got to use 3.2 |
| 22:40 | <&McMartin> | So clearly they got *some* things right |
| 22:40 | | * TheWatcher goes to stab the authors of the mediawiki API documentation |
| 22:40 | <&McMartin> | In terms of "cutting bullshit", as well as "adding features I don't need" |
| 22:40 | <&jerith> | Tamber: No, it's a sign that we've reached the point that we can recognise just how much horribleness is unnecessary, and is perpetrated by people who really should know better. |
| 22:41 | <@Tamber> | Aha |
| 22:41 | <&McMartin> | Yeah |
| 22:41 | <@TheWatcher> | Documenting the success response: great idea, thanks! Now, about what happens if it fails... |
| 22:41 | <&McMartin> | I'm just hitting the point that the specific thing that let them deprecate All The Things I Want is in fact the "knows better" solution. |
| 22:41 | | * TheWatcher should not have to break his fucking wiki just to find out what the API's response is when that happens |
| 22:42 | <&McMartin> | SCIENCE~~ |
| 22:42 | <&jerith> | TheWatcher: I love how the search API helpfully returns a "search disabled" error to let you know that you've tried to search for punctuation. |
| 22:45 | | ToxicFrog [ToxicFrog@ServerAdministrator.Nightstar.Net] has quit [Ping timeout: 121 seconds] |
| 22:46 | <&McMartin> | Oh wait |
| 22:46 | | * McMartin removes a matrix multiply from his vertex shader. |
| 22:46 | <&McMartin> | I don't *need* a modelview matrix, the camera is static and the object is dynamically generated using world coordinates directly |
| 22:46 | <&McMartin> | 2.x: undeniably better than 1.x |
| 22:50 | | ToxicFrog [ToxicFrog@ServerAdministrator.Nightstar.Net] has joined #code |
| 22:53 | | ErikMesoy is now known as ErikMesoy|sleep |
| 22:58 | <&McMartin> | Success! |
| 22:59 | | * McMartin has extracted the modelview matrix. |
| 22:59 | | * McMartin now goes to take the heightmap color selection out of the CPU. |
| 23:01 | | * jerith goes to bed. |
| 23:25 | <&McMartin> | Success, I think. |
| 23:26 | <@TheWatcher> | Provisional \o/ |
| 23:30 | | * McMartin snerks |
| 23:30 | <&McMartin> | Bonus \o/ |
| 23:31 | <&McMartin> | I was doing the Goraud lighting equation in the CPU |
| 23:31 | <&McMartin> | It turns out I was also doing it *wrong* |
| 23:31 | <&McMartin> | So I fixed that on my way to the GPU, and now it looks way better~ |
| 23:31 | <@TheWatcher> | Awesomeness. |
| 23:31 | <&Derakon> | ...why are you doing your own lightning calculations? |
| 23:31 | <&Derakon> | Shouldn't the GPU do that for you? |
| 23:32 | <&McMartin> | Dynamic terrain, so OpenGL 1.x required me to make my own normals. |
| 23:32 | <&McMartin> | Which was actually harder than just Doing The Lighting Computation By Hand. |
| 23:32 | <&McMartin> | Once I get this height map out of the CPU memory and into a texture array, I can also have the vertex shader do the normal computation, at which point I'm going to move the lighting equation into the *fragment* shader. |
| 23:33 | <&Derakon> | Is this all still for Dapper Delver? |
| 23:33 | <&McMartin> | No, this is actually me playing with chunks of Sable. |
| 23:33 | <&Derakon> | If so, I wasn't aware that procedural terrain was a part of that game. |
| 23:33 | <&Derakon> | Oh, okay. |
| 23:33 | <&Derakon> | "Let's use familiar code to get to grips with OpenGL". |
| 23:34 | <&McMartin> | "Let's evolve some OpenGL 1.1 code to OpenGL 2.1", actually :D |
| 23:34 | <&McMartin> | My grips on OpenGL 1.1 are actually quite strong, and I've since updated that to 1.5 |
| 23:34 | <&McMartin> | But I can't use the 1.5 goodies for this until I get the shaders in order |
| 23:34 | <&Derakon> | I still haven't progressed beyond 1.x. |
| 23:34 | <&Derakon> | Too fond of immediate mode. |
| 23:34 | <&McMartin> | Since the shaders are what make the CPU's idea of the geometry perfectly static, which lets me *dramatically* optimize the vertex submission. |
| 23:35 | <&McMartin> | Basically, at that point, the CPU says "here's a texture, and here's a grid of flat vertices in a square grid. Why don't you draw that?" and then the vertex shader will take that texture as a heigh tmap and turn it into terrain and correct normals |
| 23:35 | <&McMartin> | And then the fragment shader will handle lighting |
| 23:36 | <&McMartin> | And I guess there's *technically*a 1-D procedural texture that corresponds to the height map. Maybe I can make that cooler and put in a snow cap along with the tree line. |
| 23:36 | <&Derakon> | Heh. |
| 23:37 | <&Derakon> | I think I'd like dicking around with OpenGL more if it weren't so hard for me to figure out why it doesn't work when I break it. |
| 23:37 | <&McMartin> | Doing that right will *have* to happen at the fragment level, though, so. |
| 23:37 | <&Derakon> | Too often the result of bad code is just a black window with no insight into what went wrong. |
| 23:38 | <&McMartin> | Yeah, and there are entirely too many ways it can break horribly. |
| 23:40 | <&McMartin> | Anyway, once I get this stuff under control, I wouldn't mind also going back to Sable and updating its renderers |
| 23:40 | <&McMartin> | One of the ugliest bits about it is that ironically, low-poly models aren't really usable *without* pixel shader support. |
| 23:40 | <&McMartin> | Specular highlights are awful-looking in low-poly models with vertex specular computation. |
| 23:46 | | You're now known as TheWatcher[T-2] |
| 23:55 | | You're now known as TheWatcher[zZzZ] |
| --- Log closed Sun Nov 18 00:00:36 2012 |