| --- Log opened Fri Feb 17 00:00:28 2017 |
| 00:49 | | Derakon[AFK] is now known as Derakon |
| 01:41 | | * Vornicus discovers that he missed several available symmetries because his program was a little too specific about certain things. |
| 01:50 | <&ToxicFrog> | My portal placement code isn't working |
| 01:50 | <&ToxicFrog> | -> add a bunch of debugging code |
| 01:50 | <&ToxicFrog> | -> restart game |
| 01:50 | <&ToxicFrog> | -> portal placement works perfectly |
| 01:50 | <~Vornicus> | now it ... yep |
| 01:50 | <&ToxicFrog> | -> realize I forgot to save before the first test |
| 02:11 | | * Vornicus now gets 25 possibly-different things, checks them, gets indeed every single one he should have gotten in the first place, with 6 conjugate pairs like he expected. |
| 02:11 | <~Vornicus> | so, real number of symmetries on my square: 19. |
| 02:37 | <&ToxicFrog> | Ahahaha what the fuck |
| 02:37 | <&ToxicFrog> | MessageBox presents a message and up to ten buttons and waits for the user to select one |
| 02:37 | <&ToxicFrog> | The usage is: |
| 02:37 | <&ToxicFrog> | MessageBox <args> |
| 02:38 | <&[R]> | lolwut |
| 02:38 | <&ToxicFrog> | set selection to GetButtonPressed ; returns the index of the button they pressed |
| 02:38 | <&ToxicFrog> | ; do stuff based on selection |
| 02:38 | <&ToxicFrog> | *except* |
| 02:38 | <&ToxicFrog> | It takes up to 15 frames from the player clicking the button to GetButtonPressed being aware of it, because ???!?! |
| 02:38 | | McMartin [mcmartin@Nightstar-rpcdbf.sntcca.sbcglobal.net] has joined #code |
| 02:39 | | mode/#code [+ao McMartin McMartin] by ChanServ |
| 02:39 | <&ToxicFrog> | So you actually have to call MessageBox, and then loop (which is eqv to setting some state that prevents you from MessageBoxing again and then returning; the script will be called again next frame) until GetButtonPressed > -1 |
| 02:55 | <&ToxicFrog> | There also doesn't appear to be any way to split a statement across multiple lines |
| 02:55 | <&ToxicFrog> | Which is just fucking delightful when you have functions like MessageBoxEX that can take 21 arguments including the complete text of the message |
| 02:56 | <&McMartin> | ... I missed the beginning of this, what is all this then |
| 02:57 | <&ToxicFrog> | Oblivion scripting |
| 02:57 | <&ToxicFrog> | Which is a gigantic trash fire even with OBSE |
| 02:57 | | * Vornicus gives McM 5 additional symmetries which his previous script couldn't find. |
| 02:57 | <&ToxicFrog> | ...something has gone wrong here |
| 02:58 | <&ToxicFrog> | "Open which portal? Diamonds: 0 Emeralds: 0 Rubies: 1074790400" |
| 02:58 | <~Vornicus> | Whups |
| 02:58 | <&McMartin> | Okay |
| 02:58 | <&McMartin> | I was about to say "MessageBoxExW is a much simpler function than that" |
| 02:59 | <~Vornicus> | That's... somewhere in the neighborhood of 2^30 |
| 02:59 | <&ToxicFrog> | The best part is that I have no rubies, but four diamonds |
| 02:59 | <~Vornicus> | Indeed, that's 0x40100000 |
| 03:02 | | LadyOfLight` [catalyst@Nightstar-hb5vcd.dab.02.net] has joined #code |
| 03:04 | <&ToxicFrog> | Sadly, the contract of GetItemCount is that it returns the number of items of that type being held, 0 if none are |
| 03:04 | | LadyOfLight [catalyst@Nightstar-07ra9n.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 03:12 | <&ToxicFrog> | holy shit |
| 03:12 | <&ToxicFrog> | This line: |
| 03:12 | | * Azash encounters the AutoHotkey scripting interface |
| 03:12 | <&ToxicFrog> | printToConsole "diamonds: %d; rubies: %d" |
| 03:12 | <&ToxicFrog> | Produces a "mismatched quotes" parse error |
| 03:12 | <&ToxicFrog> | CAN YOU GUESS WHY |
| 03:13 | <@Azash> | Sleep function: Pass parameter X and sleep that many milliseconds |
| 03:13 | <@Azash> | Except if X is -1 |
| 03:13 | <@Azash> | Then you instead force an event queue resolution |
| 03:14 | <~Vornicus> | TF: because it seeks ';' before figuring out string literal tokens |
| 03:14 | <&ToxicFrog> | Vornicus: bingo |
| 03:14 | <&McMartin> | Vorn beats me to it |
| 03:14 | <&ToxicFrog> | Also, found the problem with getItemCount! |
| 03:14 | <&ToxicFrog> | It returns a float. |
| 03:15 | <&ToxicFrog> | (uint16_t)0.0 is apparently 1074790400 sometimes. |
| 03:15 | <&ToxicFrog> | And (uint16_t)4.0 is 0. |
| 03:15 | <~Vornicus> | What. |
| 03:15 | <~Vornicus> | Okaythen. |
| 03:16 | <&McMartin> | That only raises further questions |
| 03:16 | <&McMartin> | Of which, I think, the most relevant would be "how do you fit 1074790400 into a uint16_t" |
| 03:16 | <~Vornicus> | ^ |
| 03:16 | <&ToxicFrog> | Oh, good point |
| 03:16 | <&ToxicFrog> | I guess 'short' is uint32_t here |
| 03:17 | <&McMartin> | (There are a great many bit representations of 0.0 in IEEE 754, and doing a reinterpret_cast to int32 or int64 could give pretty exciting results) |
| 03:17 | <~Vornicus> | wait, a great many? |
| 03:17 | | * Vornicus checks. |
| 03:17 | <&ToxicFrog> | McMartin: this does however raise the question of why getItemCount returns a float in the first place |
| 03:18 | <&McMartin> | Secretly implemented in Perl and/or JavaScript? |
| 03:18 | <~Vornicus> | No, there's only two 0s. There's a lot of NaNs. |
| 03:19 | <&McMartin> | Whoops, right. |
| 03:19 | <&McMartin> | =~= WATMAN =~= |
| 03:22 | <&ToxicFrog> | McMartin: god knows what it's implemented in. It does however have distinct float, short, long, and ref types |
| 03:26 | <&McMartin> | 0x40100000 is 2.25, for the record. Does that number hold any relevance to you? |
| 03:26 | <&ToxicFrog> | none whatsoever |
| 03:33 | <&ToxicFrog> | ADDED FUN: you can only edit the script inside TESCS. The built in script editor can only open one file at a time, uses a variable-width font, and has no syntax hilighting. |
| 03:33 | <&ToxicFrog> | And you have to close the script editing window before you can save the mod. |
| 03:38 | <&ToxicFrog> | ...huh |
| 03:38 | <&ToxicFrog> | Calculated offsets for portal: (0, 75, 100) |
| 03:38 | <&ToxicFrog> | So it calls portal.moveTo player, 0, 75, 100 |
| 03:38 | <&ToxicFrog> | The player is at (-500, -240, -250) |
| 03:39 | <&ToxicFrog> | So of course the portal ends up at (1800, 1915, 7881) |
| 03:39 | <@Azash> | Seems reasonable |
| 03:40 | <&ToxicFrog> | The best part is that the portal placement code was working earlier this evening ;.; |
| 03:42 | <&ToxicFrog> | Ok |
| 03:43 | <&ToxicFrog> | apparently there is some semantic difference between "let x := pt.fPortalDistance * cos theta" and "set x to pt.fPortalDistance * cos theta" |
| 03:43 | <&ToxicFrog> | That results in exactly the same output when `printC`ing x, but different behaviour when passing x to moveTo |
| 03:44 | <@Azash> | Reference vs. literal? |
| 03:53 | <&ToxicFrog> | "huh", sez I. "why am I crouching, walking forwards, and blocking as soon as I load the game?" |
| 03:53 | | * ToxicFrog looks at the cat |
| 03:53 | | * ToxicFrog unplugs the wireless gamepad receiver |
| 03:54 | <&ToxicFrog> | ...and suddenly everything is normal again. |
| 03:54 | | celticminstrel [celticminst@Nightstar-h4m24u.dsl.bell.ca] has quit [[NS] Quit: KABOOM! It seems that I have exploded. Please wait while I reinstall the universe.] |
| 03:54 | | celticminstrel [celticminst@Nightstar-h4m24u.dsl.bell.ca] has joined #code |
| 03:54 | | mode/#code [+o celticminstrel] by ChanServ |
| 04:25 | | AverageJoe [evil1@Nightstar-lc3.hbj.189.210.IP] has joined #code |
| 04:35 | <&McMartin> | \o/ |
| 05:02 | | Derakon is now known as Derakon[AFK] |
| 05:29 | | celticminstrel is now known as celmin|sleep |
| 05:56 | | * McMartin writes a set of self-modifying routines that turn the entire upper half of BASIC's RAM into a giant ring queue. |
| 05:58 | <~Vornicus> | that's a small amount of queueueue. |
| 05:58 | <&McMartin> | Should be enough to let me do the lazy implementation of floodfill though. |
| 05:59 | <&McMartin> | I'm not completely sure it's enough for the fully general case, actually >_> |
| 05:59 | <~Vornicus> | http://imgur.com/QAulMny Hooray, <r^2, rfs> symmetry |
| 06:06 | <&McMartin> | Crypt of the Necrodancer OST makes *terrible* coding music |
| 06:07 | <~Vornicus> | 'cause you type slower to match the beat? |
| 06:08 | <&McMartin> | Very little typing ends up happening |
| 06:08 | <~Vornicus> | too busy grooving then |
| 06:13 | | * LadyOfLight` twirls |
| 06:13 | < LadyOfLight`> | I have an issue |
| 06:13 | < LadyOfLight`> | I can explain Reduct in half a page of text, but I can't actually write an evaluator for the language |
| 06:13 | | * Vornicus gives LadyOfLight` the bug reporting link. |
| 06:13 | <~Vornicus> | ah |
| 06:13 | | * LadyOfLight` hms |
| 06:14 | < LadyOfLight`> | Largely because I dislike taking the LISP cop-out of 'well, everything at the top level is /really/ just a statement' |
| 06:14 | | LadyOfLight` is now known as LadyOfLight |
| 06:15 | < LadyOfLight> | Maybe a program definition is a map and the lookup is of argument in the map |
| 06:15 | < LadyOfLight> | Nevermind, problem solved. |
| 06:15 | < LadyOfLight> | That makes a lot of sense actually |
| 06:16 | < LadyOfLight> | Although the word 'map' feels inadequate to describe a data structure which is really an infinite map of all possible symbols as keys |
| 07:08 | | Xon [Xon@Nightstar-j72.ku7.252.119.IP] has quit [A TLS packet with unexpected length was received.] |
| 07:09 | | Xon [Xon@Nightstar-j72.ku7.252.119.IP] has joined #code |
| 07:09 | | mode/#code [+o Xon] by ChanServ |
| 07:15 | | * McMartin gets his floodfill routine working, after accidentally messing up the queue code such that it actually smashed the entire program including itself |
| 07:22 | <&McMartin> | I guess blitting over the entire program is *technically* an instance of self-modifying code |
| 07:28 | <~Vornicus> | :/ |
| 07:43 | <&[R]> | Ugghhhhh |
| 07:44 | <&[R]> | s.split('asdf', 3) == s.split('asdf').slice(0, 3) |
| 07:45 | <~Vornicus> | wat |
| 07:45 | <&[R]> | I was kind of expecting it to do what PHP's explode did when given a number D: |
| 07:47 | <@macdjord|slep> | McMartin: Playing unintentional Core Wars, are we~? |
| 08:03 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has quit [Ping timeout: 121 seconds] |
| 08:09 | <&McMartin> | macdjord|slep: fixed now~ |
| 08:20 | <&[R]> | ,keepalive = this.keepalive |
| 08:20 | <&[R]> | ^ |
| 08:20 | <&[R]> | TypeError: Cannot read property 'keepalive' of undefined |
| 08:20 | <&[R]> | WTF |
| 08:22 | <&[R]> | Ah, my error, derp |
| 08:49 | | LadyOfLight [catalyst@Nightstar-hb5vcd.dab.02.net] has quit [The TLS connection was non-properly terminated.] |
| 08:50 | | LadyOfLight [catalyst@Nightstar-hb5vcd.dab.02.net] has joined #code |
| 08:51 | <&McMartin> | Seriously, even in machine language, flood fill is just hella slow on a C64 |
| 08:51 | <&McMartin> | It also uses Every Last Byte Of Memory. |
| 08:51 | <&McMartin> | This program uses every single byte from $0800 through $9FFF |
| 08:52 | <&McMartin> | And $A000 is where the BASIC ROM starts |
| 08:52 | <&McMartin> | Hum, wait, that's not true. |
| 08:53 | <&McMartin> | The range $1323-$1FFF is unused. |
| 08:53 | <&McMartin> | But other than that! |
| 08:53 | <&McMartin> | ($0800-$1322 for the program and data, $2000-$3FFF for the bitmap display, $4000-$9FFF as scratch space for the flood fill algorithm) |
| 09:02 | <~Vornicus> | that is a not small number of pages of flood fill |
| 09:03 | <&McMartin> | 3 bytes to store a screen coordinate, so that's 8,192 pixels worth |
| 09:03 | <&McMartin> | Turns out the largest shape on the Statue of Liberty pic is sliiiightly longer than that so filling it in will loop the ring buffer once. |
| 09:03 | <&McMartin> | I did not add instrumentation to determine the largest queue size at any given time. |
| 09:03 | <~Vornicus> | D: |
| 09:04 | <&McMartin> | (The whole screen is 64,000 pixels.) |
| 09:05 | <&McMartin> | Also the C64's palette for this is a bit uglier. The PC image is #0a0 on #aaa, which looks pretty nice |
| 09:05 | <&McMartin> | The C64 palette was almost literally selected at random |
| 09:06 | <&McMartin> | https://hkn.eecs.berkeley.edu/~mcmartin/retro/liberty.prg <-- here, have the C64 Best Combined Port |
| 09:32 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has joined #code |
| 09:32 | | mode/#code [+o himi] by ChanServ |
| 09:32 | | LadyOfLight` [catalyst@Nightstar-0brdhk.dab.02.net] has joined #code |
| 09:33 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has quit [Operation timed out] |
| 09:35 | | LadyOfLight [catalyst@Nightstar-hb5vcd.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 09:47 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has quit [Ping timeout: 121 seconds] |
| 09:48 | | LadyOfLight` is now known as LadyOfLight |
| 09:54 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has joined #code |
| 09:54 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 10:04 | | AverageJoe [evil1@Nightstar-lc3.hbj.189.210.IP] has quit [Connection closed] |
| 10:04 | | AverageJoe [evil1@Nightstar-lc3.hbj.189.210.IP] has joined #code |
| 10:11 | | * TheWatcher sighs |
| 10:12 | < LadyOfLight> | Hm? |
| 10:13 | | AverageJoe [evil1@Nightstar-lc3.hbj.189.210.IP] has quit [[NS] Quit: Leaving] |
| 10:15 | <@TheWatcher> | I become daily more ocnvinced that the network in this building is actually implemented using hydrated multistrand twisted interlinks with cylindrical aluminium endpoints. |
| 10:15 | <@TheWatcher> | convinced, even. |
| 10:17 | < LadyOfLight> | So more wetwork |
| 10:20 | <@TheWatcher> | Yeah. Which is frustrating when it's just slow, but rather worse when systems fail because the network has vanished temporarily. |
| 10:21 | <@TheWatcher> | And I'm having to look at making something far more fault tolerant than I ever should have needed to |
| 10:22 | <@TheWatcher> | Heyho. Stops me getting bored, I guess.... |
| 10:22 | <@TheWatcher> | (>.<) |
| 10:50 | | AverageJoe [evil1@Nightstar-lc3.hbj.189.210.IP] has joined #code |
| 11:01 | < LadyOfLight> | Heh |
| 11:02 | | AverageJoe [evil1@Nightstar-lc3.hbj.189.210.IP] has quit [[NS] Quit: Leaving] |
| 12:47 | | Kindamoody[zZz] is now known as Kindamoody |
| 12:48 | < LadyOfLight> | ...copy by default is a really weird behaviour unless you are actively modelling a cache based machine |
| 12:53 | < LadyOfLight> | Well, a machine based on memory/cache with a separate processor |
| 12:53 | < LadyOfLight> | And even then it's weird to have that at anything other than the lowest level |
| 13:06 | | LadyOfLight` [catalyst@Nightstar-0brdhk.dab.02.net] has joined #code |
| 13:06 | | LadyOfLight [catalyst@Nightstar-0brdhk.dab.02.net] has quit [The TLS connection was non-properly terminated.] |
| 13:37 | | celmin|sleep is now known as celticminstrel |
| 14:48 | < LadyOfLight`> | Hurray, today I got to use std::enable_if |
| 14:48 | < LadyOfLight`> | In the template type parameter list |
| 14:48 | | LadyOfLight` is now known as LadyOfLight |
| 15:15 | < LadyOfLight> | It was very exciting |
| 15:19 | <&jerith> | TheWatcher: Welcome to my world. |
| 15:20 | <&jerith> | At least the random reboots have stopped since we demanded that our VMs be moved to different hardware. |
| 16:02 | | LadyOfLight` [catalyst@Nightstar-jmdrvt.dab.02.net] has joined #code |
| 16:04 | | LadyOfLight [catalyst@Nightstar-0brdhk.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 16:22 | <&ToxicFrog> | That was not my smartest move. |
| 16:22 | <&ToxicFrog> | <co-worker> I have an [RPC response] here that's 350MB, which seems large, and now I'm wondering if anyone has something larger. |
| 16:23 | <&ToxicFrog> | <me> Back of the envelope math says I can get 1.5GB from my team's health monitoring, hang on |
| 16:23 | <&ToxicFrog> | *one reboot later* |
| 16:23 | <&ToxicFrog> | <me> I was right, but actually inspecting the contents of the response requires several times more RAM than my workstation has installed |
| 16:31 | < LadyOfLight`> | Smooooooth |
| 16:31 | | LadyOfLight` is now known as LadyOfLight |
| 16:46 | <&jerith> | :-D[6~[6~ |
| 16:46 | <&jerith> | Err, :-D |
| 16:47 | <&jerith> | Thanks for the escape character garbage, ssh+nyetwork. |
| 16:47 | < LadyOfLight> | Huzzah! |
| 16:57 | <&McMartin> | :o at std::enable_if |
| 17:02 | < LadyOfLight> | If it helps any, I was correctly defining the default constructor of std::pair |
| 17:12 | <&McMartin> | I'm... I'm not sure |
| 17:13 | < LadyOfLight> | Because it was giving me a compiler error when I was attempting to use a non default constructible type in a range based for loop by const reference and needed to instantiate the pair type to work out the size of it for the pointer math in iterators |
| 17:14 | < LadyOfLight> | And the constructor was not dependent and thus did not participate in SFINAE |
| 17:14 | < LadyOfLight> | ...if that helps any |
| 17:17 | < LadyOfLight> | (that moment when you wonder if you are hirable based entirely on your response to seeing that kind of error) |
| 17:19 | <&McMartin> | Yeah, my TMP knowledge is too shallow to even understand that response, sorry >_> |
| 17:19 | <&McMartin> | I think I can be the solution validator in that case but can't explain the solution or come up with it >_> |
| 17:28 | < LadyOfLight> | I was well aware that I was likely to be one of maybe three people at the company who could come close to figuring that out |
| 17:28 | < LadyOfLight> | So I rubber ducked at someone for a while |
| 17:29 | <&jerith> | I think I know what some of those words mean. |
| 17:37 | <&ToxicFrog> | What is std::enable_if? |
| 17:39 | <&McMartin> | To quote the reference manual I use: "This metafunction is a convenient way to leverage SFINAE to conditionally remove functions from overload resolution based on type traits and to provide separate function overloads and specializations for different type traits." |
| 17:40 | <&McMartin> | "SFINAE" is "Substitution Failure Is Not An Error" and is part of the rule for working out what overloads actually exist in the presence of templated functions. |
| 17:42 | <&McMartin> | From looking at it in use it looks like you're throwing a guard on whether a templated method exists given the values of its other type arguments, but I've never actually commanded the knowledge of how to do this or why it works. |
| 17:48 | | LadyOfLight` [catalyst@Nightstar-bt5k4h.81.in-addr.arpa] has joined #code |
| 18:15 | | Alek [Alek@Nightstar-cltq0r.il.comcast.net] has quit [Ping timeout: 121 seconds] |
| 18:18 | | Alek [Alek@Nightstar-cltq0r.il.comcast.net] has joined #code |
| 18:18 | | mode/#code [+o Alek] by ChanServ |
| 18:32 | <~Vornicus> | that is a not small rpc response |
| 18:54 | <&ToxicFrog> | No it is not! |
| 18:55 | <&ToxicFrog> | We have a lot of experiments, and health data for one experiment is (depending on what you ask for) somewhere between 1B and 300KB in size. |
| 18:56 | <&ToxicFrog> | (I mentioned this at standup and one of my teammates looked at me and said "you always find the most interesting corner cases") |
| 19:21 | | LadyOfLight [catalyst@Nightstar-jmdrvt.dab.02.net] has quit [[NS] Quit: Bye] |
| 19:32 | | LadyOfLight` is now known as LadyOfLight |
| 21:11 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has joined #code |
| 21:11 | | mode/#code [+o himi] by ChanServ |
| --- Log closed Sat Feb 18 00:00:29 2017 |