00:00:00 --- log: started retro/12.12.15
00:52:00 --- quit: karswell_ (Remote host closed the connection)
01:02:14 --- join: karswell_ (~coat@93-97-29-243.zone5.bethere.co.uk) joined #retro
01:11:22 --- quit: karswell_ (Remote host closed the connection)
01:21:37 --- join: karswell_ (~coat@93-97-29-243.zone5.bethere.co.uk) joined #retro
02:30:54 --- join: Mat2 (~Claude@91-65-144-133-dynip.superkabel.de) joined #retro
02:31:00 <Mat2> hello
03:22:43 --- quit: karswell_ (Remote host closed the connection)
03:28:49 --- quit: Mat2 (Quit: Verlassend)
03:32:59 --- join: karswell_ (~coat@93-97-29-243.zone5.bethere.co.uk) joined #retro
05:15:59 --- join: Mat2 (~Claude@91-65-144-133-dynip.superkabel.de) joined #retro
05:37:41 --- quit: karswell_ (Remote host closed the connection)
05:47:41 --- quit: Mat2 (Quit: Verlassend)
05:47:56 --- join: karswell_ (~coat@93-97-29-243.zone5.bethere.co.uk) joined #retro
06:05:04 --- quit: beretta (Quit: Leaving)
06:05:49 --- quit: karswell_ (Remote host closed the connection)
06:16:05 --- join: karswell_ (~coat@93-97-29-243.zone5.bethere.co.uk) joined #retro
06:46:43 <tangentsleep> morning all
06:46:47 --- nick: tangentsleep -> tangentstorm
06:58:48 --- join: Mat2 (~Claude@91-65-144-133-dynip.superkabel.de) joined #retro
06:58:57 <Mat2> hello
06:59:50 <tangentstorm> hey :)
07:00:28 <Mat2> hi tangentstorm !
07:00:59 <Mat2> im build ZenGL here beside
07:01:14 <tangentstorm> me too
07:01:31 <tangentstorm> actually i just checked it out of svn and ran the demos in lazarus
07:02:31 <Mat2> I must work the next 3 hours and will begin porting my code in the afternoon
07:03:19 <tangentstorm> i'm going to try and get demo09 (the sprite engine demo) hooked up to retro
07:03:35 <tangentstorm> see you later today then :)
07:03:55 <Mat2> which way ? Do you reserve some port addresses ?
07:04:43 <tangentstorm> i still like the block of ram idea, but maybe also a port to say "okay, you can draw the scene" or something
07:05:07 <tangentstorm> i'm just going to experiment with it i guess
07:05:39 <Mat2> ok see you later this afternoon
07:05:40 <tangentstorm> do you have a preference?
07:05:53 <Mat2> port devices would fit best to the VDP code
07:06:28 <tangentstorm> so maybe messages that say "sprite# x y setxy" ?
07:06:35 <Mat2> exactly
07:06:44 <tangentstorm> i think that actually works better with zengl too
07:06:56 <tangentstorm> ok. i'll shoot for that then
07:07:04 <Mat2> fine, see you
07:07:08 <tangentstorm> seeya
07:07:09 --- quit: Mat2 (Quit: Verlassend)
07:07:44 <carefish> grrr, almost got my vector class exposed to javascript :P
07:07:52 <carefish> also, mornin' tangy 
07:08:05 <tangentstorm> haha morning
07:20:57 --- join: kumul (~kumul@cvx-ppp-66-50-141-36.coqui.net) joined #retro
07:21:29 --- quit: karswell_ (Remote host closed the connection)
07:29:27 --- join: intothev01d (~intothev0@unaffiliated/intothev01d) joined #retro
07:31:43 --- join: karswell_ (~coat@93-97-29-243.zone5.bethere.co.uk) joined #retro
07:33:24 <tangentstorm> crc's compiling up a storm :)
07:35:13 <tangentstorm> crc: i love the idea of multiple backends. i wanted to do at least js + pascal + retro.
07:36:13 <tangentstorm> my thought was to decouple the code generation from the syntax tree by using a template system
07:37:13 <tangentstorm> either stringtemplate in particular (specificically designed for syntax trees - it's part of ANTLR, and has a python port)... or jinja2 (really nice python template engine)
07:43:16 <crc> tangentstorm: templating would be nice for this
07:49:45 <tangentstorm> crc: so i guess we're shooting for a small sprite-driven game of some kind instead of a text adventure
07:50:33 <tangentstorm> the notes are here: https://docs.google.com/document/d/17tXStyPkBdfnIl0cu_PmQBaA4ipus-vmZ7yZE0XcyKM/edit
08:04:58 <crc> tangentstorm: http://sprunge.us/FQQG has a version of fiction' with an .onDraw field and a 'draws' word for attaching code to render something to a room/item
08:06:05 <tangentstorm> :) cool!
08:06:21 <tangentstorm> crc: do you have a google account so i can invite you to the doc?
08:06:35 <tangentstorm> i'll also add you to the github repo as soon as i check this demo in
08:06:41 <crc> charles.childers@gmail.com
08:11:10 <tangentstorm> sent
08:13:58 --- quit: karswell_ (Remote host closed the connection)
08:24:14 --- join: karswell_ (~coat@93-97-29-243.zone5.bethere.co.uk) joined #retro
08:43:25 --- join: Mat2 (~Claude@91-65-144-133-dynip.superkabel.de) joined #retro
08:43:44 <Mat2> tangentstorm: How's work going ?
08:45:14 <Mat2> We have a little problem: As I see it ZenGL does not support raw keyboard input
08:47:40 <Mat2> only key detection
08:50:16 <Mat2> so you can check if a specific key code was pressed but have no routine for synchron key input (remember the problem with he key repeat rate?)
08:50:30 <tangentstorm> hmmm :/
08:50:58 <tangentstorm> well
08:52:38 * tangentstorm looks at the input demo
08:53:15 --- quit: karswell_ (Remote host closed the connection)
08:53:32 <tangentstorm> Mat: not to worry. it looks like it's just poorly documented :)
08:53:54 <tangentstorm> zgl_keyboard contains key_Down, key_Up, key_Press...
08:54:07 <tangentstorm> oh... hrm
08:58:43 <Mat2> yes, you can test if a specific key was pressed
08:59:15 <Mat2> this would mean polling all characters for input
09:01:13 <tangentstorm> You actually want this in a game engine sometimes, but... hrm...
09:02:00 <tangentstorm>   key_PPress     : procedure( KeyCode : Byte );  
09:02:10 <tangentstorm> it's listed under "callback"
09:02:29 <tangentstorm> i'm wondering if you can register that to receive the events...
09:02:39 <tangentstorm> but i don't see where it's used yet
09:03:05 <tangentstorm> if it were the private thing getting data from the os, it shouldn't be in the interface
09:03:29 --- join: karswell_ (~coat@93-97-29-243.zone5.bethere.co.uk) joined #retro
09:03:59 <tangentstorm> yeah
09:04:08 <tangentstorm> it's used in zgl_application
09:04:27 <tangentstorm> so if you assign a procedure to that variable, it'll call the procedure
09:04:42 <tangentstorm> and same for key_PRelease
09:05:37 <Mat2> ahm, and the key code is then returned in KeyCode ?
09:06:03 <tangentstorm> yep
09:06:22 <tangentstorm> right after it adds it to the array of booleans
09:07:06 <tangentstorm> in zgl_application.pas ... it's actually duplicated a couple times - once for each platform :/
09:07:17 <Mat2> I take a look
09:07:41 <tangentstorm> messy, but i think it'll work :)
09:08:13 <tangentstorm> yeah, a lot of game engines do that so you can just put all your logic in one single "onFrame" handler
09:12:40 <Mat2> ok:
09:15:43 <Mat2> there is one internal variable 'key' which hold the UTF8 code of the last pressed key
09:16:19 <Mat2> beside the used xlib routine will only return ASCII characters
09:17:10 <tangentstorm> key_PInputChar : procedure( Symbol : UTF8String ); 
09:17:26 <tangentstorm> we only care about that for typing text though
09:17:58 <tangentstorm> for controlling a sprite, we just want to know that a key has been pressed
09:18:13 <tangentstorm> like we just care about the scan code
09:18:18 <tangentstorm> not which character it produces
09:19:11 <tangentstorm> i'm kind of thinking we actually want a separate keyboard device for this
09:19:26 <tangentstorm> that works the way 
09:19:28 <tangentstorm> er 
09:19:39 <tangentstorm> the normal one for text input
09:19:51 <tangentstorm> and this one that treats the keyboard more like a set of "buttons"
09:20:30 <tangentstorm> your game code really does come out much cleaner if you just have everything happen on frame, rather than on each input event
09:20:42 <Mat2> I let the SDL keyscan routine in, it works beside the graphic initialisation
09:20:59 <tangentstorm> ok
09:21:09 <tangentstorm> do you have a github account? i'll ad you to the project
09:21:47 <Mat2> I must reactivate my old account
09:22:51 <Mat2> I think the sprite subsystem only need an OpenGL layer ?
09:23:57 <Mat2> in this case, I only need to init one from SDL and you can use the handle for your work
09:24:30 <tangentstorm> Mat2: i like zengl but there's no point having it slow us down :)
09:25:01 <tangentstorm> as long as we can agree on an interface, it doesn't matter what implements it on the backend
09:25:08 <tangentstorm> so... want to just use sdl then?
09:25:18 <Mat2> yes
09:25:40 <tangentstorm> okie doke. i'll come back to zengl when we don't have the time pressure :)
09:25:47 <Mat2> otherwise we using two graphic libraries
09:26:07 <tangentstorm> works for me
09:26:36 <Mat2> by the way, you have compiled an old version of the VDP code for win32
09:27:36 <tangentstorm> oh? I thought i grabbed the latest version before i compiled it. :/
09:27:53 <Mat2> that was the reason for his display problem
09:28:33 <Mat2> the newest one works like a real display (and offered a centred display)
09:28:39 --- join: docl (~docl@ds4403.dreamservers.com) joined #retro
09:28:39 --- quit: docl (Changing host)
09:28:39 --- join: docl (~docl@unaffiliated/docl) joined #retro
09:32:48 <tangentstorm> well, we'll get it sorted out when we compile this one, then :)
09:33:40 <tangentstorm> okay. i moved the repo to here https://github.com/goldsmile/rpld25 and invited everybody who's told me their github name :)
09:34:31 <Mat2> I need to search for my old account one moment
09:34:36 <tangentstorm> Mat2: they need to install jedi-sdl separately, right?
09:34:56 --- quit: karswell_ (Read error: Connection reset by peer)
09:35:12 <Mat2> no, the lib should be included with Freepascal 2.6
09:35:35 <Mat2> at least for the linux install package i is
09:35:45 <Mat2> ^it
09:45:10 --- join: karswell_ (~coat@93-97-29-243.zone5.bethere.co.uk) joined #retro
09:46:04 <tangentstorm> i'm going to break ngaro.pas into its own repository so we can include it as a submodule without all my other b4 junk
09:46:59 <Mat2> ok
09:47:32 --- quit: karswell_ (Remote host closed the connection)
09:52:20 <tangentstorm> https://github.com/tangentstorm/ngaro-pas
09:53:15 <tangentstorm> ah there we go
09:53:49 <tangentstorm> i need to set up the lib dir real quick before you'll be able to compile
09:59:23 --- join: karswell (~coat@93-97-29-243.zone5.bethere.co.uk) joined #retro
10:03:23 <Mat2> I'm back in one and a half hour
10:03:31 --- nick: Mat2 -> Mat2-afk
10:11:01 <tangentstorm> there's probably an easier way to give everybody access to everything than what i'm doing ... :/
10:11:30 <tangentstorm> but i guess next time around it'll already be set up, and i can just add people to the team...
10:12:36 <tangentstorm> anyway, i've added team forks for xpl, ngaro-pas, and retro
10:26:05 <tangentstorm> okay... hopefully if you're on the team at github, you can just run the 3 commands listed here and be up and running:
10:26:08 <tangentstorm> https://github.com/goldsmile/rpld25
10:27:09 <carefish> fuck yes
10:27:16 <carefish> i solved that template instatiation shit
10:27:31 <carefish> and now i can call Vector from javascript
10:37:25 <carefish> anyway, tangy :P is there anything i can help with? :)
11:00:22 --- nick: Mat2-afk -> Mat2
11:00:47 <Mat2> you can help me porting he Lord Of Midnight Engine :D
11:02:01 <Mat2> to be precise because of the time restrictions we need every helping hand for map design, graphic stuff, coding ec.
11:05:04 <tangentstorm> i've got people here distracting me... should quiet down soon. be back shortly.
11:05:56 <tangentstorm> Mat2: i did everything except merge your changes into the ngaro-pas repo. you should have rights to edit it
11:06:16 <tangentstorm> carefish: what's your github username?
11:07:34 <Mat2> thanks
11:09:09 <tangentstorm> also you should be able to just check out the top level one, run make retro, and it'll fetch the others under the "lib" directory
11:10:54 <Mat2> I'm extending at current the canvas pot for interfacing font rom, char and attribute map
11:12:05 <Mat2> after this is finished, retro-pascal is able to blit graphics to screen regardless of the initiated mode, e.g. there will be displayed in text or graphic mode
11:12:27 <tangentstorm> :) cool
11:12:42 <tangentstorm> back in a few
11:12:45 <Mat2> ok
11:13:41 <carefish> Mat2: okay, tell me what to do, where to find stuff, gotta add to this: i only know C#/C++, don't really know other languages
11:14:03 <carefish> tangentstorm: not sure if i have a github, well.. i should have one, lemme find it
11:14:42 <carefish> carefish is my github's username
11:24:15 <Mat2> at first you should compile retro from tangentstorms repro
11:25:38 <Mat2> here you find the sources for Lord Of Midnight:
11:25:41 <Mat2> http://www.icemark.com/downloads/
11:26:24 <Mat2> the graphic engine is an 16 bit assembly file for the Intel 8086 architecure
11:28:06 <Mat2> we want to get some inspirations from the math behind it and render screens with fixed sights directly from a map array
11:29:02 <Mat2> which holds every object for the game (Trees, enemies, mountains) you get the idea
11:29:40 <Mat2> It would help me if you find a nice algorithm for constructing such a map
11:30:36 <Mat2> and some routines for accessing locations (array offsets)
11:31:01 <carefish> fixed sights?
11:31:43 <Mat2> 8 or 4 view positions 45°
11:32:04 <carefish> well, that map is do-able, what i do in C# is have a globally acessable list<GameObject> and push new objects that derive from gameobject in that list in the ctor
11:32:06 <Mat2> so N,NW,W,SW,E
11:32:58 <carefish> tree : Gameobject { tree() { Globals.DynamicGameObjects.Add(this); }}
11:33:10 <Mat2> ok, sadly I don't know C# but know C and C++
11:33:14 <carefish> tree is a bad example since that's a static object
11:33:34 <carefish> so then i'd ust Globals.StaticGameObjects.Add(this);
11:33:39 <Mat2> I would prefer retro but find the language your comfortable with
11:33:51 <carefish> i need lots of guidance when using retro
11:33:59 <carefish> tried it out last night, it confused the crap out of m
11:34:00 <carefish> e
11:34:02 <Mat2> that's not the problem
11:34:26 <Mat2> I will help you out
11:34:39 <carefish> and when you say a 'fixed sights
11:35:18 <carefish> it would still be one global list where you can have a variable (or enum if you prefer) that says what it's currently pointing at
11:35:50 <carefish> and then in the render code only render the objects that have the proper N/W/E/etc flag by checking what the player currently is viewing
11:36:17 <Mat2> that is a good possibility, yes
11:36:51 <carefish> also quite fast since you won't be rendering objects that have an incorrect flag
11:37:00 <carefish> like, 3 seperate lists
11:37:21 <carefish> CurrentlyRendering / AboutToRender / NotVisible
11:37:31 <carefish> player is looking at N sight
11:37:43 <carefish> abouttorender holds NE/NW sights 
11:37:57 <Mat2> I have figured out the Lord Of Midnight engine just rendering every object by id which is stored in a fixed sized array-element of a fixed, two-dimensional map
11:38:16 <carefish> i'm grabbing tangy's retro thingy now
11:38:28 <Mat2> fine
11:38:44 <Mat2> so the playmap holding only visible objects
11:38:56 <carefish> yup, basically
11:40:04 <carefish> https://github.com/goldsmile/retro-language <- this one?
11:40:56 <Mat2> https://github.com/goldsmile/rpld25
11:41:10 <Mat2> that's the link tangentstorm posted earlier
11:41:48 <Mat2> ok, I', working now on this special retro vm with basic blitting support (I love AMIGA inspired copper lists)
11:42:25 <tangentstorm> back
11:42:32 <Mat2> wb
11:43:11 <tangentstorm> I've got a basic API sketched out under "arcade device operations"
11:43:34 <Mat2> if I had know this earlier I had include the whole AGA chipset emulation from UAE
11:43:37 <tangentstorm> let me know if it'll work for you or what needs to change
11:43:44 <Mat2> I take a look
11:43:45 <tangentstorm> :)
11:44:03 <Mat2> carefish will working on the graphic engine
11:44:18 <tangentstorm> cool
11:44:28 <carefish> okay, i've downloaded free pascal/rpld25
11:44:29 <Mat2> bu he need help with retro
11:44:45 <carefish> what next? :D 
11:44:54 <Mat2> get firm with retro :)
11:45:09 <tangentstorm> carefish: are you able to compile it?
11:45:15 <Mat2> on www.retroforth you find the documentation
11:45:50 <carefish> i'm not sure how to compile it on windows :(
11:46:07 <tangentstorm> fpc comes with make
11:46:13 <tangentstorm> so just run make
11:46:21 <tangentstorm> make retro
11:46:22 <carefish> make from the gitrepo?
11:46:25 <tangentstorm> yes
11:46:28 <carefish> ok
11:46:30 <Mat2> I'm working on the ngaro vm yet and will be back in one hour or so (so concentrate on coding)
11:47:05 --- nick: Mat2 -> Mat2-speed-codin
11:47:26 <tangentstorm> i guess i will start making some kind of simple game then
11:47:41 <tangentstorm> i will start implementing this api on the retro side
11:47:57 <carefish> C:\Projects\Git\rpld25>C:\FPC\2.6.0\bin\i386-win32\make retro
11:47:57 <carefish> The syntax of the command is incorrect.
11:47:57 <carefish> C:\FPC\2.6.0\bin\i386-win32\make: *** [init] Error 1
11:48:24 <tangentstorm> yuck
11:53:52 <Mat2-speed-codin> ok sprite engine is fine
11:54:00 <Mat2-speed-codin> for me
11:55:30 <tangentstorm> cool
11:55:44 <tangentstorm> carefish: did you install lazarus?
11:55:58 <carefish> it's got to do with UAC
11:56:39 <carefish> oh
11:56:58 <carefish> i simply downloaded the fpc installer, and bashed next->next->next/etc
11:57:59 <tangentstorm> well you can compile ngaro with the "fp" IDE that comes with freepascal
11:58:15 <tangentstorm> it's the old console-mode IDE
11:58:25 <Mat2-speed-codin> or use fpcmake
11:58:42 <carefish> downloading lazarus
11:59:16 <tangentstorm> i've got to do this too because i only have graphics on windows :)
12:00:39 <carefish> installing lazarus
12:02:26 <carefish> and now what? load up the retro.pas ?
12:03:31 <tangentstorm> yeah.. see if you can get it to compile... i'm trying the same on my end here
12:03:51 <tangentstorm> or you can just wait a few minutes... i'll get it set up :)
12:04:13 <carefish> what folder do i need to look at?
12:04:32 <carefish> rpld25/lib/retro?
12:04:44 <carefish> lib/ngaro-pas ?
12:04:46 <tangentstorm> lib/ngaro-pas 
12:04:52 <tangentstorm> that's the one we're working on
12:04:55 <carefish> ah
12:05:07 <tangentstorm> lib/retro is there so we can get the retro libraries 
12:05:31 <carefish> well, it's opened, can i can press the play button and no errors?
12:05:39 <tangentstorm> try it
12:05:53 <carefish> an empty form shows up
12:06:27 <tangentstorm> ok give me a minute then... i've only been using the command line up until now.
12:06:32 <tangentstorm> i've used lazarus, just not with this
12:06:34 <carefish> okay :)
12:06:52 <carefish> pascal doesn't seem too difficult to get a hang off
12:07:15 <tangentstorm> nah, it's a lot like java or c#
12:07:34 <tangentstorm> in fact the guy who desigend c# previously worked for many years on delphi
12:07:55 <tangentstorm> (which is what they called turbo pascal for windows)
12:10:55 <carefish> oh, nice fun fact to know :)
12:11:01 <Mat2-speed-codin> tangentstorm has written a pascal-subset to retro compiler you can use
12:12:39 <Mat2-speed-codin> (you can also create your own working language in retro if you will, that's possible)
12:12:41 <tangentstorm> carefish: do you actually have files under lib? because my windows make didn't even get that far
12:13:09 <carefish> C:\Projects\Git\rpld25\lib
12:13:17 <tangentstorm> if not, run "git submodule init; git submodule update"
12:13:25 <carefish> that folder contains 3 folders ngaro-pas / xpl  / retro
12:15:42 <tangentstorm> yeah.. they should have files inside.. run the two git commands i just posted
12:18:29 <carefish> those folders do have files in them
12:19:54 <carefish> and uhh.. i have the visual client of git
12:20:17 <carefish> i used some git.exe in AppData\etc\etc\git
12:21:12 * Mat2-speed-codin windows makes the world go round...
12:21:51 <tangentstorm> ok cool
12:22:26 <carefish> like, i didn't get compile/run errors in lazarus, so i should be fine right?
12:22:47 * Mat2-speed-codin accessing char and attribute map, vsync control and internal register access should now work
12:22:49 <carefish> http://img.is1337.eu/9d7dc8b0.png
12:22:49 <tangentstorm> is there a retro.exe?
12:23:12 <carefish> no, there is a project1.exe in some temp folder
12:23:50 <tangentstorm> i think lazarus made you a new empty program
12:23:56 <tangentstorm> that's why it's showing the form1 thing
12:24:00 <carefish> quite possibly.
12:24:36 <tangentstorm> ok to brute force it, you can copy xpl/code/stacks.pas to your ngaro-pas/src directory
12:24:51 <tangentstorm> and then just run fpc retro.pas from the command line
12:25:08 <tangentstorm> no hang on
12:29:31 <tangentstorm> okay i just added a batch file to the top level directory until i can get make sorted out
12:29:49 <tangentstorm> this will build the pascal version of retro
12:30:19 <Mat2-speed-codin> ok, it works so far
12:30:35 <tangentstorm> carefish: mkdir gen first if it's not there
12:30:45 <carefish> mkdir gen where?
12:30:53 <carefish> ngaro-pas?
12:30:56 <tangentstorm> from the rpld25 directory
12:31:06 <carefish> okay, done
12:31:19 <tangentstorm> git pull to get the batch file, and then run it
12:31:25 <tangentstorm> it should make gen\retro.exe
12:31:34 <tangentstorm> then you need retroImage from lib\retro in the same directory
12:31:40 <carefish> what batch file?
12:31:47 <tangentstorm> git pull
12:31:52 <tangentstorm> there's only one batch file
12:32:45 <tangentstorm> hang on it didn't push when i said push :/
12:33:09 <carefish> it built retro.exe
12:33:11 <tangentstorm> there it is. https://github.com/goldsmile/rpld25/blob/master/build.bat
12:33:44 <carefish> okay, i have a retro terminal/cmd window
12:33:50 <tangentstorm> cool :)
12:34:16 <carefish> and now what? :)
12:34:23 <Mat2-speed-codin> tangentstorm: we have two possibilities now: I can build the shape routines from the char map (that's the way the ti9918 VDP chip handles bitmap displays). This have he true retro effect of allowing only two colours from 256 for each 8x8 attribute box or I can blit the shapes directly (every point can be of any 256 colours). The first way have one advantage: We can reuse the graphics from LOM
12:35:14 <Mat2-speed-codin> I mean directly from the sources
12:35:18 <tangentstorm> yeah
12:35:58 <Mat2-speed-codin> what are our preferences ?
12:36:23 <tangentstorm> so: my long term goal here is to make a reusable game engine with full-color graphics and all that... that's why i'm looking an zengl and trying to get this to run on js too...
12:36:43 <carefish> so then the latter option is handier?
12:36:52 <tangentstorm> but right now this is just a prototype, so i'm fine with whatever you want to do
12:36:57 <carefish> i wouldn't know how hard it would be to create LOM graphics ?
12:37:11 <tangentstorm> if you want to reuse LOM and tie it in to the blitter you already have, let's go for it :)
12:37:23 <tangentstorm> then we can focus 100% on the engine and not worry about graphics
12:37:31 <Mat2-speed-codin> http://www.youtube.com/watch?v=ilDgNSIhW5M
12:37:35 <Mat2-speed-codin> that are the game graphics
12:38:04 <tangentstorm> carefish: lords of midnight. it's a really old game but it did some cool stuff with arranging 2d graphics to simulate a 3d scene
12:38:21 <carefish> i saw the gameplay yt vids yesterday
12:39:19 <Mat2-speed-codin> carefish: it's not easy because the shapes are designed to be build to a complete screen so there must fit well together 
12:39:34 <tangentstorm> Mat2-speed-codin: as long as we're controlling sprites from retro i'm happy. let's do it that way
12:39:45 <tangentstorm> i think that fits in better with what you want to make in general too, right?
12:41:14 <Mat2-speed-codin> somewhat because I don't need to reserve a third buffer in graphic mem for the bitmap data
12:41:35 <carefish> so tangentstorm, you had a pascal2retro compiler made?
12:41:58 <carefish> because i think i can't get the hang of retro fast enough to actually be helpful, and i think i can manage that in pascal
12:42:26 <tangentstorm> carefish: pl/0 ;)
12:43:13 <tangentstorm> https://github.com/tangentstorm/PL0-Language-Tools
12:43:58 <tangentstorm> okay... so let's regroup here... we're basically talking about porting the LOM engine to pascal from the old assembly language?
12:44:14 <tangentstorm> or PL/0 + retro, rather
12:44:47 <tangentstorm> and then making some new game with tha engine?
12:44:53 <tangentstorm> that
12:45:45 <carefish> okay, so how do i use that repo?
12:46:29 <tangentstorm> well, if that's what we're doing then you need python too
12:47:02 <tangentstorm> you just run python pl0_to_retro.py < tests/whatever.pl0 > whatever.rx
12:48:04 <Mat2-speed-codin> I have an idea: Carefish, if this would not problematic for you, you can write your code in pascal and I include the result as device into ngaro
12:48:51 <Mat2-speed-codin> this way we can use retro for scripting (tangentstorm can use his pl/0 compiler for example) and concentrate complete on the game design
12:49:19 <Mat2-speed-codin> (ngaro is the vm on which retro is build)
12:49:22 <tangentstorm> Mat2-speed-codin: can you post the LOM source code link again?
12:50:39 <Mat2-speed-codin> un moment
12:51:06 <Mat2-speed-codin> http://www.icemark.com/downloads/
12:51:16 <tangentstorm> thanks
12:51:24 <carefish> hmm, so i'll open up a new pascal project in lazarus, try to figure out how draw pic to screen and then build from there for the fixed sight/etc?
12:52:24 <Mat2-speed-codin> yes, because ngaro itself is written in pascal I can include your code as device so retro can access your interface
12:52:31 <carefish> oh, awesome
12:52:36 <carefish> well, i'll get to it
12:53:04 <tangentstorm> that site isn't loading for me
12:53:23 <carefish> in lazarus, when making a new project, what should i select?
12:53:37 <Mat2-speed-codin> tangentstorm: I have no problem accessing it
12:53:56 <carefish> app/program/console app/library/instantfpc/fpcunit console test app/fpcunit test app?
12:54:38 <tangentstorm> probably just program
12:55:09 <carefish> let's have a looksy at a few pascal thingies
12:56:18 <tangentstorm> carefish: a good start would be to merge Mat2's existing SDL backend into the ngaro-pas directory
12:56:21 <tangentstorm> https://www.assembla.com/code/avm/git/nodes/master/metro/backport
12:56:34 <Mat2-speed-codin> I suggest you should use SDL for drawing because the documentation is good and the interface will be familiar to you
12:57:24 <Mat2-speed-codin> but please wait a minute before access my repro I will upload the actual state soon
12:58:38 <Mat2-speed-codin> tangentstorm: I get an error with your data.pop routine, it accept only two parameters instead of one ?!???
13:00:52 <tangentstorm> that doesn't make sense :) looking...
13:01:56 <tangentstorm> stack.pop is a function ... it doesn't take any parameters
13:02:09 <Mat2-speed-codin> ups, sorry
13:02:13 <tangentstorm> stack.pop1( x ) will take the parameter
13:02:32 <tangentstorm> stack.pop2( t, n )   will get tos and nos into t and n
13:03:01 <Mat2-speed-codin> ok, have overread something :(
13:03:35 <tangentstorm> no worries :)
13:04:06 <tangentstorm> i still can't see the site from home but i got the source code through my server. downloading now
13:04:42 <Mat2-speed-codin> hmm, that is remarkable
13:05:03 <Mat2-speed-codin> seems like the IP address is not global accessible
13:05:03 <tangentstorm> do you want me to just clone your avm repo?
13:06:23 <Mat2-speed-codin> import romVDP.pas, romFont.pas, ng.pas and ng.ports.pas, this should work
13:06:39 <tangentstorm> ok
13:08:58 <Mat2-speed-codin> so sources are updated
13:09:38 <Mat2-speed-codin> VDP control must be wrapped into a word set for retro though
13:09:53 <Mat2-speed-codin> and I bet here are some errors ... somewhere
13:10:08 <Mat2-speed-codin> but it works at current 
13:10:39 <Mat2-speed-codin> I'm now meditate about implementing the shape functionality
13:10:40 <tangentstorm> 5,003 lines of assembly :)
13:11:36 <Mat2-speed-codin> oh, git has also merged my current JIT sources :)
13:11:52 <tangentstorm> :)
13:12:18 <tangentstorm> i look forward to reading them after this thing is over :)
13:14:58 <Mat2-speed-codin> it would help carefish if you update your repro soon so he can get reuse the VDP sources and test things out
13:15:22 <tangentstorm> ok.. doing now
13:15:45 <tangentstorm> you guys actually all have write access to everything already, btw
13:16:00 <tangentstorm> but i'll do this one. hang on.
13:20:34 <Mat2-speed-codin> crc: I was free to reserve some port addresses for an interface to a video-display processor (ports 9-20)
13:20:56 <carefish> okay, so this was my take on their 'fixed sight' idea
13:21:36 <carefish> you have a 2d array, array[8][sort-ofvariablelengthbecauseitholdswhatneedstoberendered]
13:21:45 <Mat2-speed-codin> crc: by the way how can I access ports from retro without compiling ngaro instructions ?
13:21:47 <carefish> and the first array, the 8, are the directions
13:22:14 <carefish> so you simply cycle through those and each holds all the sprites needed, and will be updated if the level changes/room changes
13:22:26 <Mat2-speed-codin> ok
13:22:34 <Mat2-speed-codin> that's simple
13:22:58 <carefish> i have no idea how to show images on the screen using pascal to show this concept
13:23:03 <carefish> googling now though :p
13:23:28 <Mat2-speed-codin> take a look at my sources in the file: romVDP.pas
13:24:19 <Mat2-speed-codin> SDL is a very basic library so you must program some routines for plotting graphic data to screen
13:24:42 <Mat2-speed-codin> don't worry about double buffering etc, this is handled already
13:25:17 <carefish> i use SDL in C++ :o does sdl have pascal bindings?
13:25:55 <Mat2-speed-codin> yes
13:26:23 <Mat2-speed-codin> there should already be installed with fpc
13:26:39 <carefish> and it's a 'simple' uses clause?
13:26:41 <tangentstorm> i need a few minutes here... i need to merge the changes
13:26:52 <tangentstorm> carefish: google freepascal meets sdl
13:27:01 <Mat2-speed-codin> right just: use SDL;
13:27:04 <Mat2-speed-codin> that's it
13:27:27 <carefish> http://img.is1337.eu/a78ad6d3.png
13:27:29 <carefish> ^ like that?
13:28:07 <carefish> ouch. my eyes bleed from that website, not only that i feel like ive travelled back in time
13:28:25 <Mat2-speed-codin> yes. (pascal have a real module concept, you can compile each unit independent and don' need to worry about library paths)
13:29:05 <Mat2-speed-codin> oh, I feel familiar because learned programming with UCSD pascal in school
13:29:06 <tangentstorm> yeah, and you can delete all that cthreads stuff
13:29:53 <Mat2-speed-codin> (ok, a this time my main language was already 6502 assembly and basic)
13:29:59 <Mat2-speed-codin> ^at
13:30:00 <carefish> this is amazing
13:30:13 <carefish> i knew it was worth it to learn SDL instead of SFML/allegro
13:30:35 <Mat2-speed-codin> I have some bad experiences with allegro
13:31:00 <Mat2-speed-codin> not with the library per se but with its building
13:32:56 <carefish> well, when i decided to never ever touch AS3 again, i did some research on what to pick next: C# XNA and C++/SDL w/ opengl came out of that research
13:34:07 <Mat2-speed-codin> I think functional programming is important in future because of the trend to multicore envirionments
13:34:35 <carefish> what does ^ mean in pascal?
13:34:53 <carefish> well, i don't study comp sci or stuff like that
13:35:10 <Mat2-speed-codin> so I'm working on a functional retro dialect which autoparallelisation support
13:35:18 <tangentstorm> pointer^  dereferences the pointer
13:35:29 <carefish> i'm on a gamedevelopment course that only cares about a finished product.. so i barely know anything about programming and what i do know is what i've taught myself :P
13:35:34 <tangentstorm> it's also used to mean control when talking about the characters... ^A, ^B, ^C..
13:35:44 <carefish> yeah tangentstorm, the second i asked that question i realized what it must've meant
13:36:16 <carefish> the pSDL stuff was ovbious: create an SDL_SURFACE that holds a pointer
13:37:19 <carefish> the := is weird to get used to
13:38:04 <Mat2-speed-codin> not really, it is simple the mathematic standart notation
13:38:40 <Mat2-speed-codin> (no one teaches at school)
13:39:57 <carefish> haha, i don't get linear algebra/algebra/whatever
13:40:08 <carefish> they don't teach me that at school i mean, it's a community college
13:40:29 <carefish> so most things that i know about vector math is also selftaught :P
13:40:39 <Mat2-speed-codin> by the way UPN notation was used in Poland as alternative to it and partly in Germany
13:41:24 <Mat2-speed-codin> what is the status of a community college ?
13:41:40 <carefish> not university
13:41:42 <carefish> not college
13:41:46 <carefish> uhh
13:42:07 <carefish> you have school for 5 ~ 12 y/o, from 12 to 16, and from 16 to 27
13:42:33 <carefish> i'm in the latter, because the educational system fucked me over and i didn't get any degrees/diploma's on the 12 to 16 school :P
13:42:46 <carefish> (don't know the proper terminology for school systems in my country in english)
13:43:25 <Mat2-speed-codin> where do you live ?
13:44:18 <carefish> the short version: middleschool/highschool they (the gov't) changed the requirements to be allowed to the the final exam, and i was in my last year that year, so i couldn't do the final exam because i didn't fit the requirements (as did all of my other classmates)
13:44:23 <carefish> Netherlands, Amsterdam
13:45:37 <Mat2-speed-codin> h, het land aan de Noordzee
13:46:42 <carefish> ja, dat land ja
13:47:12 <Mat2-speed-codin> Mijn zus spreekt Nederduits
13:48:10 <carefish> Ah, dan ken je 't fenomeen MBO wel, ik zit nu vast op een kut MBO school die geen wiskunde B/wiskunde at all geeft, ook geen programmeer lessen op een paar trucjes in Flash na, dus moet ik meeste mezelf aanleren xD
13:48:25 <carefish> also, i'm having fun with pascal.. i think i'll keep on using it on the side :)
13:48:37 <Mat2-speed-codin> good :)
13:49:10 <carefish> du sprecht Deutsch?
13:49:20 <carefish> ich spreche ein bisschen Deutsch haha
13:49:33 <carefish> (not nederduits though.)
13:50:04 <Mat2-speed-codin> I'm german but living in a multi-language family
13:51:34 <tangentstorm> mat2 : you sure those are the only files you changed?
13:51:51 <Mat2-speed-codin> have you problems compiling retro ?
13:51:56 <Mat2-speed-codin> otherwise I'm sure
13:51:58 <Mat2-speed-codin> :)
13:52:08 <tangentstorm> no problems compiling, but it doesn't work when i run it
13:52:15 <Mat2-speed-codin> uh ok
13:52:26 <carefish> ah, that's cool Mat2-speed-codin :) i've always loved languages, i can speak a few
13:52:37 <tangentstorm> i hand merged your changes.. maybe i missed something
13:52:53 <carefish> dutch/english/hebrew/german (i used to be quite good at german too... although haven't used it actively in years)
13:53:43 <Mat2-speed-codin> most germans can understand dutch more or less, it's not a so different language
13:54:13 <Mat2-speed-codin> amgentstorm: I take a look
13:55:44 <tangentstorm> hang on i skipped something
13:56:27 <carefish> http://img.is1337.eu/8d2f2d7b.png ? what am i doing wrong?
13:57:06 <Mat2-speed-codin> the const directive:
13:57:16 <Mat2-speed-codin> look at the first definaion and hen the others
13:57:22 <Mat2-speed-codin> ^definations
13:57:43 <Mat2-speed-codin> you must define a constant with CONST name = <type>
13:57:48 <Mat2-speed-codin> a variable with
13:57:58 <Mat2-speed-codin> VAR name: <type>
13:58:08 <Mat2-speed-codin> sorry CONS name = <value>;
13:58:35 <Mat2-speed-codin> so write:
13:58:48 <Mat2-speed-codin> CONST totalsprites = 256;
13:59:05 <Mat2-speed-codin> VAR direction ...
13:59:41 <carefish> hm, i'm getting debugger error messages
14:00:04 <Mat2-speed-codin> what errors ?
14:00:11 <carefish> http://img.is1337.eu/233f689d.png
14:01:06 <Mat2-speed-codin> your code seems to compile but without debug informations. This couses the debugger to crash
14:01:18 <carefish> that's a bummer
14:03:37 <tangentstorm> ok!
14:03:55 <Mat2-speed-codin> so retro starts ?
14:04:05 <tangentstorm> carefish: git submodule update ; ./build.bat
14:04:16 <carefish> hmm, even when copy/pasta'ing the example from sdl meets pascal, and making sure the director for the image existed... it still won't run
14:04:27 <carefish> LOLO
14:04:35 <tangentstorm> yep. with nice green letters on purple background. just like barney and friends :)
14:04:36 <carefish> it didn't place sdl.dll in the proper place
14:04:50 <carefish> well, place being: i should've placed it there
14:05:28 <tangentstorm> ahh.. yeah' that sdl thing is annoying on windows. dll management is really clunky in fpc on windows
14:05:30 <carefish> that wa the problem
14:06:01 <Mat2-speed-codin> corefish: sorry I don't have a Windows system for testing but bet that is one example of the feared DLL hell
14:06:07 <carefish> yeah, that's why i used static sdl libs
14:06:24 <carefish> im gonna have fun w/ pascal
14:06:30 <carefish> loving this ludum dare already :D
14:06:35 <tangentstorm> carefish: from the command line, fpc -gl adds the debug trace
14:06:48 <tangentstorm> :)
14:07:19 <carefish> can i enable that so lazarus does that by default?
14:07:26 <Mat2-speed-codin> you should try to copy all DLL's into the home directory of Lazarus (where ever this directory exist)
14:07:45 <Mat2-speed-codin> yes, just edit the compiler flags
14:08:07 <Mat2-speed-codin> for your project
14:08:11 <tangentstorm> carefish: yeah, under project options
14:09:57 <Mat2-speed-codin> carefish: the best place for DLL's is the WINDOOF, sorry WINDOWS directory (or it's virtual representation under Win Vista+7)
14:10:26 <carefish> Mat2-speed-codin: i disagree with that :) i always provide the proper .dlls in the .exe's working dir
14:10:39 <carefish> to be absolutely sure it works on every windows pc
14:11:36 <carefish> although i like my current solution alot more: statically linked, so all i have to give is a .exe :)
14:12:46 <carefish> and my lil' bro is awesome, just came home w/ a bag o' grass :D
14:12:54 <Mat2-speed-codin> that don' work for some librarys like OpenGL drivers of some prominent cpu coorporaions because there had embedded a fixed pah into there libarary files
14:13:05 <Mat2-speed-codin> ^corporaions
14:13:30 <Mat2-speed-codin> or cooporations, or corporations ???
14:13:33 <Mat2-speed-codin> however
14:14:18 <carefish> how to do a bit of OOPness in pascal?
14:14:27 <carefish> corporations
14:14:39 <Mat2-speed-codin> TYPE tName = OBJECT
14:14:47 <carefish> and for OpenGL it's fine to statically link SDL
14:14:53 <Mat2-speed-codin> <variable definations> ...
14:15:03 <Mat2-speed-codin> <methods>
14:15:06 <Mat2-speed-codin> END;
14:15:06 <carefish> same for DirectX, haven't tried dx10/11 though.
14:15:54 <Mat2-speed-codin> as written, some drivers have hardcoded a path
14:16:18 <carefish> oh :o haven't encountered that so far
14:16:53 <Mat2-speed-codin> that's one reason for the so called DLL hell
14:17:16 <carefish> yeah, i did have some issues w/ dll's a few months back
14:17:24 <carefish> but that was trying to mix static/shared libs
14:17:39 <carefish> back then i didn't know it wasn't possible, and got errors all over the place
14:18:00 <Mat2-speed-codin> specially old drivers which are needed for some obscure applications suffered from this behaviour
14:19:12 <carefish> thanks again for introducing me to pascal haha
14:19:22 <carefish> it's funny how much easier pascal is compared to C++
14:19:34 <Mat2-speed-codin> specially Delphi
14:20:21 <Mat2-speed-codin> which is used heavily in Germany for game development (in it's NET variant)
14:21:18 <carefish> i'll have a look after ld25
14:21:29 <carefish> does pascal have linked lists? or do i need to make my own?
14:21:57 <Mat2-speed-codin> I think there exist a good class in the library for that
14:22:25 <Mat2-speed-codin> but linked lists are very easy to implement in pascal
14:22:59 <Mat2-speed-codin> TYPE pList = ^tList
14:23:09 <Mat2-speed-codin> tList = RECORD
14:23:34 <Mat2-speed-codin> priour: ^tList;
14:23:45 <Mat2-speed-codin> sorry 
14:23:59 <Mat2-speed-codin> priour: pList;
14:24:04 <Mat2-speed-codin> next: pList;
14:24:21 <Mat2-speed-codin> cell: <whatever type>
14:24:24 <Mat2-speed-codin> END;
14:24:38 <tangentstorm> lib/xpl/ll.pas has a fully generic linked list
14:24:42 <carefish> do you have a sample file w/ some OOPness?
14:24:44 <tangentstorm> that works exactly that way
14:25:06 <carefish> great! i'll browse around for samples in the ld25 dir
14:25:10 <tangentstorm> whoops lib/xpl/code/ll.pas
14:26:11 <tangentstorm> some of what you'll find is code i wrote as a 16-year old. but i've been working slowly to bring it up to date
14:26:41 <carefish> haha, how old are you now? i'm 23 years old :P
14:26:51 <Mat2-speed-codin> tangentstorm: I have found the reason for the colour clash, it's just that I forgot to remap the colour values to the hardware dependent RGB values
14:27:20 <tangentstorm> <- 36
14:27:32 <carefish> haha
14:27:45 <carefish> so you've been programming for 20+ years?
14:28:10 <carefish> i started w/ C++ about 6-ish months ago, C# about a year ago
14:29:15 <Mat2-speed-codin> I've started as teenager with basic and assembly
14:30:14 <Mat2-speed-codin> but soon I've converted to forth :D
14:30:54 <Mat2-speed-codin> (and later lisp)
14:31:28 <carefish> and my school taught me the basics of AS3, which got me curious about other languages and when i popped a few veins due to debugging hell with as3 i switched to C#/C++ :P
14:32:26 <Mat2-speed-codin> there teach Actionscript ???
14:32:43 <carefish> the course used to be called Game Artist
14:33:14 <carefish> which split up into: Game Artist / Media Development, and now they're going to split up Media Development into Webdev and gamedev
14:33:32 <Mat2-speed-codin> that's the ruin of civiliation
14:34:46 <carefish> thanks tangentstorm, ll.pas shows a bit of OOPness i can use to learn from
14:35:03 <carefish> well, more like structure
14:35:05 <carefish> but it's okay
14:37:23 <carefish> what's the difference between procedure and function? they seemingly do the same
14:37:57 <Mat2-speed-codin> a function return a value (or pointer, or reference) o a caller
14:38:20 <carefish> ah yeah, now i see
14:38:25 <carefish> function list.is_empty : boolean;
14:38:30 <carefish> procedure list.drop;
14:38:35 <carefish> shouldve seen that one
14:38:55 <carefish> so procedure is a void ?
14:39:42 <Mat2-speed-codin> yes, because in C all routine definations are functions
14:40:50 <Mat2-speed-codin> the differentiation into procedures and functions is useful, because the compiler can optimizate the calling overhead away without problems for most procedures
14:41:10 <carefish> and to add ll.pas to uses, i would need to do what? USES ll; ?
14:41:50 <Mat2-speed-codin> is ll.pas a unit ?
14:41:54 <carefish> optimize btw ;)
14:42:02 <carefish> how can i know if it is?
14:42:06 <Mat2-speed-codin> thanks
14:42:10 <carefish> it says so at top of the file
14:42:28 <carefish> unit ll; { linked list support }
14:42:28 <carefish> interface uses xpc; 
14:43:06 <Mat2-speed-codin> ok, then you can compile it (fpc ll.pas) and afer that you have a automatical registered library which can be inluded with USES
14:43:28 <Mat2-speed-codin> so just write USES ..., ll;
14:43:51 <carefish> i don't need to place the resulting binary in my project dir?
14:43:57 <Mat2-speed-codin> no
14:44:22 <Mat2-speed-codin> ok, I don't know he behaviour in Windows
14:45:21 <Mat2-speed-codin> the compiler must only know the path to your binary somehow
14:45:51 <carefish> http://img.is1337.eu/99f01160.png
14:46:09 <carefish> i'll just place it in lazarus's folder 
14:46:39 <carefish> C:\FPC\lazarus\fpc\2.6.0\units\i386-win32
14:46:45 <carefish> there is where i'll put it
14:47:45 <Mat2-speed-codin> please paste line 20-26 of your ll.pas file
14:48:12 <carefish>       public
14:48:12 <carefish>         constructor create( lis : tlist );
14:48:12 <carefish>         function movenext : boolean;
14:48:12 <carefish>         property current : T read _value;
14:48:12 <carefish>     end;
14:48:12 <carefish>     listaction = procedure( var n : T );
14:48:12 <carefish>     predicate  = function( n : T ) : Boolean;
14:49:55 <Mat2-speed-codin> ok, paste the whole code for the object defination
14:50:50 <Mat2-speed-codin> the error seem to be these public statement
14:51:39 <Mat2-speed-codin> ah ok, you have a public constructor
14:51:48 <tangentstorm> carefish: the very first line in a pascal file is either "unit", "library" (for making dll's) or "program"
14:52:09 <carefish> it's not my code :) it's tangentstorm's code
14:52:51 <tangentstorm> ....
14:53:08 <tangentstorm> somehow it's conflicting with tlist from the classes unit, but that shouldn't be in scope
14:53:19 <tangentstorm> my bad.
14:53:24 <carefish> i could be doing something wrong too though
14:53:43 <tangentstorm> just search and replace tlist and change it to "ts" or something
14:53:58 <tangentstorm> hmmm...
14:54:43 <carefish> replaced all tlist w/ ts
14:54:46 <carefish> still same error
14:54:52 <tangentstorm> hang on
14:55:07 <carefish> and thanks for the crash course pascal :P
14:57:26 <tangentstorm> it's the illegal expression that's causing the problem. 
15:00:30 <tangentstorm> carefish: okay... so that ts is specializing list<t> but it's also /inside/ list<t>
15:00:41 <tangentstorm> turns out this is not supported by fpc 2.6.0
15:00:49 <tangentstorm> i'm using the fpc trunk
15:00:56 <tangentstorm> 2.7.1
15:01:16 <carefish> can i still use it then when it's compiled using 2.7.1?
15:01:22 <tangentstorm> so... you can either compile the trunk ( which is easy but annoying )
15:01:49 <tangentstorm> or you can use TList from the classes unit ( which is not generic, so you'll have to explicitly cast everything back from TObject )
15:02:23 <tangentstorm> if you like this pascal stuff: check out the compiler and just compile the latest version
15:02:35 <tangentstorm> ( you use 2.6.0 to compile 2.7.1 )
15:02:52 <tangentstorm> ( fpc is written in itself )
15:03:05 <Mat2-speed-codin> the same as retro :D
15:04:08 <carefish> oh, how does that work? 'written in itself'?
15:04:46 <tangentstorm> carefish: the fpc team uses svn , but there's an actively maintained git mirror here if you prefer: https://github.com/graemeg/freepascal
15:05:11 <tangentstorm> like i just said: each stable version of the compiler is used to compile the source for the next version :)
15:05:22 <tangentstorm> i suppose originally it was written in turbo pascal or delphi
15:05:43 <Mat2-speed-codin> tangentstorm: can you please pack a zip file with your win32 retro build (and all needed libraries) so I can test out SDL for these os ?
15:06:06 <carefish> nice :)
15:10:08 <tangentstorm> Mat2-speed-codin: sure thing... i'm afk for a bit but will do so as soon as i get back.
15:10:13 --- nick: tangentstorm -> tangentaway
15:13:09 <Mat2-speed-codin> back in a half hour
15:13:15 --- nick: Mat2-speed-codin -> Mat2-afk
15:20:53 <carefish> so it's perfectly possible to create dll's to be used by C++ code inpascal?
15:29:05 <carefish> (and vice versa?)
16:33:06 --- nick: tangentaway -> tangentstorm
16:33:21 <tangentstorm> carefish: yep
16:33:46 <tangentstorm> one of the first things i tried was packaging up a python module in pascal. worked like a charm.
16:40:28 <tangentstorm> carefish: did the graphical retro terminal build for you?
16:42:45 --- nick: Mat2-afk -> Mat2
16:43:27 <Mat2> I get some sleep
16:43:34 <Mat2> see you
16:43:44 --- quit: Mat2 (Quit: Verlassend)
16:46:55 <tangentstorm> seeya
16:47:11 <tangentstorm> wait.. mat2 : did you check your code in?
16:47:19 <tangentstorm> oh he's gone :/
16:49:28 <tangentstorm> and didn't check the code in...
16:50:53 <tangentstorm> well maybe the latest version is what's already in ng_ports.pas ... we'll work from there.
16:58:16 <tangentstorm> ------------------
16:58:42 <tangentstorm> hey guys... intothev01d's helping out with getting the sped-up js version of ngaro under test here
16:59:43 <tangentstorm> intothev01d: https://github.com/tangentstorm/ngaro-js/blob/master/src/ngaro.js#L663
16:59:50 <tangentstorm> this is the main routine that we need to expose
16:59:58 <tangentstorm> processOpcode()
17:00:05 <tangentstorm> well...
17:00:18 <tangentstorm> that's the main routine, but it only does one opcode
17:00:45 <tangentstorm> by default javascript shares a thread with the browser, unless we use these new html5 web worker things
17:01:08 <tangentstorm> without those, you have to break the processing down into little chunks
17:01:25 <tangentstorm> or else it'll lock up the whole browser :/
17:02:08 <intothev01d> so it also now needs to run those functions on a separate thread then huh
17:02:24 <tangentstorm> https://github.com/tangentstorm/ngaro-js/blob/master/src/ngaro.js#L924
17:02:34 <tangentstorm> rxProcessImage() is scheduled to run periodically
17:02:58 <tangentstorm> the WAITING state means that ngaro is waiting for you to do something (press a key, etc)
17:03:02 <tangentstorm> so there's nothing to process
17:03:32 <tangentstorm> the RUNNING means that the previous instance hasn't finished... i suppose in a single-thread environment that shouldn't ever actualy happen
17:04:10 <tangentstorm> but if either of those things are true, it exits, and otherwise it runs for however many cycles are defined in the "cycles" constant
17:04:42 <tangentstorm> so it's actually this rxProcessImage that we need to expose
17:05:24 <tangentstorm> [ we're talking about wrapping up ngaro.js as a module so we can "require" it from a command-line oriented retro.js file to be run from node ]
17:05:33 <intothev01d> so then from my understanding you would add like exports.rxProcessImage and then require it in the server.. I think
17:05:44 <tangentstorm> yep
17:05:54 <tangentstorm> but server's not a good word for whatever it is :)
17:06:21 <tangentstorm> the... repl? listener?
17:06:55 <tangentstorm> app? :) i don't know
17:07:28 <intothev01d> i just meant currently you run node server.js to get it going so that is where you would have to put the required modules and call them it seems
17:08:06 <tangentstorm> no, forget about server.js :)
17:08:14 <tangentstorm> that's *only* there to be a web server
17:08:15 <intothev01d> k. forgotten
17:08:59 <intothev01d> so you would run node app.js that would then require elements from ngaro.js and retro.js
17:09:19 <tangentstorm> well i was actually thinging retro.js would be the app
17:09:22 <intothev01d> and then you can pass in arguments via that
17:09:25 <intothev01d> oh ok
17:09:26 <tangentstorm> because we have retro.c, retro.py, etc
17:09:42 <intothev01d> so it just needs to access some elements of ngaro.js
17:09:45 <intothev01d> or all?
17:09:58 <tangentstorm> the idea is you should be able to ssh in to a machine that has node on it and run ./retro.js and it'll run and say "Retro 11.5  ok"
17:10:10 <tangentstorm> and then you'd be able to type just like anything else
17:10:14 <tangentstorm> like any of the others i mean
17:10:28 <intothev01d> ah ok
17:10:29 <intothev01d> i see
17:10:49 <tangentstorm> it needs the core vm, and then at the end, it needs to be able to do the dup f
17:10:55 <tangentstorm> to do the state dump
17:11:08 <tangentstorm> which requires access to   data, address, and image
17:11:31 <tangentstorm> although it's totally fine to put the dump routine inside the module and just expose that
17:11:55 <tangentstorm> intothev01d: do you know what a singleton is?
17:12:06 <intothev01d> eh sort of
17:12:07 <tangentstorm> like in object oriented programming terms
17:12:19 <tangentstorm> it just means that there's only one instance of the class ever
17:12:19 <intothev01d> yea only one instance can be created throughout basically right?
17:12:24 <tangentstorm> yeah exactly
17:12:47 <tangentstorm> so in languages that have modules, they perform some of the same functions as singletons
17:13:09 <tangentstorm> they encapsulate state and hide information just like an object
17:13:27 <intothev01d> oh ok
17:13:31 <tangentstorm> but there's only one of them... and no constructor so you can't make any more
17:14:17 <tangentstorm> so it would be good design to only expose three "methods" (functions) from the module for what we're doing:
17:15:04 <tangentstorm> setImage( img:array of int ); // though just making "image" public is fine too
17:15:28 <tangentstorm> rxRunImage()
17:15:49 <tangentstorm> and rxDump() : string
17:16:00 <tangentstorm> rxDumpState() maybe
17:16:35 <tangentstorm> so then you'd say   var ngaro = require("ngaro.js");  or whatever
17:16:42 <tangentstorm> and you'd have those three methods
17:17:12 <tangentstorm> so from the test runner / command line point of view that's all we need
17:18:03 <tangentstorm> well to actually run the prompt we'll need to make a new device like ngterm but for the console, but that's not important right now
17:18:33 <tangentstorm> the tests *only* look at the opcodes
17:18:41 <tangentstorm> and the effects on the stack
17:19:10 <intothev01d> nice
17:19:30 <tangentstorm> the tests for retro itself cover the ports, but i already know the ports work because i haven't changed them :D
17:19:36 <intothev01d> I'll try to get that running then
17:20:02 <tangentstorm> do you want me to take one side of the interface and you do the other?
17:20:37 <tangentstorm> ( just an offer. if not, i'll go work on something else here... )
17:23:38 <intothev01d> well i think i can work on the exports part of it and getting it to run in command line as a node app or whatever. if I get stuck I'll let you know
17:24:17 <intothev01d> but I'll start on it in a few. gtg but I'll be back 
17:52:28 <carefish> awesome! just finished watchin' a movie, back to readin 'bout pascal
18:00:07 <carefish> still looking at your linked list tangentstorm, so if i understand correctly, you first type out what it's supposed to do, and then later on you can fill them in, sort-of like C++ header/cpp?
18:01:09 <carefish> anything below implementation is 'actual' code?
18:01:30 <tangentstorm> carefish: yeah exactly... except it's usually in one file, with the INTERFACE section at the top, followed by the IMPLEMENTATION 
18:01:39 <carefish> yeah, i saw that
18:01:41 <carefish> that's neat
18:01:49 <tangentstorm> you can also have an INITIALIZATION if there's stuff to set up when the module first loads
18:02:06 <tangentstorm> and also FINALIZATION
18:02:06 <carefish> got one in the xpl folder?
18:02:12 <carefish> like, where you use that?
18:02:26 <tangentstorm> probably i would have called it BEGIN
18:02:43 <carefish> oh.. you can redefine keywords?
18:02:49 <tangentstorm> no, but before they added finalization it was just begin
18:02:56 <carefish> ooh
18:03:19 <carefish> and if you leave it empty?
18:03:22 <tangentstorm> i learned this language in 1992... and it's evolved considerably... a lot of my code is out of date
18:03:26 <intothev01d> seems that loadImage in ngaro.js doesn't take arguments? https://github.com/intothev01d/ngaro-js/blob/master/src/ngaro.js#L554
18:03:29 <carefish> like, in ll.pas i see initialization
18:03:29 <carefish> end.
18:03:46 <carefish> does that tell pascal to init it right away when used?
18:04:02 <tangentstorm> intothev01d: oh that's for loading from a browser's local cache
18:04:13 <intothev01d> alright, is there one to pass to?
18:04:29 <tangentstorm> nope.
18:04:32 <intothev01d> rxLoadCleanImage below it maybe?
18:04:36 <intothev01d> it takes a url ...
18:04:40 <tangentstorm> ...
18:04:56 <tangentstorm> yeah but it's going to want to do an http request...
18:05:11 <intothev01d> so i need to make one for process.argv that will take the arguments and load it
18:05:11 <tangentstorm> what i was picturing was really more like setImage(img)
18:05:32 <tangentstorm> carefish: it doesn't do anything. you have to actually put code in there to run something.
18:05:39 <intothev01d> brb, gonna grab a bite to eat
18:06:08 <carefish> ah, nice, do you happen to have a good tutorial website about pascal? so i needn't to bother asking stuff :P
18:06:46 <tangentstorm> intothev01d: well, probably it would just be better to have all the filesystem-related stuff in retro.js... and then we should move the web-specific stuff out to its own file too
18:07:03 <tangentstorm> there is a pdf that seems nice... i've not read all of it
18:07:04 <tangentstorm> hang on
18:07:15 <tangentstorm> but also i don't mind you asking questions :)
18:07:29 <carefish> i really wish i had found out about pascal earlier
18:07:38 <carefish> seems like a fun language
18:07:42 <carefish> like C#
18:07:48 <carefish> well, not at all like it obviously
18:07:57 <tangentstorm> http://freepascalanswers.wordpress.com/2012/09/02/freepascallazarus-book-startprog/
18:08:18 <tangentstorm> it's got a lot of nice things in it... like the SET type.
18:08:19 <carefish> but i have lots of fun with C# too, nice and easy (also having fun with C++, but that tends to give a headache or two once in a while)
18:08:32 <tangentstorm> also, btw... i read the scrollback.. for your linked list... you don't actually need a linked list.
18:08:43 <tangentstorm> you can use a variable sized array for  what you want to do
18:09:00 <tangentstorm> just say var x : array of int;  and then you can use setlength( x, 100 ) or whatever you want.
18:09:06 <carefish> oh, how do i do that?
18:09:10 <carefish> oh wow
18:09:22 <carefish> nice :)
18:09:39 <carefish> can i have the first element be of an int/char/not an image?
18:10:08 <tangentstorm> everything in an arry has to be the same type
18:10:15 <tangentstorm> but you can make a record with an array inside of it
18:10:16 <carefish> wanted to use the first index (or the last, doesn't matter that much) to determine what view the player is currently seeing
18:10:59 <tangentstorm> type direction = ( NN, NE, NW, WW, SW, SS, SE, EE, NE );
18:11:18 <carefish> like: direction[0][0] through [0][whatever] with 0 being N/1 being NW/etc
18:11:41 <tangentstorm> type whatyouwant = array[ direction ] of whatever;
18:12:00 <tangentstorm> then you can say
18:12:04 <tangentstorm> array[ NW ] = ....
18:12:08 <tangentstorm> well
18:12:17 <tangentstorm> stuff : whatyouwant;
18:12:21 <carefish> = anotherarrayholdingthesprites
18:12:23 <tangentstorm> stuff[ NW ] = ...
18:12:44 <tangentstorm> yeah so you want array[ direction ] of array of sprites;
18:12:50 <carefish> yeah
18:12:59 <tangentstorm> sprite;
18:13:06 <carefish> haha, yeah
18:13:36 <tangentstorm> you'd call the type sprite (or "TSprite" if you like the borland convention, which dominates the pascal world now)
18:13:49 <tangentstorm> i learned on the version of pascal right before they introduced it and i don't like it :)
18:13:55 <carefish> would it be wise to learn such a convention when starting out?
18:14:02 <tangentstorm> btw, if you didn't notice already, pascal is case-insensitive
18:14:11 <carefish> like, i like the idea behind the hungarian notation 
18:14:18 <carefish> but the practical use nowadays is kinda moot
18:14:45 <carefish> i've got intellisense telling me what the var is, and if it's unclear due to weird casting stuff i'm probably doing it wrong
18:14:53 <tangentstorm> well, *because* pascal is case-insensitive, you sometimes have to use prefixes or whatever to distinguish two things when you'd use capitalization in java or c#
18:15:02 <carefish> ah :)
18:15:13 <carefish> so, for example
18:15:54 <carefish> type direction  = ( NN, NE, NW, WW, SW, SS, SE, EE, NE ); type sprite = ( direction array of sprite );
18:16:26 <tangentstorm> the ( ) indicates an enumerated type, so you use it on the direction type.
18:16:29 <carefish> ehh, array[direction] of sprite;
18:16:33 <tangentstorm> yes
18:16:37 <tangentstorm> no
18:16:43 <tangentstorm> you can't say sprite = array of sprite
18:16:49 <tangentstorm> you have to name it something else
18:16:56 <carefish> ooh, so i'd need to define sprite first?
18:17:05 <tangentstorm> yes
18:17:30 <carefish> type direction  = ( NN, NE, NW, WW, SW, SS, SE, EE, NE ); type spriteDirections = ( array[direction] of sprite ); type sprite = ( SDLBITMAPwhatever );
18:18:31 <carefish> probably have sprite hold a pos vec and vel vec too, along with possible paralax code
18:18:33 <tangentstorm> how about "scene" for spriteDirections?
18:19:08 <tangentstorm> and you ONLY use the parentheses on the direction one
18:19:09 <carefish> was going to have it say: player.lookat(spritedirections[NN]);
18:19:35 <tangentstorm> well, right now you're just defining a type
18:19:48 <tangentstorm> in your example code there it would be a variable that's an instance of that type
18:20:16 <tangentstorm> you can use NN directly because that's a number just like 2 or $AF or whatever
18:20:18 <carefish> oh.. so it's still normal array iteration?
18:20:44 <carefish> spriteDirections[0] would hold all kinds of direction?
18:21:17 <tangentstorm> if you say   VAR  dir : spriteDirections ;  
18:21:28 <tangentstorm> then you'll have an array with 8 elements
18:21:53 <tangentstorm> but instead of dir[ 0 ] you would say dir[ SW ]
18:22:08 <carefish> or NN or NE?
18:22:11 <tangentstorm>  0 would give you a type error
18:22:12 <tangentstorm> yes
18:22:14 <carefish> like, wouldn't matter?
18:22:21 <carefish> as long as it's of type direction ?
18:22:29 <tangentstorm> the full syntax would be 
18:22:39 <tangentstorm> array[ NN .. NE ] of xxx
18:22:50 <carefish> aaah
18:22:51 <carefish> now i see
18:22:53 <tangentstorm> if you wanted 256 you could say
18:22:56 <carefish> that makes sense
18:22:57 <tangentstorm> array[ byte ] of xxx
18:23:02 <tangentstorm> or array[ char ] of xxx
18:23:20 <tangentstorm> array[ 'a' ...'z' ]
18:23:21 <tangentstorm> etc
18:23:32 <carefish> that's really cool
18:23:36 <tangentstorm> has to be a simple type though with fixed boundaries. can't be a string or anything
18:23:46 <carefish> 'a'..'z' to be syntactically correct?
18:24:08 <tangentstorm> you can make types like that, and you can override the  [ ] operator for indexing though, so you can get python-like syntax
18:24:24 <tangentstorm> yes, i added one too many '.' :)
18:24:33 <carefish> :D i'd love to have python's way of dealing w/ strings in pascal
18:24:49 <carefish> varname[6:]
18:24:57 <tangentstorm> you can do that
18:25:04 <carefish> cuts off first 6 (or last, been too long since i've done python)
18:25:07 <tangentstorm> varname[0..5]
18:25:09 <carefish> i loved that
18:25:17 <tangentstorm> not quite the same, but you can get the same effect
18:25:20 <carefish> awesome! :D
18:25:28 <carefish> i'll have a blast w/ pascal
18:25:34 <tangentstorm> yeah, it's cool :)
18:25:59 <tangentstorm> also when you say   ch : char   ..... if ch in 'a'..'z'  then whatever...
18:26:10 <tangentstorm> 'a'..'z' will be treated as a set
18:26:25 <tangentstorm> so it doesn't do    ch >= 'a'  AND ch <='z'
18:27:10 <tangentstorm> it does:   1 SHL ord(ch)   AND    some long constant where the a'th through z'th bits are set to 1
18:27:31 <tangentstorm> meaning it turns it into a bitmask
18:28:02 <tangentstorm> you can say   d : direction;   IF d IN [ N, S, E, W ] THEN...
18:28:23 <tangentstorm> that's easier to see because it'll just use a single 32-bit cell in the cpu.
18:28:44 <tangentstorm> whereas with the char type you need 256 characters, and it may not actually be as efficient
18:28:46 <tangentstorm> so like
18:28:51 <carefish> so to store pSDL_SURFACES in the direction stuff: type sprite = ( pSDL_SURFACE ) type direction  = ( NN, NE, NW, WW, SW, SS, SE, EE, NE ) type spriteDirections = ( array[NN..NE] of sprite );
18:29:23 <carefish> 1 SHL ord(ch) < wut?
18:29:25 <tangentstorm> type sprite = PSDL_SURFACE; direction = ( ... ); spriteDIrections = array[direction] of sprite;
18:29:46 <carefish> ah, ( ) denote enums when used with type?
18:29:48 <tangentstorm> yes
18:29:53 <tangentstorm> otherwise it's a syntax error
18:30:01 <tangentstorm> and... you don't have to keep saying type over and over
18:30:17 <carefish> and to add more to sprite?
18:30:45 <tangentstorm> type sprite = record x, y : int32; img : psdl_surface end;
18:30:45 <carefish> type sprite = PSDL_SURFACE, vector, vector ?
18:30:49 <carefish> ah
18:30:51 <carefish> awesome
18:30:52 <tangentstorm> or class
18:31:02 <carefish> what's the difference between record and class?
18:31:07 <carefish> record is like struct?
18:31:22 <tangentstorm> mostly yeah
18:31:46 <tangentstorm> struct and also union, but with cleaner rules
18:32:18 <tangentstorm> and you can have a file of a record type and just read and write them directly, which is handy... so they're fixed width
18:32:31 <carefish> i barely know what a union is, also don't know alot about struct.. tend to use 'class' in C#/C++
18:33:00 <tangentstorm> just use class everywhere, and you'll be fine :)
18:33:01 <carefish> ahh, that's why you sometimes use structs
18:33:13 <carefish> no need for proper instantiation?
18:33:18 <tangentstorm> right
18:33:19 <carefish> proper being actually 'new'ing it
18:33:24 <tangentstorm> right
18:33:31 <carefish> awesome :D
18:33:44 <carefish> so a vector (math vector) should be a struct?
18:33:52 <tangentstorm> no it should be an array
18:34:20 <carefish> oh, w/ 3 (or 4 if you're into vector4) elements?
18:34:37 <tangentstorm> yeah
18:34:38 <carefish> and then use the same trick as with the directions
18:35:08 <carefish> so you could have, for funsies, a vector[x][x][x]?
18:35:15 <carefish> ehh
18:35:24 <tangentstorm> array of array of array of whatever.
18:35:34 <carefish> well, that wasn't what i meant
18:35:45 <carefish> vector[1..4]
18:36:07 <tangentstorm> array[ 1..5 ] of array[ -5..10 ] of array[ 12 .. 15 ] of char;
18:36:12 <tangentstorm> no
18:36:37 <tangentstorm> only string and array can have dimensions, as far as i know
18:37:02 <carefish> i'm really liking what i'm reading
18:37:32 <carefish> gonna read a bit more about pascal, specifically look at how to write the syntax, and i'll get going w/ the sprites
18:37:38 <tangentstorm> yeah, it's pretty clean and mathematical, and lets you think about your algorithm rather than the computer
18:37:49 <tangentstorm> but you can actually also  add asm blocks with pure assembly language
18:38:20 <tangentstorm> if you don't mind, i'm going to implement some simple routines to show the surfaces
18:38:27 <carefish> go ahead
18:38:33 <tangentstorm> but i well leave the directional stuff to you
18:38:40 <tangentstorm> i just want to get something on the screen :)
18:38:49 <tangentstorm> cool
18:38:55 <carefish> i'll just make it when i think i understand the syntax well enough
18:39:18 <tangentstorm> ok :)
18:39:34 <tangentstorm> feel free to post work in progress to pastebin or gist or something if you want some help
18:39:41 <carefish> sure :)
18:40:52 --- join: intothev01d_m (~intothev0@unaffiliated/intothev01d) joined #retro
18:42:21 <carefish> oh wow, if/else is so easy to read :D
18:44:00 <carefish> a float in C++ is a Single in pascal?
18:44:24 <tangentstorm> ...
18:44:42 <tangentstorm> erm
18:44:48 <tangentstorm> maybe? :)
18:44:55 <tangentstorm> so.. retro has no floating point numbers
18:45:15 <tangentstorm> forth in general uses fixedpoint.
18:46:06 <tangentstorm> back in my day we used the type "real" which *should* be a floating point number at your processor's native size just like integer should be and cardinal should be
18:46:21 <tangentstorm> but freepascal is delphi-compatible, and delphi is tp compatible and tp was 16 bits
18:46:29 <tangentstorm> so...
18:46:33 <tangentstorm> that's all i know :) 
18:47:06 <tangentstorm> ususally if you use the word "freepascal whatever" in google, you can find the right page in the docs or the wiki though
19:01:08 <carefish> oh so that's what the T is for, so you know it's not a variable
19:01:36 <carefish> so there isn't an IDE that tells you what it is if your cursor hovers over it? (text cursor or mouse pointer)
19:04:43 <tangentstorm> carefish: lazarus has a bunch of code analysis tools but the user interface doesn't really expose them like that :/
19:08:19 <carefish> aww, that's too bad :(
19:12:18 <tangentstorm> agreed. i'm kind of working up to making an IDE that does that kind of stuff, probably as a semi-commercial product
19:12:34 <tangentstorm> lazarus is cool but i make games and web pages, not GUI apps.
19:13:24 <tangentstorm> also i want something that's more aimed at personal computing... things you would do as a hobbyist or in your own small business or something
19:13:35 <tangentstorm> or in a classroom
19:14:21 <tangentstorm> that's what the compiler is all about
19:14:48 <tangentstorm> i'm making a language that cleans up some of the junk that's accumulated in delphi and borrows some nice stuff from python and haskell and prolog and smalltalk
19:15:09 <tangentstorm> and of course retro :)
19:16:06 <tangentstorm> and between the portability of retro/ngaro and free pascal itself, it'll run pretty close to everywhere. :)
19:30:45 <intothev01d_m> hey tangentstorm, so I have the node module thing working with a simple hello world. I'm going to push it so you can see what I'm thinking about for the rest of the functions
19:30:55 <tangentstorm> ok
19:36:24 <intothev01d_m> https://github.com/intothev01d/ngaro-js
19:36:29 <intothev01d_m> that's my fork
19:36:58 <intothev01d_m> anyway, i just put a function at the top and then call it in retro.js once it's been set up as a module. I think I should be able to do that with any of the other functions as well
19:37:30 <intothev01d_m> that way you can just run "node retro.js" like you were talking about and have it do whatever
19:37:56 <tangentstorm> that works for me
19:38:17 <tangentstorm> but for the functions that already exist, just put exports.rxWhatever = rxWhatever
19:38:23 <intothev01d_m> yea
19:38:32 <tangentstorm> wait...
19:38:35 <tangentstorm> does that compile?
19:38:45 <intothev01d_m> ?
19:38:46 <tangentstorm> exports.hello = function.hello() { } 
19:38:47 <tangentstorm> +
19:38:49 <tangentstorm> ?
19:39:00 <tangentstorm> er function hello() i mean
19:39:02 <intothev01d_m> it worked
19:39:06 <tangentstorm> !!
19:39:36 <intothev01d_m> it's not function.hello
19:39:43 <intothev01d_m> it's function hello()
19:39:45 <tangentstorm> yeah
19:39:55 <tangentstorm> i didn't know you could use a named function that way
19:39:57 <tangentstorm> cool
19:40:06 <intothev01d_m> yea i didn't either :D
19:40:09 <intothev01d_m> I just tried it
19:40:09 <tangentstorm> in that case, as long as it works in the browser, go for it
19:40:22 --- quit: carefish (Ping timeout: 252 seconds)
19:40:34 <intothev01d_m> well i don't know about browser, all that does is make it print hello in the console 
19:40:42 <intothev01d_m> like you said about ssh and all and not using the server
19:40:47 <tangentstorm> right
19:40:49 <tangentstorm> hmm
19:41:37 <tangentstorm> well... no big deal if it doesn't work.
19:41:47 <intothev01d_m> in the browser?
19:41:50 <intothev01d_m> or in general
19:41:51 <intothev01d_m> lol
19:41:54 <tangentstorm> we can always just break it into 2 statements if the browsers complain i mean
19:42:16 <tangentstorm> function hello(){ } ; exports.hello = hello;
19:42:18 <tangentstorm> hmm
19:42:40 <intothev01d_m> that might be the correct way there, put them at the bottom like that?
19:43:00 <intothev01d_m> I've seem them do it inline as well with the function, just not with a named function
19:43:09 <tangentstorm> yeah
19:43:11 <intothev01d_m> so like exports.hello = function() {}
19:43:16 <tangentstorm> right
19:44:06 <tangentstorm> well, how about put it at the bottom for now, since we know that works, and then if/when we find out the way die hard javascripters do it, we'll do what they do :)
19:44:35 <intothev01d_m> yea i think that might be the best way anyway
19:45:01 <tangentstorm> also in the browser, there won't be an "exports"
19:45:06 <tangentstorm> so it'll raise an error
19:45:17 <tangentstorm> so you need a test like on line 6
19:45:55 <tangentstorm> you could say if (typeof exports == "undefined") { exports = {}; } or something
19:46:16 <tangentstorm> new Object(); if you prefer... pretty sure that's the same thing
19:46:26 <tangentstorm> exports = new Object(); i mean
19:46:43 <tangentstorm> then it would just be silently ignored
19:46:46 <intothev01d_m> so at the top just do exports = new Object(); so it won't throw errors
19:47:03 <tangentstorm> yeah but only if it's not already defined
19:47:35 <tangentstorm> either that or we make the browser side use modules too :)
19:48:07 <tangentstorm> but that's scope creep, so let's just ignore it for now and use the dummy object :)
19:48:30 <intothev01d_m> sounds good
19:49:16 <tangentstorm> thanks for doing this, btw :)
19:50:25 <intothev01d_m> np. i don't know if I'm doing it right or can get it working but I'll try
19:51:33 <intothev01d_m> so i need to create a setImage, then expose the existing ProcessImage, then make a imageDump right?
19:51:44 <tangentstorm> yep
19:52:44 <tangentstorm> it might be easier just to create those two functions to do nothing or set the image to [] for starters
19:53:01 <tangentstorm> then you can focus just on retro.js and not worry about the ngaro internals yet
19:53:29 <tangentstorm> hopefully you'll get 30 failing tests
19:55:08 <intothev01d_m> where would they go in the current ngaro.js code?
19:55:12 <intothev01d_m> line #'s
19:55:24 <intothev01d_m> I'll just put them in as placeholders like you said
19:57:51 <tangentstorm> intothev01d_m: doesn't matter to me
19:57:57 <intothev01d_m> k
19:58:26 <tangentstorm> there is some amount of logic to the layout though... maybe with the other loaders? 
19:58:49 <tangentstorm> i say it doesn't really matter because i probably *will* break this into modules later and so it'll all get re-arranged :)
20:01:48 <intothev01d_m> so I'm looking at retro.py. Am I porting all of that over to retro.js?
20:04:04 <intothev01d_m> sorry, still a bit confused as to what the role of retro.js is
20:10:15 <tangentstorm> ok :)
20:10:23 <tangentstorm> one sec
20:11:51 <tangentstorm> ok so if you look under http://bazaar.launchpad.net/~crc-x/retro-language/retro/files/head:/vm/
20:12:06 <tangentstorm> ( or in github or whatever .. .they're all mostly in sync )
20:12:18 <tangentstorm> you'll see there are a BUNCH of implementations of retro/ngaro
20:12:25 <tangentstorm> they all do pretty much the same thing
20:13:03 <tangentstorm> under complete/ you can see it implemented in go, lisp, several versions in c, c#, f#, php, py, ruby, retro itself, and scheme
20:13:29 <tangentstorm> under embedded, there are implementations for microcontrollers like arduino or rasberry pi or whatever
20:13:41 <tangentstorm> the original js one lives under web
20:13:48 <intothev01d_m> gotcha
20:14:06 <tangentstorm> experimental has like a bunch of variations on the complete ones... sockets, sdl, etc
20:14:24 <tangentstorm> partial has whatver fs is...
20:14:42 <tangentstorm> looks like some kind of forth dialectß
20:14:58 <tangentstorm> java, lua, perl, scheme
20:15:05 <tangentstorm> i guess the .s file i called scheme was actually assembly
20:15:24 <tangentstorm> anyway.... who knows if these things all actually work the same way?
20:15:32 <tangentstorm> they're all made by different people
20:15:40 <intothev01d_m> lol
20:15:52 <tangentstorm> and like... we figured out that one of the opcodes ( WAIT ) is redundant
20:16:00 <tangentstorm> at least... it appears to be
20:16:08 <tangentstorm> IN and OUT are sufficient for what it does
20:16:25 <tangentstorm> so it would be nice to actually drop it from all the machines, provided it doesn't break them
20:16:50 <tangentstorm> you can't make sweeping changes like that in a project without test coverage
20:17:06 <tangentstorm> now there's a test suite inside retro itself
20:17:16 <tangentstorm> but if your vm's messed up you can't trust retro :)
20:17:41 <tangentstorm> so...
20:18:04 <tangentstorm> to augment the tests in retro, i wrote that python file which tests the vm itself
20:18:22 <tangentstorm> so... the role of retro.js is to do what all the other retro.xxx does, but in javascript
20:19:02 <tangentstorm> which has the nice benefit of porting retro to node and rhino and all the various web browsers, and every app that runs javascript everywhere :)
20:19:11 <intothev01d_m> yea
20:19:20 <tangentstorm> so that's the point of ngaro.js
20:19:54 <intothev01d_m> and so retro just handles the interface between the user and the vm?
20:20:06 <tangentstorm> i'm almost positive i broke something in it while i was refactoring
20:20:07 <intothev01d_m> plust it's higher level language too right?
20:20:12 <tangentstorm> yeah
20:20:46 <tangentstorm> retro is both the repl (the ok prompt) and the set of words that have been predefined
20:21:25 <tangentstorm> i really don't have an easy way to explain what forth is... you just have to play with it and you'll see it
20:21:30 <tangentstorm> retro is a forth :)
20:22:13 <tangentstorm> a forth is a set of words and a loop that evaluates words as you enter them, and two stacks
20:22:27 <tangentstorm> that's pretty much it
20:23:13 <intothev01d_m> gotcha
20:23:13 <tangentstorm> but... it's also enough that you can make pretty much any language you want out of it
20:23:24 <tangentstorm> it's sort of like lisp without all the syntax :)
20:23:42 <intothev01d_m> interesting
20:23:44 <tangentstorm> and the order is different
20:24:14 <tangentstorm> so like if you don't like ":" you can say  ": def : ;"
20:24:16 <tangentstorm> hmm
20:24:29 <tangentstorm> you may need a quoting operator in there because ":" is an immediate word
20:24:34 <tangentstorm> i'd have to experiment
20:24:44 <tangentstorm> but in principle you can rename that word
20:24:52 <tangentstorm> or make an alias or whatever
20:25:26 <tangentstorm> immediate words write stuff to ram that become part of the dictionary
20:25:37 <tangentstorm> there's no one compiler or interpreter in forth
20:25:42 <tangentstorm> but any little word can be a compiler
20:26:06 <tangentstorm> ":" is a compiler that waits for you to type the next word and then makes a new dictionary header with that name and leaves the compiler on
20:26:25 <tangentstorm> by "the compiler" i really just mean a switch in the listener
20:26:41 <tangentstorm> when it's on, what you type gets compiled to instructions immediately
20:26:51 <tangentstorm> when it's off, what you type gets executed immediately
20:27:30 <tangentstorm> but in both cases it's just looking up each word in the dictionary and then either making a call to that address or writing that address as the next instruction
20:27:50 <tangentstorm> ( in ngaro, if you enter a number above 30, it's treated as a call to a function at that address)
20:28:16 <tangentstorm> so you really have as many instructions as you have addresses
20:28:39 <intothev01d_m> nice
20:28:55 <tangentstorm> and if you execute opcode 555, the instruction pointer gets pushed onto the address stack and then set to 555
20:29:50 <tangentstorm> then when you execute the RETURN instruction (written ; or ;; in forth (the shorter one also turns off the compiler)) then when that executes, it pops the old address off the address stack
20:30:20 <tangentstorm> most languages use one stack where everything is mixed together, but in forth they're separate and it makes things very simple and direct
20:31:12 <tangentstorm> you can picture the dictionary as if it were the following python structure:
20:32:04 <tangentstorm> [ ("word", "docs for word", SomeWordClass, [ instructions ]) .... ]
20:32:04 <intothev01d_m> is that the idea of the seperate arrays then? for data and instructions
20:32:32 <tangentstorm> well i should clarify that... the "image" contains data too.
20:32:58 <tangentstorm> because you can define strings and constants and whatever you want
20:33:05 <tangentstorm> and that all lives in the heap
20:33:22 <tangentstorm> historically "heap" was a kind of tree structure but it has come to mean "dynamic memory"
20:34:02 <tangentstorm> meaning you can just allocate some ram for whatever you want and then it's "in the heap"
20:34:08 <tangentstorm> and thats' where the dictionary lives
20:34:30 <tangentstorm> which is what's in the retroImage file too
20:34:40 <tangentstorm> in fact that's all the image is
20:35:10 <tangentstorm> the first two cells hold 0. a JUMP instruction 2. the location of the  main routine
20:35:17 <tangentstorm> er that would be cell 1
20:35:25 <tangentstorm> then 2 is a pointer to the last defined word
20:36:07 <tangentstorm> so you can look at that pointer and from there it's a linked list that you can walk to see all the words that have been publically defined
20:36:13 <intothev01d_m> so what the heck does divmod actually do in python?
20:36:21 <tangentstorm> in fact that's what happens if you type "words" in retro
20:36:22 <intothev01d_m> just divide numbers then mod first by second?
20:36:35 <tangentstorm> forth has no floating point numbers
20:36:51 <tangentstorm> so for division, it returns two results
20:37:18 <tangentstorm> the integer division sort of like python does (but i'm not sure if python rounds to 0 or negative infinity)
20:37:22 <tangentstorm> retro rounds toward 0
20:37:30 <tangentstorm> and then mod
20:37:38 <tangentstorm> which is the remainder
20:37:57 <tangentstorm> ( i think you were around for a long talk about modulo on lpmc a while back, right? )
20:38:09 <tangentstorm> so that's what it's doing in the python version
20:38:22 <tangentstorm> and all the versions, hopefully :)
20:38:25 <intothev01d_m> it's just moding a with b?
20:38:35 <tangentstorm> let me look at the actual code
20:38:39 <intothev01d_m> ok
20:38:46 <intothev01d_m> I'm looking at retro.py
20:38:57 <intothev01d_m> rxDivMod function
20:39:37 <tangentstorm> help(divmod)
20:39:49 <tangentstorm> i never knew that was there :)
20:39:59 <tangentstorm> i always just used / and % :)
20:40:54 <intothev01d_m> so divide a / b and then what? mod that value by b or?
20:41:14 <tangentstorm> divmod(x, y) -> (div, mod)
20:41:22 <tangentstorm>  Return the tuple ((x-x%y)/y, x%y).  Invariant: div*y + mod == x.
20:41:55 <tangentstorm> the first value there (x-x%y)/y is going to be an integer
20:42:04 <tangentstorm> because x%y is the remainder
20:42:21 <intothev01d_m> oh i see what it's doing now
20:42:22 <tangentstorm> so if you have  26%5 the remainder is 1
20:42:42 <tangentstorm> and 26-1/5 is going to be an integer because the remainder is removed
20:42:44 <tangentstorm> cool
20:43:15 <tangentstorm> -26/5 is -6 in python
20:43:24 <tangentstorm> but +26/5 is 5
20:43:34 <tangentstorm> so python rounds toward negative infinity
20:45:01 <tangentstorm> in retro, -26 5 /mod .s   prints -1 -5
20:45:34 <tangentstorm> crc told me this is called "symmetric division" and what python does is assymmetric division
20:45:45 <tangentstorm> asymmetric
20:46:28 <tangentstorm> so that's what the extra stuff in rxDivMod is doing
20:49:39 <tangentstorm> intothev01d_m: http://bazaar.launchpad.net/~tangentstorm/retro-language/docwork/revision/399.2.3#vm/complete/retro.py
20:49:57 <tangentstorm> these are the changes i made to retro.py to implement what you're now doing for the js version
20:50:48 <tangentstorm> it's mixed in with some other stuff that's not relevant though ( like the termio stuff, which is for direct keyboard access )
20:53:10 <intothev01d_m> in that divmod function in python it returns 2 values there, how would I do that in js
20:53:32 <tangentstorm> return an array
20:53:50 <tangentstorm> but all that logic is already implemented
20:54:22 <tangentstorm> https://github.com/intothev01d/ngaro-js/blob/master/src/ngaro.js#L802
20:54:26 <tangentstorm> divmod in js
20:55:54 <tangentstorm> but since this is ngaro, instead of returning things, it adds the values to the data stack
20:57:24 <tangentstorm> i suspect the pascal version probably mis-implements this opcode, btw. the -26/5 example would be a good thing to add to the test suite. :)
21:00:29 <intothev01d_m> yea
21:00:31 <intothev01d_m> it would
21:01:39 <tangentstorm> okay... well, i've been at this all day, and need some sleep... hopefully tomorrow i'll actually write some code
21:02:35 <intothev01d_m> alright
21:02:36 <intothev01d_m> cya
21:03:15 <tangentstorm> *** Mat2 / carefish *** if you're reading the logfile : i've been looking through the sdl stuff but haven't added any code... whoever shows up first tomorrow feel free to take the lead and run with the sdl stuff.
21:03:51 <tangentstorm> i guess i have their email addresses from the doc... okay, well... will do that next time :)
21:04:04 <tangentstorm> see you guys
21:04:10 --- nick: tangentstorm -> tangentsleep
21:38:31 --- quit: intothev01d_m (Quit: intothev01d_m)
21:43:29 --- quit: intothev01d (Quit: intothev01d)
22:44:51 --- quit: kumul (Quit: WeeChat 0.3.9.2)
23:59:59 --- log: ended retro/12.12.15