00:00:00 --- log: started retro/12.10.07 04:35:47 --- quit: ChanServ (*.net *.split) 04:45:14 --- join: ChanServ (ChanServ@services.) joined #retro 04:45:14 --- mode: hubbard.freenode.net set +o ChanServ 05:02:46 you guys are gonna like this ... :) 06:06:41 so so close, but my brain is melting and i'm about to fall asleep at the keyboard 06:06:48 https://github.com/sabren/b4/tree/master/b4a 06:08:12 i wrote a very simple assembler in perl with a simple macro system based on capturing the write pointer like what retro and forth do 06:08:46 then i hand coded a ngaro image that uses the one character dispatch idea 06:09:51 and left 256 slots so you can fill them in with whatever you want 06:10:23 the idea is that your retro is a blank slate but it has the ability to send one instruction at a time 06:11:34 so it's forth but instead of keeping the dictionary in the image, you keep it in whatever's programming the image 06:12:10 anyway. it's neat. probably doesn't work yet, but it's still neat. 06:12:22 have a good day. i'll be asleep. :) 08:03:17 --- join: kumul (~kumul@cvx-ppp-66-50-130-239.coqui.net) joined #retro 09:31:37 --- quit: erider (Changing host) 09:31:37 --- join: erider (~chatzilla@unaffiliated/erider) joined #retro 09:32:16 hi all 09:59:32 hi erider 09:59:39 hi docl 10:35:56 tangentstorm: what are you using that perl script for? 12:32:09 --- join: ncv (~quassel@89.123.22.68) joined #retro 12:32:09 --- quit: ncv (Changing host) 12:32:09 --- join: ncv (~quassel@unaffiliated/neceve) joined #retro 13:30:08 --- quit: ncv (Remote host closed the connection) 13:38:49 --- join: ncv (~quassel@unaffiliated/neceve) joined #retro 14:36:28 --- quit: ncv (Ping timeout: 264 seconds) 14:36:37 hey all 14:38:05 erider: it's an assembler for any kind of 32-bit binary file. 14:39:36 erider: if you feed ngaro.b4a into it, it should produce a very small image file 14:42:26 hi tangentstorm 14:42:35 heya :) 14:44:44 * docl is looking at the perl code now 14:49:34 * oPless reads the buffer. 14:49:49 oooh tangentstorm …. fpga on a usb stick :D 14:50:50 where's this greenarray jobby sold in singles? 14:51:41 oh, there . 14:51:45 eerk 1.8v 14:54:33 * oPless has always wanted to fiddle with fpgas and build his own cpu from scratch 14:54:42 the greenarray chips are qfn, so soldering by hand may be problematic :( 14:54:54 oh there's breakoutboards :) 14:55:56 yeah.. the schmartboard guys have a video of how to do the soldering... from what i understand, their little boards have most of the solder already in place or something. 14:56:34 tangentstorm: that would help a lot 14:56:34 that fpga would be fantastic to build for though. 14:56:59 but heh. too many projects on at the moment. 14:57:31 oPless: i've never used that usb fgpa one 14:57:35 * crc will take a closer look at greenarrays in a few months 14:57:42 i think docl found that... i have this: http://papilio.cc/ 14:58:01 but i don't know how to use it yet. it's still in the box :) 14:58:50 crc: you can also run the simulator, which comes with colorforth... assuming you can cope with colorforth :) 15:00:21 https://plus.google.com/115310700190572540735/posts <- animation of the simulator from colorforth 15:00:55 the way those machines work... you only talk to one node in the array directly. 15:01:00 tangentstorm: I think I see what your code is doing. neat! 15:01:13 thanks :) 15:01:29 each node in a greenarray can talk to its neighbors, or if it's on the edge, you can wire some device into it 15:02:10 so what you have to do is have your pc talk to one of the nodes, and basically write a little "virus" that carries any code you want from neighbor to neighbor 15:02:51 the chips have a mode where you can basically set the instruction pointer to a port instead of to ram 15:03:49 so.. what i'm trying to do with https://github.com/sabren/b4/blob/master/b4a/ngaro.b4a is provide that same ability for ngaro 15:05:23 what is "b4" ? 15:05:38 tangentstorm: I could cope with colorforth if the keyboard interface is qwerty or dvorak. I never was able to get the keyboard interface in the original colorforth. 15:06:43 oPless: it was going to be a forth dialect, until i found retro... now maybe it's a bootloader/bootstrapper for retroforth? :) 15:07:42 crc: yeah. the current version from greenarrays has a qwerty mode enabled by default. it's still completely crazy though :) 15:08:08 yeah, I was going to make a tiny VM before I ran across this (actually I've half written three or four, and I get distracted a lot so they rarely get finished) 15:08:33 hence my dirty C# port of the java 15:08:38 version 15:08:47 <- never once got distracted by anything ;) 15:08:59 haha 15:09:18 damn you tangentstorm, I want to fiddle about with fpgas now 15:10:12 the other use cases for this thing are a) emulate other virtual machines dynamically in retro (esp. python and some public domain pascal machines) 15:10:36 b) provide an implementation of meta.rx for bootstrapping retro without retro 15:12:24 i was thinking c) a UTF-8 decoder too but i'm kind of thinking it may be simple enough to do that in plain assembly language 15:12:39 ngaro assembly i mean 15:17:06 ... the reason it doesn't work yet is that i have no way to make a forward reference in the assembler :/ 15:23:33 i think that if the very first jump calls a routine that queries the image size and jumps to the end of the image - 2, then when you're done with the code you can leave an area that's just full of instructions to go back and patch all the forward references 15:24:22 when it's done, it clears that memory out for use as the input buffer 15:25:02 and updates byte 0 to the start of the freshly compiled interpreter 15:25:42 so you basically leave a trail for ngaro to follow so it can finish the compilation itself the first time it boots 15:26:07 and then all your compilers for whatever language you want are simple one pass compilers 17:41:09 http://pastebin.com/Uv42NNAy woot. 17:44:55 cool :) 18:11:46 just implemented the devices in experimental too. 18:12:04 no idea if they work at all though 18:19:15 http://pastebin.com/aJ0VFyZ3 18:19:25 enough for tonight methinks 18:19:54 ack, 2:20am 18:20:07 nice 18:20:15 can it do tcp stuff? 18:21:43 allegedly. yes, I just implemented the devices added in http://code.google.com/p/retro-language/source/browse/#git%2Fvm%2Fexperimental%2FCSharp%253Fstate%253Dclosed 18:22:12 cool 18:22:21 there's a few things I need to sort out too 18:22:43 but it's enough to boot an image up 18:26:11 I might even be able to port this back to *cough* java 18:26:59 and have identical implementations on both platforms. 18:28:05 would be nice to deploy to android and appspot 19:05:24 oPless: what am I looking at? a c# version with your pluggable devices? 19:05:34 ja 19:05:45 and it's 3am, damnit 19:06:58 very cool... makes me want to get it running in unity3d :) 19:07:01 get some sleep! 22:17:54 --- join: ncv (~quassel@unaffiliated/neceve) joined #retro 22:29:28 --- quit: kumul (Quit: WeeChat 0.3.9) 23:14:12 --- quit: ncv (Ping timeout: 255 seconds) 23:59:59 --- log: ended retro/12.10.07