00:00:00 --- log: started retro/12.10.27
00:39:08 --- join: tangentstorm (~michal@108-218-151-22.lightspeed.rcsntx.sbcglobal.net) joined #retro
00:44:09 <tangentstorm> anyone awake in here?
00:45:21 <tangentstorm> i'm working on the test suite for ngaro, and starting to suspect that the "loop" opcode is never actually used anywhere.
00:49:18 <tangentstorm> hmm. okay. i take that back. it's just never used by name. 
00:51:38 <tangentstorm> bbl
00:51:39 --- quit: tangentstorm (Quit: leaving)
05:59:55 <crc> foucist: I'm not afraid of colorforth, I just prefer punctuation
06:05:45 <crc> I implemented a colorforth mode over retro in the past, but never really found a use for it
06:06:42 <crc> (as an aside, the prefixes in parable [and to a lesser degree, retro] correspond generally to the use of colors in colorforth)
07:22:21 --- quit: ivan`` (Ping timeout: 246 seconds)
07:24:17 --- join: ivan`` (~ivan@unaffiliated/ivan/x-000001) joined #retro
12:17:09 --- join: tangentstorm (~michal@108-218-151-22.lightspeed.rcsntx.sbcglobal.net) joined #retro
13:27:47 --- join: Mat2 (~claude@91-65-144-133-dynip.superkabel.de) joined #retro
13:27:52 <Mat2> hello
13:30:01 <tangentstorm> heya :)
13:32:10 <Mat2> hi tangentstorm
13:32:28 <Mat2> the parallela funding has reached his goal
13:36:31 <Mat2> I think about buy one board (64 cores + an ARM dual core and 1 gB ram is very attractive)
13:40:00 <tangentstorm> yeah i saw that ! :)
13:40:21 <tangentstorm> i'm interested in it. i just don't have any money coming in at the moment.
13:41:39 <Mat2> there plan to offer there board for less than $100
13:41:53 <Mat2> let us see
13:42:04 <tangentstorm> i think he said $99 in the video
13:42:38 <tangentstorm> still out of my price range ... i stopped doing client work a while back to focus on this project, so i'm like... dirt poor at the moment :)
13:45:05 <Mat2> ups
13:45:50 <Mat2> seems I'm not the only one with flexible time management :)
13:47:14 <Mat2> an alternative would be this:
13:47:18 <Mat2> http://www.duinomite.com/
13:58:54 <Mat2> tangentstorm: I have assembled a small interface for handling text and semigraphic though a character and a attribute map. Have you some special requirements for text processing ?
14:06:46 <tangentstorm> not sure what you mean
14:08:25 <tangentstorm> actually, i do have special requirements :)
14:08:53 <tangentstorm> lately, i've been making pretty heavy use of the 32 lowest ascii characters
14:09:43 <Mat2> for terminal control ?
14:09:55 <tangentstorm> in particular ^A, ^B, ^C ( for hierarchical documents, like lisp/xml ), ^G ( as sort of a "todo" marker ), ^L for page breaks in code
14:10:26 <tangentstorm> and characters 28, 29, 30, 31 for tabular data
14:10:40 <tangentstorm> also terminal control though, yeah ( character 27 / escape )
14:11:30 <tangentstorm> so my requirement is to be able to change the font and make these characters visible. 
14:11:36 <tangentstorm> not sure if that's what you meant.
14:11:58 <Mat2> that was my question, yes 
14:12:23 <tangentstorm> i also plan to make use of a non-unicode character set, accessed with the shift in / shift out control codes
14:12:31 <tangentstorm> i forget what their codes are
14:13:05 <tangentstorm> http://www.cs.tut.fi/~jkorpela/chars/c0.html
14:15:58 <tangentstorm> my plan was to add a new ngaro device, or possibly extend the character generator device, and give you the ability to change the font for a particular character
14:16:45 <Mat2> changing the font would require SDL or a similar library
14:17:21 <Mat2> the first retro versions based on ngaro used this method
14:17:57 <Mat2> possibly you will find useful code there
14:21:14 <Mat2> ok, I will add some functionality for font selection
14:21:32 <Mat2> and palette support
14:23:46 <Mat2> in my repro you will find a subdirectory named backport. Next week you will find a ngaro version with a new device; That is the backport of my navm code for character generation
14:24:55 <Mat2> in this directory
14:30:29 <Mat2> need some sleep, ciao
14:30:35 --- quit: Mat2 (Quit: Verlassend)
14:41:29 <tangentstorm> :)
16:49:43 <tangentstorm> so, anyone: what's a good unit test for the jump instructions?
16:53:14 <tangentstorm> i guess maybe something like the test for if in tests/base.rx     [ -1 [ 1 ] [ 2 ] if ] expected: { 1 }
17:09:31 --- join: Z_Mass (~zachary@bas1-cornwall24-1242467096.dsl.bell.ca) joined #retro
17:17:56 --- join: kumul (~kumul@173.215.130.73) joined #retro
17:24:38 <tangentstorm> heya Z_Mass! :)
17:34:49 <Z_Mass> Hey :)
17:41:05 <tangentstorm> you want a quick go job? :)
18:02:21 <Z_Mass> tangentstorm: Sure, why not.
18:11:32 <tangentstorm> cool :)
18:11:55 <tangentstorm> basically, i'm making a test suite for the virtual machine
18:13:33 <tangentstorm> i'm trying to make sure they all work the same way, and fix mine ( which doesn't )
18:14:39 <tangentstorm> so... the mission is to add a feature to the go version that will dump the content of the two stacks and the image when the program exits
18:17:14 <Z_Mass> I can try, but I really have no idea how the retro code works
18:18:01 <tangentstorm> well, what it should be is three arrays of integers
18:18:15 <tangentstorm> one for each stack and one for the ram
18:18:37 <Z_Mass> Gotcha. Is the Go code in the lpmc repo, or somewhere else?
18:19:05 <tangentstorm> yep. one sec
18:19:15 <Z_Mass> k
18:20:09 <tangentstorm> hrm
18:20:17 <tangentstorm> https://github.com/LearnProgramming/retro-language/tree/master/vm/vm/complete/go
18:20:29 <tangentstorm> it shouldn't be vm/vm though... i must have screwed something up
18:21:21 <tangentstorm> ugh. that repo's a complete mess
18:21:45 <tangentstorm> i tried to just push the version from google code and github wouldn't let me
18:23:20 <tangentstorm> git clone https://code.google.com/p/retro-language/    
18:23:37 <tangentstorm> i'll have to redo the lpmc one
18:24:59 <tangentstorm> here's what the python version looks like : http://bazaar.launchpad.net/~tangentstorm/retro-language/docwork/view/head:/vm/complete/retro.py#L402
18:25:32 <Z_Mass> Cool, I'll clone that and have a look. :)
18:25:48 <tangentstorm> cool :)
18:37:16 <Z_Mass> So would this be going in main.go?
18:37:51 <tangentstorm> i just got go-mode installed for emacs.. just starting to look at this
18:38:10 <tangentstorm> i think maybe so
18:38:25 <tangentstorm> those flag.XXX stuff look like command line arguments
18:40:55 <tangentstorm> var dump_after = flag.Bool("dump", false, "dump stack and ram after running for ngarotest.py")
18:41:06 <tangentstorm> something like that maybe to get --dump ?
18:42:33 <Z_Mass> Hmm, never used the flag package, will have to check it out.
18:43:42 <tangentstorm> how come sometimes the code uses = and sometimes := ?
18:44:18 <Z_Mass> := is an assignment within a function
18:45:02 <tangentstorm> i'm looking at line 105  err = vm.Run() ... right above it is vm := ...
18:46:02 <Z_Mass> err is declared on line 68
18:46:30 <tangentstorm> oh so := is like declare + assign?
18:46:48 <Z_Mass> := is like using auto in C++. = is for re-assigning a value, or declaring with a type
18:46:57 <Z_Mass> Yep
18:47:21 <tangentstorm> cool
18:49:46 <tangentstorm> hrm
18:49:54 <Z_Mass> ?
18:50:03 <tangentstorm> looking at core() in core.go 
18:50:22 <tangentstorm> it looks like data and addr might be internal to that function
18:50:39 <tangentstorm> whereas ram is attached to the vm via   vm.img
18:51:21 <tangentstorm> this version of the vm has support for multiple concurrent instances in their own goroutines
18:51:52 <tangentstorm> i'm wondering if they all share the same image but different stacks?
18:52:58 <docl> tangentstorm: have you read this book? http://www.catb.org/esr/writings/taoup/
18:52:58 <Z_Mass> Not too sure, lol
18:53:15 <Z_Mass> Is there like a wiki of some sorts on this?
18:53:16 <tangentstorm> heya docl. i've read parts of it
18:53:16 <docl> it has some good discussion of OO vs other paradigms
18:53:34 <docl> I'm reading through it slowly now. good stuff.
18:54:07 <docl> I'm up to chapter 5
18:54:24 <tangentstorm> Z_Mass: there was a wiki for retro but it was covered in spam... there's a bunch of docs, but probably not specifically for the go version
18:55:39 <tangentstorm> docl: esr's a pretty smart guy. i read through a lot of his stuff back when the cathedral and the bazaar came out
18:55:55 <Z_Mass> Gotcha. Will probably have to read up on how it works a little more before I can write some tests for it :P
18:56:06 <tangentstorm> Z_Mass: i can talk you through it if you can supply the go syntax :)
18:56:23 <tangentstorm> oh
18:56:31 <tangentstorm> sorry. i haven't explained this very well :)
18:56:36 <docl> tangentstorm: yeah me too. this is a bit more recent than that though.
18:57:02 <tangentstorm> Z_Mass: what i'm shooting for here is one test suite that applies to all the implementations of the vm
18:57:40 <tangentstorm> so i'm writing the tests in python here: http://bazaar.launchpad.net/~tangentstorm/retro-language/docwork/files/head:/test/ngaro/
18:58:30 <tangentstorm> ngarotest.py has a little assembler in it that generates an image for each test
18:58:33 <Z_Mass> tangentstorm: I can try but I'm no Go expert ;). I think I'm a little too tired tonight but in the next few days for sure :)
18:59:12 <Z_Mass> So each languages version will use a python test suite?
18:59:57 <tangentstorm> yeah... if you look at the makefile in the directory i just linked, you can see where i'm testing both the python and pascal versions
19:00:36 <tangentstorm> ngarotest.py generates a bunch of little image files, then invokes whatever virtual machine you want and tells it to run that image and --dump
19:00:54 <tangentstorm> image meaning like... a snapshot of ram
19:01:36 <tangentstorm> i'm just using python for the tests because python ships with a testing library
19:02:23 <Z_Mass> I see. I've gotta head out for a bit, but I'll look this stuff over and hopefully I can help you out this week. 
19:02:34 <tangentstorm> the idea is that no matter which vm you're using, if you pass in an image and --dump you should get the same output
19:02:38 <tangentstorm> ok
19:02:59 <tangentstorm> cool. it's no rush or anything. just thought you might want something to do :)
19:03:16 --- quit: Z_Mass (Quit: Leaving)
19:07:31 <tangentstorm> docl: not sure this is the book i was thinking of. thanks for the link :)
19:25:41 <docl> tangentstorm: the history of computing stuff made me think of you. also the discussion of object-oriented programming.
19:26:32 <tangentstorm> where does he talk about that?
19:26:47 <tangentstorm> rule of composition maybe?
19:27:00 <docl> one of the early chapters. history I think, maybe chapter 2?
19:27:31 <tangentstorm> http://www.catb.org/~esr/writings/taoup/html/unix_and_oo.html
19:29:13 <docl> that must be the one
19:31:48 <tangentstorm> that reads like advocacy for pascal to me... i must be kind of biased :)
19:31:57 <docl> :)
19:33:31 <tangentstorm> i think unix itself does a lot of what OO is meant to offer. you don't really need an OO language inside it.
19:33:56 <docl> yes, that is basically what I got out of it.
19:34:04 <docl> hiding one process from the other
19:34:34 <docl> with the sole exception of the textual output stream
19:35:22 <tangentstorm> that's the part i don't like :)
19:35:36 <docl> (well, mostly textual.)
19:35:47 <tangentstorm> stream good, textual bad i guess
19:37:27 <docl> unix pipes can deal with binary data too, from what I understand.
19:38:18 <tangentstorm> yeah, it's just bytes
19:38:55 <tangentstorm> i'm just grumbling about the line-oriented nature of everything. 
19:39:43 <docl> the fact that newlines usually mean something?
19:41:52 <tangentstorm> no, that everything is passing strings around instead of structured data
19:42:05 <tangentstorm> so you have to have parsers for everything
19:42:20 --- join: obobo_ (~chatzilla@dyn-dsl-ly-98-124-43-42.nexicom.net) joined #retro
19:43:55 --- quit: obobo (Ping timeout: 244 seconds)
19:43:57 <docl> hmm. he cites non-textual records as a major source of opacity, and a form of premature optimization.
19:44:00 --- nick: obobo_ -> obobo
19:44:26 <docl> textual stuff is more transparent.
19:45:04 <tangentstorm> i don't mean raw binary data
19:46:41 <tangentstorm> i mean it would be nicer if instead of everything focusing on lines, you had pipelines for something like xml, json, etc.
19:47:04 <docl> json good, xml bad (imo)
19:47:05 <tangentstorm> you can do that now, but the standard tools don't know about those formats, only lines
19:48:22 <docl> yeah he notes that xml is good for deeply nested stuff, but doesn't play well with traditional unix tools.
19:48:34 <tangentstorm> json > xml :) ... not sure i want to go so far as to say json is good :)
19:48:47 <docl> :)
19:49:29 <tangentstorm> yaml > json > xml
19:49:45 <docl> colon-delimited text maybe, like unix password files
19:50:09 <tangentstorm> but yeah... the problem is with the tools though.. they're all line-oriented, so they can't see that structure
19:50:31 <tangentstorm> but what if you want to put colons in your text ?
19:50:44 <docl> \:
19:51:11 <tangentstorm> http://www.cs.tut.fi/~jkorpela/chars/c0.html
19:51:59 <tangentstorm> characters 1,2, and 3 seem to me like they'd do the job xml or lisp does, for nested structures.
19:52:54 <tangentstorm> 28 - 31 give you a delimiter to put between tables, groups ( header row vs content rows ), records, and fields
19:53:36 <tangentstorm> no conflicts with anything you'd want to type by hand, since they're invisible
19:54:13 <docl> invisible is part of the problem though... you don't necessarily know they are there.
19:54:32 <tangentstorm> how do you know tabs and spaces and linefeeds are there, though? :)
19:55:29 <tangentstorm> i'd probably show something. they look like ^A ^B ^C in vim and emacs
19:55:47 <docl> by how they move stuff around I guess
19:55:53 <tangentstorm> as soon as i get that terminal workign i'm going to make up some glyphs for them
19:55:53 <docl> yeah
19:56:37 <tangentstorm> i have my emacs set up to show only the information between ^L (form feed, i think)
19:56:49 <tangentstorm> so i can page back and forth without scrolling
19:56:59 <tangentstorm> i think it's really readable
19:57:22 <tangentstorm> i was worried compilers would gripe but they don't seem to
23:03:10 --- quit: obobo (*.net *.split)
23:03:11 --- quit: timg (*.net *.split)
23:03:42 --- quit: kumul (*.net *.split)
23:43:31 --- quit: karswell (*.net *.split)
23:43:36 --- quit: oPless (*.net *.split)
23:43:51 --- quit: ivan`` (*.net *.split)
23:44:16 --- quit: yiyus (*.net *.split)
23:44:24 --- quit: saper (*.net *.split)
23:45:24 --- quit: foucist (*.net *.split)
23:52:46 --- join: foucist (~foucist@ps14150.dreamhost.com) joined #retro
23:52:47 --- join: ivan`` (~ivan@unaffiliated/ivan/x-000001) joined #retro
23:52:47 --- join: karswell (~coat@93-97-29-243.zone5.bethere.co.uk) joined #retro
23:52:47 --- join: saper (saper@wikipedia/saper) joined #retro
23:52:47 --- join: oPless (~oPless@lart.doosh.net) joined #retro
23:52:47 --- join: yiyus (1242712427@je.je.je) joined #retro
23:59:59 --- log: ended retro/12.10.27