00:00:00 --- log: started retro/13.02.25
08:55:55 --- join: kumul (~mool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
09:18:31 --- join: kumool (~mool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
09:19:20 --- quit: kumul (Ping timeout: 256 seconds)
09:44:07 --- join: kumul (~mool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
09:46:32 --- quit: kumool (Ping timeout: 256 seconds)
09:46:51 --- join: kumool (~mool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
09:48:44 --- quit: kumul (Ping timeout: 246 seconds)
09:49:29 --- join: ncv (~quassel@unaffiliated/neceve) joined #retro
09:52:39 --- quit: kumool (Ping timeout: 252 seconds)
09:53:52 --- join: kumool (~mool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
10:11:13 --- join: kumul (~mool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
10:13:44 --- quit: kumool (Ping timeout: 256 seconds)
10:20:38 --- join: kumool (~mool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
10:22:43 --- quit: kumul (Ping timeout: 252 seconds)
10:27:29 --- join: kumul (~mool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
10:29:41 --- quit: kumool (Ping timeout: 246 seconds)
10:47:30 --- join: kumool (~mool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
10:49:38 --- quit: kumul (Ping timeout: 246 seconds)
10:59:39 --- join: Mat2 (~claude@91-65-144-133-dynip.superkabel.de) joined #retro
10:59:49 <Mat2> hello 
11:00:55 <tangentstorm> hey Mat2 
11:01:03 <Mat2> hi tangentstorm
11:01:12 <Mat2> howe's going ?4
11:01:29 <Mat2> sorry, what is going on ?
11:02:06 <tangentstorm> good i guess. kind of frustrated that i haven't shipped anything :)
11:02:47 <tangentstorm> other than that just great.
11:03:01 <Mat2> I guess that mean you are an ebay seller
11:03:06 <tangentstorm> haha
11:03:17 <tangentstorm> no i meant shipping any of this code that i've been working on for a year and a half
11:03:39 <Mat2> ah, I know this source of frustation
11:05:58 <Mat2> by the way, the problem with the SDL terminal code was simply a bug with the DirectX drivers for windows supporting 8 bit modes. As I heard it work fine on Windows 8
11:06:39 <tangentstorm> huh!
11:07:07 <tangentstorm> well, it was frustrating, but i'm kind of glad we had those problems because it lead me to try zengl :)
11:07:39 <Mat2> it can be zengl suffers from the same bug
11:08:16 <Mat2> otherwise it is a nice alternative to SDL for Pascal
11:08:51 <Mat2> anyhow, I am not sure about further development of these library
11:12:18 <tangentstorm> i'm probably going to port it to retro pascal
11:12:32 <tangentstorm> assuming i ever get to that point :)
11:12:54 <tangentstorm> what are you up to?
11:13:08 <Mat2> for sure, you will [yoda mode]
11:13:34 <tangentstorm> :)
11:14:19 <Mat2> Saiwa progresses well
11:14:53 <Mat2> the state system is a bit complex but that can be factored after finishing
11:15:06 <Mat2> I mean factored out
11:15:12 <tangentstorm> are you still porting all the retro routines?
11:15:50 <Mat2> no, I have an initial function-set of around 20 routines
11:16:00 <Mat2> at current
11:16:18 <tangentstorm> is that enough to make the dictionary and all that?
11:16:22 <Mat2> yes
11:16:43 <tangentstorm> i don't suppose it's online? :)
11:17:04 <Mat2> at 1 march in its cureent state
11:17:10 <Mat2> ^current
11:18:15 <tangentstorm> i will be patient :)
11:18:46 <Mat2> at current experimenting with syntax features
11:19:44 <Mat2> for example, instead of closures I have typed conditionals which are bound to the function state (and type save)
11:21:20 <tangentstorm> i don't understand 
11:21:49 <Mat2> in retro, you can define closures without application:
11:22:03 <Mat2> [ dup * ] :test
11:22:12 <Mat2> sorry,
11:22:20 <Mat2> : test [ dup * ] ;
11:22:38 <Mat2> : foo test do ;
11:22:45 <Mat2> or simply
11:22:49 <Mat2> [ dup * ]
11:22:56 <Mat2> : foo do ;
11:23:42 <tangentstorm> i'm surprised that works
11:25:04 <Mat2> in the last example that is dangerous because the closure is not bound to a specific application
11:26:43 <tangentstorm> yeah i thought you were saying it did a lot more than it does
11:27:10 <Mat2> moreover, the listener can only call the closure where a jump would be faster for the interpreter
11:27:26 <Mat2> ^yes
11:27:44 <Mat2> a conditional like these:
11:28:02 <Mat2> | 0n * | do :test
11:28:19 <tangentstorm> that's a conditional?
11:28:24 <Mat2> yes
11:28:29 <tangentstorm> what's the conditon?
11:29:11 <Mat2> the do function test the stack state and execute the conditional only depend on it
11:29:21 <Mat2> ^depending on iot
11:29:40 <Mat2> | 0n * | 1 | 1 < do :test
11:29:50 <Mat2> that is more intuitive
11:29:55 <tangentstorm> is 0n the condition?
11:30:04 <Mat2> no: 1 <
11:30:21 <Mat2> will execute | 1 | if tos < 1
11:30:40 <Mat2> in all other cases | 0n * | is executed
11:31:16 <Mat2> *if the type and stack state is compatible (only one stack item on the stack in this example)
11:31:28 <Mat2> 0n duplicates tos
11:31:39 <tangentstorm> would not have guessed that
11:31:53 <Mat2> 1n convert tos to a natural value
11:32:01 <Mat2> 2n is swap
11:32:05 <Mat2> 3n is rot
11:32:22 <Mat2> 1i would convert tos to an integer
11:32:28 <tangentstorm> would you use that in actual code, or is this just for bootstrapping?
11:32:56 <Mat2> no, it will used for all code inclusive the bootstrapping one
11:33:22 <tangentstorm> that's the first time you've said something terrible :D
11:33:27 <tangentstorm> i would never remember all that
11:34:13 <Mat2> I write a book here beside coding: "The Saiwa programming language - An introduction"
11:34:13 <tangentstorm> is there a mnemonic system there that i'm not seeing?
11:34:20 <Mat2> yes
11:34:31 --- join: kumul (~mool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
11:34:56 <Mat2> the first character is the stack item (0 means duplicating tos, 1 = tos, 2 = second stack item, 3 the third)
11:35:07 <Mat2> the next character is the type of these item
11:35:10 <tangentstorm> hrm... ok... i guess i can see the pattern for 1n 2n 3n, but 0n surprises me
11:35:55 <tangentstorm> does 0+ mean the same as 2 * ?
11:36:13 <tangentstorm> have the same effect i mean
11:36:36 <Mat2> the second character is the type (for conversion)
11:36:49 <Mat2> i = integer
11:36:52 <Mat2> n = natural
11:36:54 <Mat2> c = char
11:36:57 <Mat2> s = string
11:37:00 <Mat2> p = pointer
11:37:03 --- quit: kumool (Ping timeout: 276 seconds)
11:37:26 <tangentstorm> what does   10 0s  do then?
11:37:34 <tangentstorm> produce   10 "10" on the stack?
11:38:05 <Mat2> 10 is pushed as natural value on the stack
11:38:43 <Mat2> and 0s convert these value to a string, replacing the value with a pointer to the resulting string
11:38:55 <Mat2> and duplicate that
11:39:06 <tangentstorm> so "10" "10"
11:39:08 <Mat2> yes
11:39:55 <tangentstorm> i find prefixes confusing 
11:40:09 <tangentstorm> i would call what you just described s0
11:40:15 <tangentstorm> because the s happens first
11:40:35 <tangentstorm> also it wouldn't look like a number anymore 
11:41:19 <tangentstorm> but now that i understand the pattern, i do like the idea
11:42:10 <Mat2> the parser interpreting prefixes and postfixes, I have choosen the postfix character here because it is natural for a RPN language to specifiy the number before its action
11:43:01 <tangentstorm> oh because   3s  doesn't 's' the top item on the stack
11:45:02 <Mat2> that was my reason. Saiwa convert all types if needed automatical (and this syntax ensures that always the type is guessed right or must be choosen explicite)
11:45:26 <Mat2> so the type-sytem is as strict as in Pascal
11:45:27 <tangentstorm> okay, it was surprising at first but it seems consistent. cool.
11:46:09 <tangentstorm> pascal's type system is too strict! :)
11:46:20 <tangentstorm> at least without generics
11:48:05 <Mat2> the language need a strict type system because the parser can compile routines to parallel executed routines this way if possible
11:49:19 <Mat2> so Saiwa is able to parallelize code automatical without hints from the programmer
11:49:26 <tangentstorm> cool
11:49:49 <tangentstorm> i guess it's not that pascal's system is too strict it's that it's not powerful enough
11:50:22 <tangentstorm> like i'm looking at some old ISO pascal code, and the guy had to duplicate 3 functions
11:50:50 <tangentstorm> they all walk through an array of records and look up values by name
11:50:58 <Mat2> I think every type system have some drawbacks
11:50:59 <tangentstorm> but three different record types
11:52:04 <tangentstorm> even though the code is word for word identical :/
11:53:24 <Mat2> on the other (the very dark side), the type system of C is inconsistent and it is impossible to check up against some common type conflicts
11:54:07 <Mat2> one of the restrictest type-system I know is that of ADA
11:54:21 <tangentstorm> i've never used it but i've looked at it
11:54:37 <Mat2> that's because ADA support coroutines and parallel processing I guess
11:55:04 <Mat2> it's also a very type-rich system
11:55:16 <tangentstorm> yeah.
11:55:55 <Mat2> the type system of Saiwa is a compromise of both because it's late evaluated and you convert most types
11:56:45 <tangentstorm> it does sound like fun :)
11:57:57 <Mat2> :D at current there exist only a generic pointer type for example. I am thinking of extending it with specific pointer types (pointer to integer for example)
11:59:41 <tangentstorm> what about structured types like records?
12:01:14 <Mat2> I think to found a better alternative to structures, evaluating functions
12:01:32 <Mat2> (because Saiwa is a somewhat pure functional language)
12:02:35 <tangentstorm> like what? even haskell has records
12:02:56 <tangentstorm> they're just not mutable
12:08:15 <Mat2> I know. The thing is: In haskell a record is just syntactical sugar for defining functions
12:08:59 <Mat2> That is not a dependancy of the core language
12:09:15 <tangentstorm> i guess i don't understand what a defining function is
12:10:03 <tangentstorm> i mis-parsed that, didn't i? :)
12:10:26 <Mat2> a function which return a function executing the functionality for accessing a given "record" element
12:11:00 <tangentstorm> hmm.. ok
12:12:01 <Mat2> Forth do not have records either (but you can create words implementing records)
12:12:24 <tangentstorm> yeah. i'm writing a database :)
12:12:24 <Mat2> in Saiwa you can create functions implementing records at demand
12:12:46 <Mat2> or to be specific: The functionality of records
12:13:41 <tangentstorm> well... like i said, i look forward to trying it :)
12:13:55 <tangentstorm> i don't want to just keep pestering you with questions :D
12:16:56 <Mat2> because it is a uncommon language I will write a book beside "Programming in Saiwa - an introduction"
12:17:54 <Mat2> and because practice is the best teacher, I have an colleque who waiting using it for his 3D modelling
12:19:53 <tangentstorm> i'll definitely try it
12:22:48 <Mat2> thanks, that are 3 users (I really thing this language has potential)
12:39:23 <Mat2> hmm, division by zero is not good at all ;)
12:40:16 <Mat2> should the parser generate an error or should I encode it as result=zero instead ?
12:44:44 <tangentstorm> result=NaN maybe?
12:45:02 <tangentstorm> or if you're using crazy symbols why not the infinity one?
12:48:24 <Mat2> hmm, so much possibilites
12:56:42 --- quit: kumul (Read error: Connection reset by peer)
13:37:19 --- quit: ncv (Remote host closed the connection)
14:03:59 --- join: ncv (~quassel@92.83.227.80) joined #retro
14:03:59 --- quit: ncv (Changing host)
14:03:59 --- join: ncv (~quassel@unaffiliated/neceve) joined #retro
14:16:20 <Mat2> get some sleep, ciao
14:16:27 --- quit: Mat2 (Quit: Verlassend)
14:44:40 --- quit: ncv (Remote host closed the connection)
14:49:12 --- join: kumul (~mool@76.26.237.95) joined #retro
15:18:16 --- quit: kumul (Ping timeout: 245 seconds)
16:01:14 --- join: kumul (~mool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
16:36:37 --- quit: kumul (Ping timeout: 272 seconds)
16:51:29 --- join: kumul (~mool@76.26.237.95) joined #retro
17:57:43 <tangentstorm> crcx: i've written a couple assemblers for ngaro, in various languages and with different names for the opcodes
17:58:02 <tangentstorm> this is the latest, ported from ngarotest.py: https://github.com/sabren/b4/blob/master/b4a/b4a.js
17:59:08 <tangentstorm> i think i might add ( ) syntax for stack comments and maybe something to allow writing strings and sequences of numbers
17:59:44 <tangentstorm> for my course i want something that looked more like assembly so it would be visually distinct from the forth level
18:00:58 <tangentstorm> i plan to port the core interpreter from commentary.txt/kernel.rx to this 
18:03:21 <tangentstorm> i wasn't sure if you'd want it for the release because it breaks from the naming conventions. you're welcome to it but i wouldn't be at all bothered if you didn't want it. ;)
18:04:44 <tangentstorm> i don't think the ngaro.js changes or the retro.pas stuff are going to be ready and shippable in time for 11.5 though. :/
18:06:11 <tangentstorm> if you want mineswpr, i have two versions: one in org-mode (kind of useless without emacs) and the other is the tangled retro source (kind of messy)
18:06:33 <tangentstorm> i'm really not happy with this but i don't have anything better yet
18:08:31 <tangentstorm> i've also been slowly working on https://github.com/sabren/b4/blob/master/ref/retro-trail.org
18:09:18 <tangentstorm> which is commentary.txt + kernel.rx partially put back together in org-mode
18:09:30 <tangentstorm> org-mode can generate kernel.rx from this file
18:10:31 <tangentstorm> and can also generate syntax-highlighted source: lpmc.ne.sabren.com/retro-language/doc/retro-trail.html
18:10:39 <tangentstorm> er http://lpmc.ne.sabren.com/retro-language/doc/retro-trail.html
18:10:51 <tangentstorm> clearly not finished
18:13:21 <tangentstorm> Sorry for just dumping all this here... I guess my point is if you can prioritize any of that, I'd be happy to focus on some part of it :D
18:13:38 <tangentstorm> also a few things here: https://github.com/sabren/b4/tree/master/rx
18:15:24 <tangentstorm> and this is a tool to syntax-highlight and inspect retro code, but mostly limited to the kernel.rx at the moment. not sure how useful it is in general, but it's kind of pretty. :)
18:15:27 <tangentstorm> https://github.com/sabren/b4/blob/master/go/untangle.py
18:15:37 <tangentstorm> the output in the console i mean. not the code :(
18:15:39 <tangentstorm> </rant>
19:43:16 --- quit: kumul (Ping timeout: 245 seconds)
20:43:27 --- join: kumul (~mool@76.26.237.95) joined #retro
20:55:40 --- join: dessos (~dessos@c-174-60-176-249.hsd1.pa.comcast.net) joined #retro
21:14:06 --- quit: kumul (Ping timeout: 245 seconds)
21:56:39 --- quit: tangentstorm (Ping timeout: 244 seconds)
21:58:38 --- join: tangentstorm (~michal@108-218-151-22.lightspeed.rcsntx.sbcglobal.net) joined #retro
22:27:19 --- quit: harrison (Ping timeout: 264 seconds)
22:48:19 --- nick: tangentstorm -> tangentsleep
22:58:01 --- join: ncv (~quassel@92.83.227.80) joined #retro
22:58:01 --- quit: ncv (Changing host)
22:58:01 --- join: ncv (~quassel@unaffiliated/neceve) joined #retro
23:53:15 --- quit: ncv (Remote host closed the connection)
23:59:59 --- log: ended retro/13.02.25