00:00:00 --- log: started retro/13.01.30
03:54:26 --- quit: karswell (Ping timeout: 245 seconds)
07:29:31 --- join: impomatic (~digital_w@148.54.112.87.dyn.plus.net) joined #retro
09:30:52 --- join: ncv (~quassel@unaffiliated/neceve) joined #retro
11:49:04 <tangentcode> hm.
11:50:16 <tangentcode> so... i can see how to implement a "halt" function that drops everything retro is doing and basically reboots 
11:50:32 <tangentcode> but i suspect it's already in here somewhere
11:50:46 <tangentcode> maybe something to throw an exception or error?
11:50:58 <tangentcode> anyone know what it might be called?
11:58:50 <tangentcode> hmm... 0; actually fits the bill pretty nicely...
12:44:40 --- join: Mat2 (~claude@91-65-144-133-dynip.superkabel.de) joined #retro
12:44:51 <Mat2> hello !
12:45:12 <tangentcode> hey Mat2  :)
12:45:21 <Mat2> hi tangentstorm 
12:45:31 <Mat2> what's new ?
12:46:54 <tangentcode> i'm trying to make a minesweeper game in retro
12:47:08 <tangentcode> it's hard!
12:47:22 <tangentcode> not the minesweeper part... the translating it to retro part :)
12:47:29 <tangentcode> hard for me i mean :)
12:47:33 <tangentcode> what about you?
12:48:00 <Mat2> I'm coding the hashed dictionary routines for metro at current
12:48:28 <Mat2> as always performance matters in this scope
12:49:16 <Mat2> but it progresses
12:50:23 <Mat2> (the bad news is, that because of the way retro handles strings I can not port it)
12:52:36 <tangentcode> what's the problem?
12:53:35 <Mat2> for efficient hashing, the retro strings need to be packed and this compensates the performance advantage
12:54:50 <Mat2> apart from this problem, the retro language relate on features of it's linked list approach
12:57:03 <tangentcode> oh right. you'd need to use the bytestrings module to do it in retro
12:58:09 <Mat2> metro is using hashed arrays for dictionarys
12:58:20 <Mat2> ^dictionaries
12:58:21 <tangentcode> library/bstrings.rx
12:58:42 <tangentcode> cool :)
12:59:05 <tangentcode> crc and i were talking about doing something like that when discussing case-insensitivity a while back
12:59:25 <tangentcode> since making it case insensitive would hurt the current compare
12:59:39 <Mat2> the changes would be worth the effort for sure
13:00:37 <Mat2> the language has types and this will include hash tables
13:00:56 <tangentcode> handy
13:01:50 <tangentcode> to really use b-strings quickly, you probably want to be able to read and write individual bytes to ram
13:03:26 <tangentcode> hmm
13:03:37 <Mat2> that's not the problem because 3 instructions are bundled so and masking inclusive store or load operations are free
13:04:07 <Mat2> (this simplifies the compiler by the way)
13:04:31 <tangentcode> but when you're talking about the dictionary, speed isn't a huge issue, because a few extra calculations are still much faster than the speed at which even a fast typist can type
13:04:58 <tangentcode> i mean in retro/ngaro
13:05:13 <Mat2> speed matters for immediate functions for examples
13:06:30 <tangentcode> why? usually they only walk the list once, at compile time, and then they just write the address, right?
13:07:34 <Mat2> no, there are cases, where code-generating functions need to dynamical parse word addresses
13:08:39 <Mat2> think of structure defining words for example. These should be extensible
13:09:13 <Mat2> or prototype like object-orientated objects
13:12:10 <tangentcode> hmm. i guess so
13:19:50 <tangentcode> i want to make an    enum| a b c d |    like   variables| a b c d |  but that assigns the numbers 0, 1 , 2, 3 etc as it goes.
13:22:26 <tangentcode> but i don't need to because it's already there!
13:27:14 <Mat2> how have you implemented the enum| word ?
13:50:01 <tangentcode> needs enum' : enum| 0 enum'enum| ;
13:50:08 <tangentcode> :)
13:50:13 <Mat2> *g*
13:50:45 <Mat2> ok, in metro you can define that word simply with:
13:51:08 <tangentcode> http://code.google.com/p/retro-language/source/browse/library/enum.rx
13:51:22 <Mat2> 0 u1 [1+] iota :enum
13:52:10 <Mat2> result is an array with values 0,1..n
13:52:14 <tangentcode> u1 defines a variable?
13:52:34 <tangentcode> hmm
13:52:57 <tangentcode> no, i don't want an array. i want to define a bunch of constants
13:53:02 <Mat2> u1 picks the second stack element (replacement for stack)
13:53:25 <Mat2> these are constants because metro is functional language and do not have variables
13:53:46 <Mat2> you can access all elements though dictionary entries at default
13:54:07 <Mat2> 20 enum :test
13:54:40 <tangentcode> after    enum| a b c d |   ...     d c b a .s     should print:   <0> 0 1 2 3 4
13:54:59 <Mat2> {a1 a2 a3 ...} test label
13:55:39 <Mat2> ok:
13:56:06 <Mat2> 4 enum :test {a b c d} 'test label
13:56:27 <Mat2> d c b a will print 0 1 2 3 4
13:57:02 <tangentcode> i can't visualize how that works yet
13:57:15 <tangentcode> what does : do ?
13:57:27 <tangentcode> is it a prefix?
13:57:39 <Mat2> create an dictionary entry (word defination)
13:58:07 <Mat2> { } defines symbolic values and
13:59:20 <Mat2> 'test label extend the content of test with label properties for each numerical value
13:59:32 <Mat2> (this feature I have borrowed from lisp)
14:01:14 <Mat2> probably I will create an simplified set of word definations for symbolic definations
14:03:59 <tangentcode> a b c d are properties of test? like indexes?
14:04:40 <Mat2> sort of, yes
14:05:29 <tangentcode> it's like test.a test.b and so on?
14:06:01 <Mat2> yes
14:06:03 <tangentcode> or like  a = lambda : test[0]
14:06:33 <tangentcode> i can underst and the lambda version
14:06:51 <tangentcode> i don't understand how they can be test.a and test.b
14:06:59 <tangentcode> because
14:07:58 <tangentcode> 4 enum :test 4 enum :other {a b c d} 'test label {a b c d} 'other label
14:08:08 <Mat2> for every constant defination, each cell can hold a value and property label which compiles to code to push that value onto the stack if parsed
14:08:23 <tangentcode> better:
14:08:30 <tangentcode> 4 enum :test 4 enum :other {a b c d} 'test label {d c b a} 'other label
14:09:08 <Mat2> oh, I se your confucsion
14:09:10 <tangentcode> are a b c d  properties of test, or of other? or do they each have their own?
14:10:19 <Mat2> the parser binds each property to the last referenced constant defination so a b c d
14:11:06 <Mat2> can be properties for both test and other without problems
14:12:33 <tangentcode> interesting. so how would i get the "test" ones back?
14:12:45 <Mat2> 'test
14:13:28 <Mat2> a b c d <- test.a test.b test.c test.d
14:13:33 <Mat2> 'other
14:13:54 <Mat2> a b c d <- other.a other.b other.c other.d
14:14:07 <Mat2> a b c d <- other.a ....
14:14:08 <tangentcode> so ' doesn't put anything on the stack..
14:14:21 <Mat2> exact
14:14:57 <tangentcode> so
14:15:17 <tangentcode> hrm
14:15:28 <tangentcode> interesting.
14:22:05 <Mat2> ' define a label value
14:22:20 <Mat2> you can also define:
14:23:08 <Mat2> 4 enum 'a 'b 'c 'd :test
14:24:08 <Mat2> the parser test the type of each token at run-time, so:
14:24:09 --- quit: ncv (Read error: Connection reset by peer)
14:24:19 <Mat2> 3.14 :pi
14:24:26 <Mat2> <- create a constant
14:24:47 <Mat2> 3 0.14 + :pi
14:25:09 <Mat2> <- create a word which returns 3.14 onto the stack
14:25:41 <Mat2> both, constant and code definations are defined though the same prefix
14:26:23 <Mat2> that is possible because the parser know the state and type of each token
14:26:51 <Mat2> and can compile code dependent of it
14:27:33 <Mat2> this way, type conversions are handled automatical too, so in most cases ther is no need for explicit type declarations
14:29:25 <Mat2> $2F 0.42 + <- unsigned integer + float = float
14:34:25 <tangentcode> sounds pretty cool. i will definitely try it out :)
14:34:54 <tangentcode> but for now i need to sleep. see you!
14:35:04 --- nick: tangentcode -> tangentsleep
14:37:27 <Mat2> ciao
14:37:45 --- quit: Mat2 (Quit: Verlassend)
14:42:19 --- join: kumul (~Kumool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
15:36:42 --- join: kumool (~Kumool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
15:39:30 --- quit: kumul (Ping timeout: 248 seconds)
15:40:22 --- quit: impomatic (Quit: impomatic)
15:46:53 --- join: impomatic (~digital_w@148.54.112.87.dyn.plus.net) joined #retro
16:16:04 --- quit: impomatic (Quit: http://corewar.co.uk)
16:51:47 --- join: kumul (~Kumool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
16:54:57 --- quit: kumool (Ping timeout: 264 seconds)
16:58:28 --- join: kumool (~Kumool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
17:01:06 --- quit: kumul (Ping timeout: 248 seconds)
17:01:35 --- join: arescorpio (~arescorpi@190.55.93.118) joined #retro
17:31:34 --- join: kumul (~Kumool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
17:34:10 --- quit: kumool (Ping timeout: 248 seconds)
18:13:18 --- join: kumool (~Kumool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
18:15:46 --- quit: kumul (Ping timeout: 248 seconds)
18:25:38 --- join: kumul (~Kumool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
18:28:02 --- quit: kumool (Ping timeout: 248 seconds)
18:50:34 --- join: kumool (~Kumool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
18:53:06 --- quit: kumul (Ping timeout: 248 seconds)
19:19:27 --- quit: kumool (Quit: Leaving)
19:23:29 --- join: kumul (~Kumool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
19:40:05 --- join: kumool (~Kumool@c-76-26-237-95.hsd1.fl.comcast.net) joined #retro
19:42:37 --- quit: kumul (Ping timeout: 244 seconds)
19:43:38 --- quit: arescorpio (Quit: Leaving.)
21:42:50 --- quit: kumool (Quit: Leaving)
21:43:43 --- nick: tangentsleep -> tangentcode
23:59:59 --- log: ended retro/13.01.30