00:00:00 --- log: started retro/12.08.04
05:15:34 --- join: salvager_ (cc6fab08@gateway/web/freenode/ip.204.111.171.8) joined #retro
05:16:07 <salvager_> Hello everyone
05:16:09 <salvager_> Good Morning
05:17:59 <salvager_> I have a doubt about a retro command,i was wondering if you guys can clear it :)
05:25:08 <crc> ask :)
05:26:58 <salvager_> I have just started learning retro and i am reading this doc (http://retroforth.org/docs/An_Introduction_to_Retro.html) right now 
05:27:22 <salvager_> I am trying to execute this command ': scale ( x-y ) a @ * ;' as a part of example in the doc
05:28:03 <salvager_> The moment i type ': scale ( x-y) a' it asks me 'a ?'
05:28:11 <salvager_> I type '1000 variable: a' 
05:28:18 <salvager_> but it still asks 'a ?'
05:28:27 <salvager_> Am i doing anything wrong ?
05:28:53 <crc> no; the first two lines in that section should have been removed
05:29:09 <crc> the a ? means retro can't find a word or variable named a in the dictionary
05:29:45 <salvager_> yes, so shouldnt '1000 variable: a' define 'a' ?
05:29:59 <crc> yes, as long as you ended the previous definition first
05:30:18 <salvager_> oh, but it doesnt let me end the previous definition
05:30:26 <crc> ;
05:30:45 <crc> when it says a ?, just hit ; and a space to terminate the bad definition
05:31:15 <salvager_> Yes, i am doing that to terminate the definition
05:31:52 <salvager_> So, should i define 'a' first and then define the expression ?
05:32:00 <crc> yes
05:32:35 <CIA-91> retro: 03crc <crc@questor> * rcrc@questor-20120804123203-mt7xsrn1y8apwtjh 10rx/ (3 files in 2 dirs): fix example in hyperstatic global environment section of introductory text
05:34:20 <salvager_> Thank you, i have another question if you dont mind ...
05:35:42 <salvager_> should i use ': 1000 variable: a' to define the variable ...without the ':' it wont let me type the whole definition ..the whitespace escapes the definition of the expression
05:35:59 <crc> no, variables are not defined using :
05:36:58 <salvager_> When i try something like this '1000 variable: a' the listener wont let me type after '1000' it goes to a new line
05:37:20 <crc> that's ok
05:38:30 <crc> it's a side effect of how the interpreter was designed
05:38:52 <crc> it starts a new line after each token is processed
05:39:17 <salvager_> yes, so when i type '1000' it saves the number in the stack
05:40:19 <salvager_> Then, how can variables be created when new lines comes after processing each token ?
05:41:33 <salvager_> Sorry, i was little confused with the interpreter .. I might be spamming this chat room :)
05:41:40 <crc> the newlines are a display issue only
05:41:44 <crc> 1000
05:41:49 <crc> value goes on the stack
05:41:51 <crc> variable: a
05:42:06 <crc> pulls the 1000 off the stack, makes 'a', and sets its value to 1000
05:42:25 <salvager_> when i type 'variable: a' it goes to new line after 'variable:'
05:42:50 <crc> that shouldn't happen 
05:42:55 <crc> which retro version?
05:43:11 <salvager_> new one 11.4
05:43:16 <salvager_> let me try again
05:43:50 <salvager_> the first two commands worked, 
05:44:09 <salvager_> which command should i use, to print the value of variable ?
05:44:17 <crc> a @ putn
05:45:20 <salvager_> yes it worked, thank you for update in documentation
05:45:58 <crc> no problem; sorry for the confusion
05:46:17 <salvager_> i realized that i should not worry about new lines created after each token. I should go on and type the full expression :) 
05:47:09 <salvager_> The example in the docs looked like it was typed in the interpreter, so i was trying to type the same thing and i got confused when i did not see the same output as the example doc
05:48:04 <crc> I'll clarify this in the docs soon
05:48:26 <crc> (I'm at work currently, so time is limited until this evening)
05:48:49 <salvager_> No problem, thanks for the quick updates and clarifying my doubt 
05:48:53 <salvager_> Retro looks like fun 
05:48:54 <salvager_> :)
05:49:22 <salvager_> Have a good day, i might spam this IRC if i have more questions but i will try to understand and not to post too many doubts
05:49:55 <crc> feel free to ask. it may take a while for me to answer, but I'll do my best to help
05:51:05 --- nick: tangentstorm-afk -> tangentstorm
05:51:30 <salvager_> Thank you, i will be working with this more over next 2 days :) 
07:53:25 --- quit: salvager_ (Quit: Page closed)
10:04:13 --- nick: tangentstorm -> tangentstorm-awa
10:06:23 --- join: salvager_ (cc6fab08@gateway/web/freenode/ip.204.111.171.8) joined #retro
10:07:34 <salvager_> Hi crc, i am trying to use console in retroforth 
10:07:51 <salvager_> when i type the command needs console' it shows me all the documentation 
10:09:14 <salvager_> but when i type (console' red) or (console' onRed) or (: console onRed) it does not work
10:28:18 <salvager_> I have got the commands working like this 
10:28:35 <salvager_> ^console'(use the appropriate function here )
10:32:43 <salvager_> but still doing a command like ^console'onRed is not changing anything 
10:38:38 --- join: impomatic (~digital_w@31.185.139.168) joined #retro
12:31:19 --- nick: tangentstorm-awa -> tangentstorm
13:07:24 <crc> the documentation will show, since it's embedded in the source
13:07:34 <crc> use either:  with console'
13:07:43 <crc> (to add it to the search order)
13:07:54 <crc> or  ^console'<function>
13:08:28 <crc> if you enter console' directly, it'll replace the default dictionary
13:09:30 * crc is going to move to a flat dictionary in retro12; the search order bits are becoming annoying to me
13:09:46 <tangentstorm> heya @crc :)
13:09:52 <crc> hi tangentstorm 
13:09:55 <tangentstorm> flat is nice
13:10:53 <crc> I've been doing a flat dictionary in parable (my other language project) and finding it cleaner overall
13:11:16 <saper> is there currently a dictionary stack (or search list)?
13:11:38 <crc> there's a search list
13:11:50 <tangentstorm> i'm still just reading docs at the language level.. i haven't found this yet... i know "words" shows the global words... but how do i see the list of modules, and the list of words in the modules... and how do i see the definition of a word? :)
13:12:22 <crc> words ending in a ' are vocabularies
13:12:34 <crc> add them to the search order, then do words to see what's in them
13:12:43 <tangentstorm> with "need" ?
13:13:00 <tangentstorm> hrm
13:13:03 <crc> needs vocabulary'
13:13:10 <crc> will load the vocabulary from the library
13:13:13 <crc> with vocabulary'
13:13:18 <crc> adds it to the search order
13:13:34 <tangentstorm> and without removes the last one?
13:13:35 <crc> without removes the last one added from the search order
13:13:37 <crc> yes
13:13:49 <crc> and global returns to just the main dictionary
13:15:49 <tangentstorm> so... : showvoc ..... and i'm stuck. i can't define what i want to define without quoting the words "needs"
13:16:43 <tangentstorm> so now i want to go quote the word needs... and i have to find the word for quote... and showvoc is the tool i was hoping would help me with that :D
13:18:38 <tangentstorm> not trying to whine and complain... the answer is RTFM ... but I'm thinking there ought to be a word like "help" that gives this basic "how to explore" information up front
13:21:49 <crc> : showContentsOf ' .chain words "%%" find drop @d->xt do ;
13:21:55 <tangentstorm> it's frustrating, because i can see the steps in my head to define showvoc... take a name, "need" it, "with" it, list the words that are new.. and pop it, unless it was already in the old list
13:21:56 <crc> showContentsOf console'
13:22:22 <crc> you'd still need to load the vocabulary with needs first though
13:23:01 <tangentstorm> yeah.. so to me it's "dup needs with ... output result ... without"
13:23:54 <tangentstorm> i don't care that the words are new.. that's part of the fun... it's that i don't know how to find them with out searching through the manual 
13:24:17 <crc> there's not any method currently
13:24:47 <tangentstorm> yeah, but you're showing me how to make one right now :)
13:26:08 <tangentstorm> so.. okay... i see this code, and i probably should have known ' from ans forth... but then i see .chain and wonder what it means
13:26:22 <tangentstorm> closest i had was that see command that moved off into casket
13:26:50 <tangentstorm> is there another way to see the definition?
13:29:14 <crc> the only decompilers are in examples/autopsy.rx; they use one of the libraries IIRC
13:29:35 <crc> .chain is the class handler for vocabularies
13:36:06 <crc> http://sprunge.us/CcNg
13:36:21 <crc> that's a pretty minimal decompiler using the dissect' library
13:36:34 <tangentstorm> thanks
13:37:00 <crc> no problem
13:37:01 <tangentstorm> my brain's just churning here trying to catch up with you :)
13:37:52 <tangentstorm> i see the debugger in autopsy.rx... very cool.. i added an external one to the pascal vm implementation too
13:39:11 <tangentstorm> ok ... one more question.. i type words and there are lots of words! :D
13:39:35 <tangentstorm> what's retro for: ls | grep "foo"
13:40:44 <crc> no equivilent to ls; there's a crude grep for files in examples/util
13:41:52 <tangentstorm> sorry.. in python i can say dir() to list defined variables... and  [ word for word in dir() if word.startswith("x") ]
13:42:06 <tangentstorm> like.. just how do i filter the output of "words" so i can find the needle in the haystack?
13:42:46 <impomatic> $G 1% N: 2 F: ( N. F. \ 0 > ^ F. 1 + F: ) F. N. < [ F. ! "" #G, N. F. /;] @
13:42:51 <impomatic> Prime factors in Mouse!
13:43:47 <tangentstorm> @crc : like in the /demo/ page, if i type "words"... i don't see ANY words whose names end with "'"... because they scrolled off the screen
13:44:20 <tangentstorm> oh not true there's d' and ' :)
13:44:52 <tangentstorm> and --' ... it's like a word search :D
13:45:21 <tangentstorm> __' i guess
13:45:32 <crc> hmm, this will involve some thinking
13:47:04 <impomatic> Wrong channel, sorry.  In case you're wondering Mouse is a 70's RPN stack based language http://en.wikipedia.org/wiki/Mouse_(programming_language)
13:47:08 <tangentstorm> ... the name of the word isn't in the record that contains its definition, is it?
13:47:33 <tangentstorm> impomatic : looked pretty forthy to me. i wouldn't have known. :)
13:47:51 <crc> the name is in the dictionary, which has the pointers to the definitions
13:48:25 <crc> a minimalistic words would be:  : words last [ d->name puts space ] ^types'LIST each@ ;
13:48:42 <crc> (though that doesn't track search order, only the global dictionary)
13:49:07 <tangentstorm> the dictionary is like a python dictionary then? like a hash?
13:49:11 <tangentstorm> hashmap?
13:50:47 <tangentstorm> ok. now it makes sense why there's a disassembler but not a decompiler
13:51:47 <crc> http://sprunge.us/SZNO
13:51:50 <crc> "'" words
13:52:00 <crc> would then display all words with a ' in their name
13:53:39 <crc> there is a source decompiler in autopsy (the 'see' there has three viewing options, source is one of them)
13:53:45 <crc> it's not flawless though
13:54:17 <crc> dictionary is a linked list, not a hashmap
13:54:40 <tangentstorm> ...
13:54:53 <crc> the structure is covered in the docs  (link to prior, link to class, link to definition, zero terminated name)
13:56:35 <tangentstorm> what's the advantage of linking to the definition rather than just putting it right there inline?
13:57:29 <crc> this makes it possible to discard headers that don't need to be kept (e.g., for embedded targets like arduino with limited memory)
13:58:32 <tangentstorm> headers meaning...?
13:59:19 <crc> the name, pointer fields, etc
13:59:48 <tangentstorm> so a record in the dictionary list
14:00:01 <crc> yes
14:00:06 <tangentstorm> gotcha
14:05:53 <tangentstorm> I guess my conception of what a dictionary is looks something like this:  (link to prior, link to {class/defining-word/whatever}, size, fixed-length-zero-padded name, definition )
14:06:51 <tangentstorm> so if i know the definition, it's always a fixed offset jump to find the metadata
14:07:05 <tangentstorm> is that kind of what you meant by  flat list?
14:09:47 <crc> the name is not fixed length
14:10:14 <crc> but otherwise yes
14:17:25 <tangentstorm> because there's a forth instruction pointer separate from the vm's... forth IP jumps to the definition, not to the actual code... 
14:17:56 <tangentstorm> so you just need a cell saying how long the string is so you can jump, or just scan through it until you find the null.
14:18:28 <tangentstorm> wait
14:18:56 <tangentstorm> your forth pointer can't be working that way or your assembler wouldn't say to call numbers
14:18:59 <tangentstorm> disassembler
14:19:03 <tangentstorm> it would know the name
14:19:18 <crc> the simple disassembler doesn't do name resolution
14:19:33 <crc> the source decompiler in autopsy does
14:19:41 <crc> but it's longer
14:24:12 <tangentstorm> okay. thanks for answering all my questions. I hope i'm not annoying you too much :)
14:24:37 <crc> not at all :)
14:25:15 <tangentstorm> wait one more question
14:25:43 <tangentstorm> i was writing the game design doc most of the morning while salvager was trying to get the console working
14:26:10 <tangentstorm> he seemed to be on the right track from what he was saying but couldn't get any colors to show up.
14:27:44 <tangentstorm> the theory is c, c#, python, ruby and maybe php should all have port 8, right?
14:28:21 <tangentstorm> nm.. you already told me about the comparision page.. sorry.. i'm frazzled :)
14:41:31 <tangentstorm> @crc you know... we could add another field to the dictionary for a documentation string and then just /generate/ the docs directly from the dictionary... add another vm option to test whether docs are included so it's still easy to shrink.
14:42:51 <tangentstorm> kind of the basic idea for the b4 language i want to build is that it's basically a relational database, and the forth dictionary is just one table.
14:44:05 <saper> good for porting to AS/400
14:45:21 <tangentstorm> since that's so not forthlike, i was just paying attention to ngaro and ignoring retro... but since we wound up using it for the game i seem to be learning it anyway. :)
14:45:27 <crc> tangentstorm: port 8 still depends on ANSI support from the terminal in most cases
14:46:06 <crc> re: dictionary and documentation: I'm exploring this option
14:47:39 <tangentstorm> also, the quotes and the lexer have me torn... as a user of the language, i like your choices... but in introducing people to forth from scratch i like the simplicity of ." even though it looks completely silly to me :)
14:48:21 <tangentstorm> ok.. back to work. thanks again. :)
14:48:48 <crc> np
14:49:05 * crc is off to make supper; will check back in later
15:04:27 --- join: Kumul (~Kumul@cvx-ppp-66-50-52-16.coqui.net) joined #retro
15:20:05 --- quit: Kumul (Remote host closed the connection)
15:25:12 --- join: ricericericerice (~puppy@74-142-222-186.dhcp.insightbb.com) joined #retro
17:32:46 --- nick: tangentstorm -> tangent-afk
18:09:33 --- nick: tangent-afk -> tangentstorm
18:23:27 --- join: Kumul (~Kumul@cvx-ppp-66-50-52-16.coqui.net) joined #retro
19:52:44 --- join: arescorpio (~arescorpi@222-206-17-190.fibertel.com.ar) joined #retro
21:19:26 --- quit: ricericericerice (Quit: Leaving)
22:39:12 --- quit: arescorpio (Read error: Connection reset by peer)
23:48:41 --- quit: impomatic (Quit: http://RetroProgramming.com)
23:59:59 --- log: ended retro/12.08.04