00:00:00 --- log: started forth/09.07.14
01:04:58 --- quit: aguaithefreak (Read error: 54 (Connection reset by peer))
01:14:31 --- quit: mathrick (Read error: 110 (Connection timed out))
01:19:09 --- quit: gogonkt (Read error: 110 (Connection timed out))
01:25:54 --- join: mathrick (n=mathrick@wireless.its.sdu.dk) joined #forth
01:32:02 --- join: aguaithefreak (i=aguai@114-44-21-47.dynamic.hinet.net) joined #forth
01:37:22 --- quit: sunwukong ("bye")
01:46:30 --- join: nighty__ (n=nighty@210.188.173.245) joined #forth
01:54:05 --- quit: mathrick (Read error: 60 (Operation timed out))
02:03:23 --- quit: maht (Read error: 104 (Connection reset by peer))
02:07:49 --- join: maht (n=maht__@85-189-31-174.proweb.managedbroadband.co.uk) joined #forth
02:19:51 --- quit: saon (Read error: 110 (Connection timed out))
02:21:25 --- join: mathrick (n=mathrick@wireless.its.sdu.dk) joined #forth
03:12:53 --- quit: mathrick (Read error: 110 (Connection timed out))
03:20:38 --- join: GeDaMo (n=gedamo@212.225.107.61) joined #forth
03:37:36 <schme> Anyone hacked up a forth for the beagleboard ?
03:44:09 --- join: X-Scale2 (n=email@89-180-180-228.net.novis.pt) joined #forth
03:47:15 --- quit: TR2N (Nick collision from services.)
03:50:18 --- nick: X-Scale2 -> TR2N
03:52:54 --- quit: nighty__ (Client Quit)
04:39:11 --- join: saon (n=saon@75.102.128.133) joined #forth
04:58:10 --- quit: GeDaMo ("Leaving.")
05:03:51 --- join: mathrick (n=mathrick@wireless.its.sdu.dk) joined #forth
05:22:31 --- quit: mathrick (Read error: 110 (Connection timed out))
05:24:21 --- part: saon left #forth
05:29:54 --- join: mathrick (n=mathrick@users177.kollegienet.dk) joined #forth
05:51:44 --- join: jauaor (n=araujo@gentoo/developer/araujo) joined #forth
06:30:01 --- join: gogonkt (n=info@59.38.200.244) joined #forth
07:27:13 --- quit: mathrick (Read error: 104 (Connection reset by peer))
07:27:16 --- join: _mathrick (n=mathrick@users177.kollegienet.dk) joined #forth
07:29:01 --- quit: aguaithefreak (Read error: 104 (Connection reset by peer))
07:29:58 --- nick: _mathrick -> mathrick
07:31:31 --- join: aguaithefreak (i=aguai@114-44-21-47.dynamic.hinet.net) joined #forth
08:02:21 --- quit: nighty^ (Read error: 54 (Connection reset by peer))
08:04:08 --- join: nighty^ (n=nighty@x122091.ppp.asahi-net.or.jp) joined #forth
08:35:16 --- quit: Quartus` (Read error: 60 (Operation timed out))
08:38:03 --- join: GeDaMo (n=gedamo@212.225.107.61) joined #forth
08:42:41 --- quit: aguaithefreak (Read error: 104 (Connection reset by peer))
08:43:02 --- join: aguaithefreak (i=aguai@114-44-21-47.dynamic.hinet.net) joined #forth
10:29:21 --- join: I440r (n=mark4__@c-69-136-171-118.hsd1.in.comcast.net) joined #forth
10:36:35 <ment> what exactly does DOES> ?
10:37:12 <GeDaMo> It replaces the default behaviour of a CREATEd word
10:37:28 <I440r> ment does> is kind of complex
10:37:46 <I440r> and while the above description is kind of close it doesnt help much hehe
10:37:53 <GeDaMo> :P
10:38:05 <I440r> does> is used IN a crating word to create a word and define its behaviour
10:38:15 <I440r> for example a constant. wjhen executed returns its value
10:38:26 <I440r> a possible definition for the word constant might be
10:38:38 <ment> it's possible to express does> in simpler "words"?
10:38:38 <I440r> : constant ( n1 --- ) create , does> @ ;
10:38:49 <I440r> so i could then say 0 constant zero
10:39:14 <I440r> the word constant would create the word called "zero" and comma in its value
10:39:32 <I440r> when i execute the word zero everything after does> is executed. that feteches and returns the constants value
10:39:47 <I440r> dont try understand how it does this, just understand what it does so that you can use it
10:40:21 <schme> indeed "see does>" is probably not much help ;)
10:40:47 <I440r> no lol
10:40:50 <ment> well i'm fiddling around jonesforth, so i don't have any words missing there unless i implement it myself :)
10:41:10 <I440r> what OS do you use?
10:41:17 <ment> linux
10:42:10 <I440r> it will probably NOT help much but read
10:42:14 <I440r> http://www.isforth.com/kern.php?t=compile&p=19
10:42:34 <I440r> and the next page after it. understnding ;uses and ;code would help understand does>
10:42:51 <I440r> but those are VERY isforth specific. 
10:43:06 <I440r> is your PC an x86 machine?
10:43:20 <ment> yep
10:43:26 <I440r> if so i would personally recommend isforth (im the autnor so im biased)
10:43:47 <I440r> tho... im rabidly anti ans so most people here would not recommend isforth :)
10:43:50 <ment> it's small and asm so i can give it a shot :)
10:44:01 <I440r> you code asm then?
10:44:44 <ment> well anything lowlevel
10:45:34 <I440r> my kinda coder :)
10:45:53 <schme> isforth! that was the name of it.
10:45:55 <I440r> the www.isforth.com site is not complete yet but its getting there. its on hold atm tho cuz im job searching
10:46:03 <schme> I440r: I have been googling for it for 2 weeks! I forgot the name.
10:46:06 <I440r> not worked in months
10:46:09 <I440r> lol
10:46:14 <I440r> its in the topic (or was)
10:46:27 <I440r> yup. last url in this channels topic
10:46:28 <I440r> hiding
10:46:34 <schme> aaargh
10:46:41 <schme> There it is.
10:46:42 <I440r> lol
10:46:50 <schme> I never /topic
10:47:47 <jauaor> http://nopaste.info/d6e79e1522.html
10:47:56 <schme> I started working on a half arsed x86-64 long mode port of it, but I happened to rm it by accident.
10:47:56 <jauaor> an example of an imperative construct
10:48:18 <I440r> oopts
10:48:23 <schme> (which I need to use it to write an assembler for arm so I can port it again :P)
10:48:38 <I440r> yes i need an arm assembler too :(
10:48:47 <schme> gas ;)
10:49:05 <I440r> and isforth is majorly in need of an x86 assembler extension but i cant write it
10:49:14 <I440r> ive tried MANY times
10:49:30 <schme> writing x86 assemblers seems like a road to insanity.
10:49:31 <ment> x86 assembler extension like asm/disasm written in forth?
10:50:27 <schme> Well thanks for the link there. Seems I must go enjoy real life :(
10:50:27 <I440r> yes but i will not accept an assembler extension that only worked backwards
10:50:35 <I440r> i am very picky
10:50:49 <schme> I440r: Good thinking. backwards assembly is ugleeh.
10:50:56 <I440r> any assembler extension for isforth would HAVE to be able to take 100% of the existing soruces with only MINOR changes
10:51:10 <I440r> like the injection of a space or two here or there
10:51:16 <ment> backwards means it's able to disasm forth core and nothing else?
10:51:28 <I440r> backwards means traditional forth assembler
10:51:32 <I440r>  5 # ax mov 
10:51:38 <I440r> instead of mov ax, # 5
10:51:57 <I440r> i would only accept the latter for isforth (tho the former could be an option in the assembler too)
10:52:06 <I440r> i wrote a nice 8051 assembler that works both ways
10:52:12 <ment> postfix asm seems like neat idea
10:52:25 <I440r> wrote an assembler, disassembler and emulator
10:52:38 <I440r> there are many forth assemblers that support it. i cant use any of them
10:52:56 <I440r> the assembler that comes with win324th or what ever its called supports it
10:53:03 <I440r> but that assemblers license is unacceptable 
10:53:08 <ment> do you have NIH complex? :)
10:53:08 <I440r> or i would ninja it
10:53:13 <I440r> NIH ?
10:53:20 <ment> Not-Invented-Here
10:53:24 <I440r> no
10:53:41 <I440r> i would use the assembler that came with FPC if it would work in isforth but it wont
10:53:59 <I440r> because in FPC coded definitions go in a separate segment to colon definitions and text and headers
10:54:12 <I440r> they are all separate. i only separate code from headers
10:54:29 <I440r> and the FPC assembler relies heavilly on the way FPC works.
10:54:50 <I440r> i would use the one from win32 forth but its PURE gpl. and that would taint my license
10:55:22 <I440r> i was given permission to use the one in bigforth i think. but it is totally unuseable
10:55:35 <I440r> i dont know how ANYONE could use that gobbldegook lol
10:55:49 <I440r> and its also bass ackwards
10:56:16 <I440r> if someone wrote a useable assembler for isforth i would LOVE to use it. but i have some quirks that have to be catered to lol
10:56:38 <I440r> tho... if it worked i could mayb modify it to cater to those (maybe)
10:57:06 <I440r> i have some phobias. like cluster fuck unreadable code. makes me shudder
10:59:27 <ment> btw. in which segment allocates forth memory for malloc, when heap is occupied by HERE?
11:00:26 <I440r> in isforth?
11:01:00 <I440r> isforth has ONE block of memory that i split in tw. the lower half is code (list space - pointed to by here)
11:01:02 <I440r> the other is head space
11:01:06 <ment> well in anyforth :)
11:01:15 <I440r> isforths "allocate" word allocates a NEW heap from the OS
11:01:24 <I440r> im going to assume thats how other forths do it to
11:01:42 <I440r> tho... my allocate word is actually a full blown memory allocator
11:01:59 <I440r> it allocates a heap and you can allocate pieces of said heap with a granularity of 16 bytes
11:02:38 <I440r> if you use the linux syscall to allocate a buffer your going to have a granularity of 4k.  allocate 50 bytes you get 4k buffer
11:03:00 <I440r> also. in linux, dont think in terms of segments. its all a flat address space
11:03:06 <I440r> think of it as sections
11:03:17 <ment> well i thought that compilation of words doesn't perform explicit allocations other than HERE +!
11:03:46 <TreyB> You get an allocation of PAGE_SIZE, depending on the CPU, right?
11:03:48 <I440r> yes. MOST forths dont have a malloc type allocator
11:04:06 <I440r> if you need 40 bytes for your decompressor to decompress your data you usually "allot" it
11:04:14 <ment> I440r: that 4K you get from mmap? :)
11:04:18 <I440r> that puts it in forths heal
11:04:20 <I440r> yes
11:04:25 <ment> I440r: also the alloc in isforth that's phk?
11:04:55 <I440r> my allocate word uses mmap to allocate a large heap. you can then use allocate to bite of chunks of that as needed
11:04:59 <I440r> phk?
11:05:22 <ment> phkmalloc, nevermind
11:05:33 <I440r> isforths allocate word is a full blown memory manager
11:05:46 <I440r> think of LINUX as the BIOS :)
11:06:01 <I440r> and isforth is the OS (tho its kind of lacking in that regard as yet :)
11:06:04 <TreyB> A BIOS that doesn't suck :-)
11:06:10 <I440r> heh
11:06:32 <I440r> isforth doesnt use any external libraries. it does everything by using syscalls.  
11:06:45 <I440r> so memory allocation in isforth would have to use mmap.
11:06:51 <ment> I440r: my question was - what kind of full blown memmory manage? :) first-fit/best-fit/buckets of same-sized chunks broken from 4K page...
11:06:58 <I440r> i didnt like having a granularity of 4k so i wrote a memory sub allocator
11:07:24 <I440r> its not QUITE a "best fit"
11:08:11 <I440r> i have linked lists of blocks of free memory) that are sorted into buckets of various memory size ranges
11:08:57 <I440r> if you try allocate a 30 byte buffer your going to get a 32 byte buffer. which will be from bucket 2.  bucket 2 can have any block of memory from 32 bytes to 63 bytes in size 
11:09:18 <TreyB> Pool-based?
11:09:19 <I440r> if i sorted the buckets smallest to largest then my allocator would be a true best fit
11:09:21 <I440r> yes
11:09:29 <I440r> i dont bother with that
11:10:00 <TreyB> Sorted the buckets?
11:10:06 <I440r> if the first block of memory in the bucket is larger than you requres its split into 2 buckets and one is given to you. the other is then added to an appropriate bucket
11:10:10 <I440r> no
11:10:31 <I440r> each bucket contains descriptors for various memory buffers
11:10:35 <TreyB> Ah, so not pool-based.
11:10:45 <I440r> it IS and it isnt heh
11:11:23 <I440r> ok i have to organize my thoughts lol
11:11:43 <I440r> http://www.isforth.com/ext1.php?t=memory
11:11:57 <ment> what happends when i try to allocate 16K buffer?
11:11:58 <I440r> or you can read that. im not sure if that description does the allocator justice tho
11:12:00 <I440r> thers alot in there
11:12:04 <I440r> you get a 16k buffer
11:12:08 <ment> from mmap?
11:12:16 <I440r> if there is not enough space in any existing pool a new one is created
11:12:27 <I440r> no. you get it from my allocator
11:12:52 <I440r> if you ask for a buffer thats larger than any available block a new pool is allocated and you get a piece or ALL of that
11:13:01 <ment> ok, then i'm going to dealloc it, how do you know how large it is and to whom return it?
11:13:30 <I440r> you deallocate a buffer by passing its address to "free"
11:13:47 <I440r> free will search all pools for an allocated buffer of that size and will deallocate it
11:14:05 <I440r> if that deallocation deallocates the last buffer in the heap then the heap itself is returned to the OS
11:14:23 <I440r> and as you deallocate buffers they are merged with adjacent free buffers
11:14:35 <I440r> so if you have  [ free ] [ allocated ] [free ]
11:14:39 <I440r> and deallocate the middle buffer
11:14:47 <ment> ok, i'm going to read the sources :)
11:15:02 <I440r> a NEW buffer is creates with a size equal to all 3 buffers
11:15:09 <I440r> eep
11:15:37 <I440r> its some complex code but ive been working on making it more readable. not sure how successfull i was at that lol
11:16:05 <I440r> src/ext/memory.f is where you want to look tho
11:16:15 <ment> i'm looking in there right now
11:16:17 <I440r> http://www.isforth.com/ext1.php?t=memory  <-- suggest uyou read that and the sources together
11:16:48 <ment> the basic problem i was trying to solve was that mmap returns non-continuous range of pages
11:17:15 <TreyB> Physical or logical pages?
11:17:22 <I440r> ya well you probably have heap randomization turned on
11:17:29 <I440r> you can fix that. in fact... to run isforth you HAVE to
11:17:37 <ment> TreyB: logical
11:17:38 <I440r> echo 1 >/proc/sys/kernel/randomize_va_space 
11:18:05 <ment> well to properly lookup page sizes i had to maintain i growing hashtable
11:18:21 <I440r> mmap was never suitable for allocating buffers for doing DMA (err there might be a parameter you can pass i forget)
11:18:32 <ment> (walking long list of allocated pages just isn't an option)
11:18:44 <I440r> they probably arent long
11:18:59 <I440r> unless you are allocating thousands of 16 byte buffers
11:19:22 <ment> which just might happend, when you running lisp interpretr atop of forth
11:19:41 <I440r> tho. i had also considered hasing the buffers in this way so that deallocation could be faster
11:19:54 <I440r> wouldnt take much of a change to add it
11:20:10 <I440r> my "buckets" could be pointers to hash tables
11:20:48 <I440r> so all 16 byte buffers get added to the first bucket but that bucket refers to a hash table of N lists of buffers
11:20:57 <ment> btw looking into src/bot, are you going to port isforth to DS? :)
11:21:04 <I440r> just hash the address in question (not sure how to do that :)
11:21:07 <I440r> nintendo?
11:21:11 <ment> yep
11:21:15 <I440r> lol i was working on that yes
11:21:27 <ment> that's just the only game i was missing from ds
11:21:27 <I440r> but i need an arm assembler
11:21:45 <I440r> i have a hacked nintendo ds, a dserial2 cart and an xport-ii cart
11:21:59 <I440r> that xport-ii cart cost almost twice as much as the ds itself did
11:22:28 <I440r> i want to be able to telnet into a ds from my linux box and be at an isforth terminal
11:22:30 <ment> what for do you need xport cart? :)
11:22:43 <I440r> FORTH!!!!!!!!!! lol
11:22:48 <ment> you have built-in wifi
11:23:08 <I440r> and the drivers for it werent completed when i started working on this
11:23:19 <I440r> i needed a way to communicate with the ds and the xport solved that
11:23:29 <I440r> tho... the dserial-ii solves it in a better way
11:23:48 <I440r> i would NEED to completely rewrite the wifi drivers in asm/forth
11:23:53 <I440r> in order to use them
11:24:05 <I440r> no fuckin way im using the C version
11:24:30 <ment> why not?
11:24:36 <I440r> tho sgstair did an freekin awesome job RE'ing the hardware
11:24:52 <I440r> because i want to have the FORTH do it
11:25:13 <ment> ah, programming language xenofobia :)
11:25:17 <I440r> its just hardware. you poke in this value over here. enable that interrupt over there and ding! you got wifi
11:25:26 <I440r> im not a fan of C
11:25:36 <I440r> but thats not WHY i want it done in forth
11:25:43 <I440r> i want it in forth because i want to learn HOW.
11:25:52 <I440r> but this is all just a pipe dream atm
11:26:03 <I440r> need an arm assembler ext for isforth first :0
11:29:38 <ment> asm is just detail, the bigger problem is how to design the gui (and on-screen keyboard is just not the answer)
11:30:12 <I440r> which is why i wanted to telnet in to the ds
11:30:31 <I440r> the ds key and emit would do I/O accross the telnet session
11:30:55 <I440r> there would be words defined to input and output locally too. the touch screen and the display etc
11:31:09 <ment> the point of having forth on ds is that you can play around while you commute, not to telnet from your notebook to ds and run forth there
11:31:13 <I440r> so you could write and test a game and the two I/O methods would not be mashing on each other
11:31:40 <I440r> that could also be an option with the right extension. "stand alone mode"
11:31:54 <I440r> but that would require a touch screen keyboard be the input
11:32:01 <I440r> and the other display would be the output
11:32:26 <I440r> you could dedicate the "start" button as a "reset this thing so i can break into my code"
11:32:50 <I440r> or some vulcan nerve pinch button combination 
11:33:27 <ment> but onscreen keyboard sucks (for writing anything but name of your pokemon character)
11:33:50 <I440r> how would you suggest. the ds does not have a serial port so you cant attach some keyboard
11:34:09 <I440r> and a touch screen keyboard is better than the lefty righty uppy downy select letter meothd thats built in
11:34:24 <ment> first, you can write and debug entire forth in the emulator (no xport needed)
11:34:42 <I440r> and how do you USE that forth on the ds?
11:34:52 <I440r> sure... you can write the entier forth and its application on the pc
11:35:01 <I440r> then move it to the ds and use the app
11:35:03 <ment> you have cursor keys (and L) for the left hand, stylus in the right
11:35:08 <I440r> but thats NOT developing an app ON the ds
11:35:16 <I440r> suxors
11:35:32 <I440r> the touch screen keyboard for input is superior to that method
11:35:37 <ment> with cursor keys you can select an input dictionary (all words, latest defined, most used, conditionals ... )
11:35:39 <I440r> you can type quite fast that way
11:35:46 <ment> and with L there's keyboard
11:35:47 <I440r> ugh
11:36:04 <I440r> i would become very annoyied with kaving to use that method lol
11:36:09 <ment> i've used lisp interpreter on pal which worked that way
11:36:14 <ment> palm
11:36:19 <I440r> it would be faster to just TYPE  ": blah my new definition goes here ;
11:37:15 <ment> well download zxds and run z80 forth in there to check it really sucks :)
11:38:33 <I440r> lol
11:38:44 <I440r> thers already a ds forth btw
11:39:07 <I440r> but i dont think the sources are available. cant remember
11:39:11 <ment> there is? give a link
11:40:48 <I440r> cant find one
11:41:07 <I440r> but the #dsdev channel or maybe #gbadev has the author
11:41:10 <I440r> not on this server tho
11:41:20 <I440r> i forget where they are located but im sure you know of them
11:43:50 <I440r> yea blitz.org
11:44:01 <I440r> irc.blitz.org #dsdev and #gbadev
11:44:54 <I440r> thers also a #dslinux there :)
11:44:56 <I440r> btw
11:45:06 <ment> i'll ask there
11:45:30 <schme> Ah ds. I am looking for a forth for ARM. Any ideas?
11:45:53 <I440r> well im gona go back to job searching, if u have any questions about isforth PM me here or email the address in the isforth release :)
11:46:01 <I440r> ya. write one is the usual answer lol
11:46:06 <I440r> which was on my todo list
11:46:08 <schme> indeed.
11:46:23 <I440r> tho... there are some good free arm forths out there
11:46:37 <I440r> ive seen a couple. some of them are for the acorn tho :/
11:46:37 <schme> I'd like one for me beagleboard.
11:47:02 <schme> But ya. I guess writing one is where it's all at (:
12:29:45 --- quit: GeDaMo (hubbard.freenode.net irc.freenode.net)
12:29:45 --- quit: TR2N (hubbard.freenode.net irc.freenode.net)
12:30:25 --- join: GeDaMo (n=gedamo@212.225.107.61) joined #forth
12:30:37 --- join: TR2N (i=email@89-180-180-228.net.novis.pt) joined #forth
13:04:39 <kspaans> schme: Have you looked at muForth? I think it's for ARM.
13:05:13 <kspaans> Ah, cool apparently it targets ARM and AVR.
13:48:03 --- join: jewel_ (n=jewel@dsl-247-203-169.telkomadsl.co.za) joined #forth
14:38:00 --- quit: jewel_ (Read error: 113 (No route to host))
15:01:29 --- quit: GeDaMo ("Leaving.")
17:06:52 --- join: nighty__ (n=nighty@210.188.173.245) joined #forth
20:05:20 --- quit: I440r (Read error: 110 (Connection timed out))
20:06:02 --- join: I440r (n=mark4__@c-69-136-171-118.hsd1.in.comcast.net) joined #forth
21:35:35 --- quit: I440r ("Leaving")
21:49:11 <ment> is there a default forth looping construct which uses return stack for storing counter?
22:38:19 --- quit: jauaor ()
23:59:59 --- log: ended forth/09.07.14