00:00:00 --- log: started forth/06.03.19 00:16:29 --- quit: JasonWoof ("off to bed") 01:11:15 well, theoretically you don't need fp, you only need registers with an appropriate size, to get fine resolutions. 01:14:24 I know that I need to support fp/simd simd isn't something where I have a bad feeling, but with fp I have a bad feeling, it's a feature of a CPU which is attached to it with a coprocessor, it's only there because some like calculating with real numbers. 01:15:36 I think integers fit very good to computers, they can be used directly as indexes in memory where fp needs to get converted. 01:38:07 like PoppaVic says I see a shitload of issues 01:41:08 why are CPUs so dirty, why does OOP rock the world and why the hell is finding a clean solution so difficult? 03:09:21 --- join: neceve (n=Clau@unaffiliated/neceve) joined #forth 05:28:39 --- join: PoppaVic (n=pete@0-1pool47-119.nas30.chicago4.il.us.da.qwest.net) joined #forth 06:24:02 hmm... 06:24:30 oxygene: you alive, chief? 06:53:21 --- join: Lars_G (n=lars@unaffiliated/lars-g/x-000001) joined #forth 07:23:38 --- quit: Lars_G ("Leaving") 07:24:47 --- part: saon left #forth 07:40:41 --- quit: Quartus (Read error: 104 (Connection reset by peer)) 07:49:11 damn 07:55:24 bah, boring 07:55:49 yeppers, and I don't "bore" 07:57:53 floating point math is somehow crazy, I'm not wondering why there is the issue of 'types' 07:58:55 simple, Forth was never designed as more than a register-general system 07:59:41 --- join: saon (i=1000@unaffiliated/saon) joined #forth 07:59:44 I'm dancing all over "typing" even now... It's no easy chore, let me tell you 07:59:56 yeah, generally types are something which is inflexible 08:00:55 right, Forth took the other path. Nothing else does. 08:01:52 yes, I'm also stumbling over it, but I think I'll say that there is only one type and that is the accumulator length. 08:03:25 well, that gets painful as well. 08:05:00 I'm trying to see, even today, if we can deal in typing and types and classes, etc. 08:05:27 I believe I have found the worst-case wedge I need to flog shit into shape 08:06:53 and that would be? 08:07:21 * PoppaVic sighs 08:08:12 why do you sighs? 08:08:22 It has to do with the damned stacks actually acting like streams, and every word just blithely chewing and spitting whatever they think they see. The compiler almost needs prototypes or some locals-variant 08:09:11 My notes are suggesting nearly a paradigm-shift, and I *STILL* want to embrace as much as I can 08:11:36 a paradigm-shift? 08:13:36 yeah.. I think forth is busting up on the reefs of "compiler and interpreter are too trusting and variant" 08:15:19 well, I don't understand really what you mean. I see only the tip of the iceberg of what you mean. 08:15:37 virl: as-is, shit.. you can't even guarentee that 10 inputs are eaten and shit into 2. You can envision the variants 08:17:14 I'm trying to work my way thru all the issues, keeping foremost the IDEA of extensiblity and parse/compile and organization. 08:21:25 shit into 2? some of your speech phrases aren't really understandable by one like me who gets taught british english. do you mean that a word pops 10 cells from the data stack and then it pushes 2 cells on it? 08:22:39 sorry 08:23:16 I've about concluded that forth likes to treat stacks as bi-directional streams.. Eating and shitting into them 08:23:31 yes 08:24:17 and, there is no (current) plan to codify such that words are written and deployed "properly" - folks just trust. 08:25:16 So, I'm trying to decide how much "testing/verify/trust" is really required, and how that affects at least the "compiler" 08:25:52 ..granted, while interactive, sheesh - unless we want to slow down everything to a trickle, there is little we can do 08:26:38 virl: throw me a full email address, I'll drop you this developing doc written in html 08:27:24 08:27:31 I got peeved and switched from plain-text notes all over, but the doc is still a bit sloppy. 08:29:02 so your problem is to manage the number of inputs to a word and it's outputs, nice overhead to every word. 08:30:02 Yeah, but it should ONLY matter at compiletime 08:31:04 incoming 08:31:13 ...splash 08:32:56 Today, I am beginning to cogitate a "compiler-stack" and more musing over "types" 08:35:17 I also think that headers need a few more fields and checking 08:38:54 virl: also... I am seriously considering relocating "headers" into a database.. I think we are fighting too hard to make "fast and compact" without considering "portable". 08:56:55 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #forth 09:00:08 --- quit: tathi (Client Quit) 09:00:19 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #forth 09:00:30 hi, tathi 09:00:43 --- join: JasonWoof (n=jason@pdpc/supporter/student/Herkamire) joined #forth 09:00:43 --- mode: ChanServ set +o JasonWoof 09:00:44 hi 09:01:14 Thinking of a snooze ;-) I've been tippy-typing notes all morning. 09:02:07 --- quit: GoHst10 (Remote closed the connection) 09:02:31 ..hell, I even sent virl a copy of the html-notes... I dunno' if we shall ever agree, but I keep seeing "connections". 09:03:16 --- quit: PoppaVic ("Pulls the pin...") 09:04:07 --- join: Infight9 (i=WINNT@12-208-98-237.client.insightBB.com) joined #forth 09:04:07 hi :) 09:05:03 hi JasonWoof 09:05:18 and did you something in factor? 09:05:56 --- join: PoppaVic (n=pete@0-1pool46-68.nas30.chicago4.il.us.da.qwest.net) joined #forth 09:06:06 I compiled it 09:06:13 sorry for delay 09:06:27 compiled? 09:06:40 virl asked if I did something with factor 09:06:49 ahh 09:07:01 I could not really warm to it 09:07:59 I have this strange behavior sometimes 09:08:04 I install a program and start it up 09:08:16 indending to "check it out" 09:08:29 but then I start it up and I don't have anything I'm trying to accomplish 09:08:43 so I stare at it for a while and fiddle with the commands/buttons I see 09:08:51 and then I don't use it 09:10:00 and the same you did with factor? 09:10:06 yep 09:10:28 I wanted to poke around the graphic interface 09:10:38 but then I found out that the graphic interface doesn't work on my box 09:10:52 hehehe 09:11:44 yes it also doesn't like my box, well I ned to load the glxcore, but I'm too lazy to do that. 09:13:35 works on mine :) 09:16:43 the last time I tried the ui, it was incredible slow, even a swing java application is faster. 09:17:49 what are your feelings about this image thingy? which you need to load at startup 09:18:17 where is this? 09:18:26 like gforth does? 09:19:20 or smalltalk 09:21:23 smalltalk is a mess.. Just spare me 09:22:31 I don't really know about the factor image thing 09:27:58 you don't know about? 09:28:42 everytime you start it you need to type ./f 09:32:36 well, I know that 09:32:45 and I know it takes like 30 minutes or something to make 09:32:59 but that's about it 09:35:09 --- join: crc (i=crc@pool-70-110-205-24.phil.east.verizon.net) joined #forth 09:37:39 hi crc 09:42:06 JasonWoof, what are your feelings about it? 09:42:40 iirc, both gforth and a few other load an "imagefile" - and it's no more than a sec or 2 09:43:40 hi virl 09:46:42 PoppaVic, yes, factor also takes a sec or 2, but it needs to compile 5-10 minutes or something like that, only to rectify 09:47:06 hmm, a poor 'loader' or 'linker' 09:47:08 I only wanted to rectify that 09:59:50 I hear that the long compilation time is because of the optomizing compiler 10:00:00 and because much of the compiler is running in some sort of interpret mode 10:00:17 virl: I don't have much feelings about the image file because I don't even know what it is 10:00:29 I'm annoyed that it takes 20-30 minutes to generate 10:00:53 it's raw machine code I think. 10:01:06 I think they mean "relocatable" or "virtual" and the time is all ratty-code 10:01:31 --- join: TheBlueWizard (i=TheBlueW@ts001d0798.wdc-dc.xod.concentric.net) joined #forth 10:07:47 I don't think factor's image generation is actually that slow. 10:08:05 it's just that it's something that a user actually does occasionally. 10:08:22 IIRC generating a whole new squeak image takes a long time, for instance 10:10:14 tathi: they STILL mean an "image" akin to an "object"? 10:10:16 also, if you know that some file changed, you can just load that part of the source code into your current image 10:10:36 I don't think you actually NEED to rebuild the image for very many things 10:10:47 huh? 10:10:53 the usual meaning of image 10:11:21 instead of loading, relocating, and liking an executable, then running its initialization code, etc. 10:11:27 right 10:11:27 you just save the whole memory image of the running process 10:11:46 running system 10:12:06 They should be adjusting the "image", saving it, and then reloading the "image" and re-readjusting it. 10:12:35 like I said, I don't think you have to generate a new image from scratch very often 10:14:04 yeah, a binload should only be created when the CORE changes. 10:18:46 --- join: segher_ (n=segher@dslb-084-056-186-006.pools.arcor-ip.net) joined #forth 10:30:23 --- quit: segher (Read error: 110 (Connection timed out)) 10:55:35 --- quit: PoppaVic ("Pulls the pin...") 10:58:47 --- quit: neceve (Remote closed the connection) 11:03:08 what's the f script do? 11:05:04 f script? 11:13:30 factor has a program named f that you use to run images 11:13:39 f is the runtime 11:13:50 I was just wondering... if the image is machine code binary then what do you need ./f for? 11:14:04 iirc, it has the basic primitives and handles loading the image 11:14:28 I'm not sure if the image is a pure memory dump 11:37:13 --- part: TheBlueWizard left #forth 11:48:42 the image is a dump of the garbage collected data heap, and the code heap 12:41:23 --- quit: Cheery ("Leaving") 13:08:13 the image is garbage? 13:08:39 yes, just like everything you say 13:33:14 thanks my god 14:18:40 ... 14:18:42 lol 14:19:46 --- mode: ChanServ set +o crc 14:38:10 only gods can say that everything someone says is garbage, otherwise they would be garbage too. 14:45:38 that's garbage! 14:46:12 --- quit: warpzero ("Lost terminal") 14:46:19 --- join: warpzero (n=warpzero@wza.us) joined #forth 14:59:09 --- join: sproingie (n=chuck@64-121-2-59.c3-0.sfrn-ubr8.sfrn.ca.cable.rcn.com) joined #forth 15:09:06 --- quit: uiuiuiu (Remote closed the connection) 15:09:10 --- join: uiuiuiu (i=ian@dslb-084-056-250-154.pools.arcor-ip.net) joined #forth 15:16:57 --- quit: Snoopy42 (Read error: 145 (Connection timed out)) 15:26:31 --- quit: virl ("Verlassend") 15:32:48 --- join: neceve (n=Clau@unaffiliated/neceve) joined #forth 15:34:35 --- quit: sproingie ("Konversation terminated!") 15:46:11 --- join: Snoopy42 (i=snoopy_1@dslb-084-058-134-226.pools.arcor-ip.net) joined #forth 17:11:56 --- join: segher (n=segher@dslb-084-056-156-070.pools.arcor-ip.net) joined #forth 17:14:18 --- quit: segher_ (Read error: 110 (Connection timed out)) 17:30:21 --- quit: tathi ("leaving") 18:12:52 --- quit: uiuiuiu (zelazny.freenode.net irc.freenode.net) 18:12:52 --- quit: warpzero (zelazny.freenode.net irc.freenode.net) 18:12:52 --- quit: crc (zelazny.freenode.net irc.freenode.net) 18:12:53 --- quit: ccfg (zelazny.freenode.net irc.freenode.net) 18:12:53 --- quit: Jim7J1AJH (zelazny.freenode.net irc.freenode.net) 18:12:55 --- quit: Infight9 (zelazny.freenode.net irc.freenode.net) 18:12:55 --- quit: saon (zelazny.freenode.net irc.freenode.net) 18:12:55 --- quit: madgarden (zelazny.freenode.net irc.freenode.net) 18:13:11 --- join: crc (i=crc@pdpc/supporter/active/crc) joined #forth 18:13:11 --- join: uiuiuiu (i=ian@dslb-084-056-250-154.pools.arcor-ip.net) joined #forth 18:13:11 --- join: warpzero (n=warpzero@wza.us) joined #forth 18:13:11 --- join: Infight9 (i=WINNT@12-208-98-237.client.insightBB.com) joined #forth 18:13:11 --- join: saon (i=1000@unaffiliated/saon) joined #forth 18:13:11 --- join: madgarden (n=madgarde@Toronto-HSE-ppp3708221.sympatico.ca) joined #forth 18:13:11 --- join: ccfg (n=ccfg@dsl-roigw1-fe8ade00-21.dhcp.inet.fi) joined #forth 18:13:11 --- join: Jim7J1AJH (n=jim@221x115x224x2.ap221.ftth.ucom.ne.jp) joined #forth 18:13:11 --- mode: irc.freenode.net set +o crc 19:43:51 --- quit: neceve (Remote closed the connection) 20:19:48 --- join: Raystm2_ (n=Raystm2@adsl-69-149-51-195.dsl.rcsntx.swbell.net) joined #forth 20:35:08 --- quit: Raystm2 (Read error: 110 (Connection timed out)) 23:50:25 --- join: Cheery (i=Henri@a81-197-45-47.elisa-laajakaista.fi) joined #forth 23:59:59 --- log: ended forth/06.03.19