00:00:00 --- log: started forth/15.03.25 00:41:15 --- join: pgomes (~pgomes@ip923476d3.dynamic.kabel-deutschland.de) joined #forth 00:49:07 --- quit: atommann (Ping timeout: 265 seconds) 00:55:16 --- quit: pgomes (Ping timeout: 246 seconds) 00:59:00 --- join: ASau (~user@46.115.27.197) joined #forth 01:01:13 --- join: atommann (~atommann@58.251.2.94) joined #forth 01:25:27 --- join: impomatic (~chatzilla@61.205.125.91.dyn.plus.net) joined #forth 02:41:02 --- join: true-grue (~grue@95-27-141-28.broadband.corbina.ru) joined #forth 03:13:44 --- join: nighty-_ (~nighty@hokuriku.rural-networks.com) joined #forth 03:31:01 --- join: impomatic_ (~chatzilla@61.205.125.91.dyn.plus.net) joined #forth 03:33:40 --- quit: impomatic (Ping timeout: 250 seconds) 03:33:41 --- nick: impomatic_ -> impomatic 03:46:11 --- join: mnemnion (~mnemnion@c-98-210-219-91.hsd1.ca.comcast.net) joined #forth 03:50:40 --- quit: mnemnion (Ping timeout: 246 seconds) 04:04:30 --- quit: atommann (Ping timeout: 264 seconds) 05:02:56 --- join: vsg1990 (~vsg1990@cpe-67-241-148-119.buffalo.res.rr.com) joined #forth 06:12:23 --- join: Zarutian (~Adium@168-110-22-46.fiber.hringdu.is) joined #forth 06:17:37 --- join: kumul (~mool@adsl-64-237-236-59.prtc.net) joined #forth 06:29:13 --- join: atommann (~atommann@222.248.72.156) joined #forth 06:34:10 --- join: johnmark_ (~johnmark@c-73-51-235-57.hsd1.il.comcast.net) joined #forth 07:19:03 --- join: fantazo (~fantazo@213.129.230.10) joined #forth 07:20:12 --- join: mnemnion (~mnemnion@c-98-210-219-91.hsd1.ca.comcast.net) joined #forth 07:49:19 --- quit: mnemnion (Read error: Connection reset by peer) 07:49:43 --- join: mnemnion (~mnemnion@c-98-210-219-91.hsd1.ca.comcast.net) joined #forth 08:53:43 --- quit: luptenschteiner (Ping timeout: 264 seconds) 09:08:46 --- quit: darkf (Quit: Leaving) 09:11:35 --- quit: atommann (Quit: Leaving) 09:15:33 --- quit: johnmark_ (Quit: Leaving) 09:20:24 --- join: probonono (~User@ppp103-111.static.internode.on.net) joined #forth 09:20:24 --- quit: probonono (Changing host) 09:20:24 --- join: probonono (~User@unaffiliated/probonono) joined #forth 09:20:32 --- quit: fantazo (Quit: Verlassend) 09:38:36 --- join: johnmark_ (~johnmark@c-73-51-235-57.hsd1.il.comcast.net) joined #forth 10:09:46 --- quit: mnemnion (Remote host closed the connection) 10:49:53 --- quit: impomatic (Read error: Connection reset by peer) 11:17:04 --- join: mnemnion (~mnemnion@2601:9:4300:438:10dd:f629:ee9c:fca1) joined #forth 11:25:01 --- join: pgomes (~pgomes@ip923476d3.dynamic.kabel-deutschland.de) joined #forth 11:50:09 yunfan: Were you asking about MIPS microAptiv? 11:51:10 --- quit: pgomes (Ping timeout: 255 seconds) 11:51:48 It is used in the PIC32MZ series CPUs. 12:25:06 --- join: impomatic (~chatzilla@61.205.125.91.dyn.plus.net) joined #forth 12:40:37 --- quit: Zarutian (Quit: Leaving.) 13:25:12 --- join: Mat4 (~claude@ip188619bb.dynamic.kabel-deutschland.de) joined #forth 13:25:16 hello 13:26:10 --- join: MrMobius (~Joey@c-68-45-16-225.hsd1.nj.comcast.net) joined #forth 13:32:47 Hi Mat4 13:37:44 --- quit: MrMobius (Quit: HydraIRC -> http://www.hydrairc.com <- Po-ta-to, boil em, mash em, stick em in a stew.) 13:40:20 hey 13:56:26 --- quit: mnemnion (Remote host closed the connection) 14:02:24 --- quit: vsg1990 (Quit: Leaving) 14:08:24 hi bluekelp 14:08:34 and backer 14:10:14 I've just found coreboot as good alternative to UEFI (mainly because there exist a GRUB version for it) 14:27:31 --- join: pgomes (~pgomes@ip923476d3.dynamic.kabel-deutschland.de) joined #forth 14:43:35 --- quit: Mat4 (Quit: Verlassend) 15:06:41 --- quit: nighty-_ (Quit: Disappears in a puff of smoke) 15:13:34 --- quit: true-grue (Read error: Connection reset by peer) 15:31:27 --- join: nighty-_ (~nighty@hokuriku.rural-networks.com) joined #forth 15:40:16 --- join: vsg1990 (~vsg1990@cpe-67-241-148-119.buffalo.res.rr.com) joined #forth 15:49:58 --- quit: bjorkintosh (Ping timeout: 256 seconds) 16:14:48 --- join: mnemnion (~mnemnion@c-98-210-219-91.hsd1.ca.comcast.net) joined #forth 16:20:25 --- quit: nighty-_ (Ping timeout: 246 seconds) 17:53:54 --- quit: irsol (Ping timeout: 255 seconds) 17:55:33 --- quit: pgomes (Ping timeout: 264 seconds) 17:56:14 --- join: irsol (~irsol@unaffiliated/contempt) joined #forth 18:12:47 --- join: MrMobius (~Joey@c-68-45-16-225.hsd1.nj.comcast.net) joined #forth 18:18:49 --- quit: vsg1990 (Quit: Leaving) 18:19:52 --- join: saml_ (~saml@cpe-24-102-97-97.nyc.res.rr.com) joined #forth 18:20:44 --- join: xyh (~xyh@2001:250:3002:5550:6ea1:cc0f:bcb2:b187) joined #forth 18:24:01 backer: yes 18:24:29 backer: looks like the chip is too expensive campared to arm series 18:52:53 --- quit: xyh (Remote host closed the connection) 18:54:13 --- join: xyh (~xyh@2001:250:3002:5550:6ea1:cc0f:bcb2:b187) joined #forth 18:58:49 --- join: darkf (~darkf___@unaffiliated/darkf) joined #forth 19:05:52 --- quit: proteusguy (Ping timeout: 265 seconds) 19:18:03 --- join: proteusguy (~proteusgu@ppp-110-168-229-137.revip5.asianet.co.th) joined #forth 19:18:03 --- mode: ChanServ set +v proteusguy 19:56:49 --- join: ASau` (~user@46.114.150.159) joined #forth 20:01:00 --- quit: ASau (Ping timeout: 272 seconds) 20:02:48 --- join: atommann (~atommann@58.251.2.94) joined #forth 20:11:27 --- quit: saml_ (Quit: Leaving) 20:17:13 --- nick: ASau` -> ASau 21:00:53 --- quit: johnmark_ (Quit: Leaving) 21:15:27 yunfan: Really? Compared to what in the arm series? 21:18:12 cortex m? 22:20:16 Cortex-M4? 22:21:03 m3? 22:21:11 i've no idea 22:21:23 maybe its just the dev-board price too high 22:21:42 Here's an STM32F427 @ $17.33 in single qty, 256K SRAM and 2M FLASH, 168MHz: http://www.digikey.com/product-detail/en/STM32F427VIT6/497-14049-ND/4357419 22:22:34 Here's a PIC32MZ2048 @ $12.95 single qty, 512K SRAM and 2M FLASH, 200MHz: http://www.digikey.com/product-detail/en/PIC32MZ2048ECG064-I%2FPT/PIC32MZ2048ECG064-I%2FPT-ND/4500837 22:23:26 $9.89 in 25 qty. ST is $11.42 in 500 qty. 22:23:57 Though maybe there are other cortexes that are more bang for the $... 22:24:19 Why were you interested in microAptiv? 22:25:23 --- quit: xyh (Ping timeout: 256 seconds) 22:25:44 backer: because i'd like to try mips's new design 22:26:15 i like its concise designs 22:26:30 backer: also what you talk is the chip price? 22:26:33 I am aiming to release a Forth for 32-bit MCUs (currently PIC32) within a few weeks. But I have yet to look into microAptiv... 22:26:38 i am refer to dev-board :[ 22:27:17 backer: do you have some optimizing trick on mips? 22:27:40 yunfan: It is trivial to build your own board if you have modest soldering skills. I have built several around the PIC32s, both in DIP and TQFP packages. 22:27:43 backer, a forth for PIC32? that will be cool. 22:28:17 TQFP is easy to solder. 22:28:22 backer: but i dont have soldering skills. i am software guys :] 22:28:43 It has been running very well using standard MIPS32r2. My only optimizing "trick" is to use subroutine threading and extensive inlining. 22:28:47 yunfan, just watch some youtube videos and practice. 22:28:56 atommann: maybe i should join some hackerspace activities for learning that skill :] 22:29:18 backer, your PIC32Forth is written in C or Assembler? 22:29:21 yunfan: Now that I think about it, those STM32F4 dev boards are dirt cheap... 22:29:37 yunfan, trust me, it's not as hard as you think. 22:29:54 backer: yep, arm does well on marketing 22:30:05 atommann: will try that later 22:30:32 atomann: It is written in all assembler. I started working on a native C kernel for use under Linux, but that is a lower priority. 22:31:03 backer, I see. 22:31:51 also PIC32MX[12]xx are available in 28-pin DIP for the solder-phobic 22:31:54 http://armpit.sourceforge.net/ is also in assembler. 22:32:08 backer: since mips has so many generel registers, could forth use them as the cache of stack? 22:33:04 i dont like arm, the official's document are full of c examples 22:33:09 atomann: Wow, thanks for that link! 22:33:47 but these embed domain c code are full of port definiton, initialzation and entry register 22:34:52 The bulk of the Forth code (the "colon" words) are human-readable Forth that I am translating into assembler. That is then assembled along with the rest of the kernel bits. 22:36:34 yunfan: I haven't really figured out a good way to use registers as a stack cache. I keep the top of both stacks in registers. 22:39:15 backer: mips has 32 generel registers, you might need 8 or 12 as abi or other stuffs, the rest could be a shadow of stack 22:39:48 but the problem is do they support address register file by ofset? 22:39:54 looks like avr support that 22:40:42 No, not that I know of. 22:42:27 I don't need to reserve registers for MIPS ABI. If I want to make a call into some C routine (which I can do), I just save my registers before calling. 22:43:38 backer: memory save/load cost much time 22:44:30 yunfan: It depends on how often you have to call something written in another language. Presently I'm using it for an embedded application that is 100% Forth and assembly. There is no saving going on there. 22:45:58 But I built a version for Linux/MIPS userspace and was able to make calls into libc without a problem. 22:47:08 --- quit: kumul (Ping timeout: 246 seconds) 22:48:38 backer: i am more interesting of using forth as a embed script language for high performance service, this require to communicate to other libc based libraries 22:50:26 yunfan: I'm using six registers now, plus temporaries t0, t1, t2... 22:51:01 backer: how much ram do your forth cost when booting 22:51:28 I bet you even in that case that saving/restoring several registers would be entirely negligible. 22:52:29 i agree 22:57:33 Have to check about RAM usage to be sure. AVAIL shows I am using 7k on startup, but I know I reserved a 4k buffer in there, and there may be some other buffers that are larger than need be. 22:58:19 Code compiled into flash runs out of flash, and newly compiled code goes into SRAM (though can be appended to the flash dictionary for persistent storage). 23:02:01 Has anyone seen RodgerTheGreat? 23:03:01 and in the 7k totally ram usage, what features does your forth support? 23:03:14 it looks like your forth cost lesser than a vanilla lua vm 23:06:17 Well basically it is based on eForth. So if you take a look at what eForth provides as a language, add platform support and a dictionary that spans RAM and FLASH - that's what you (will) get. 23:06:41 And there's support for writing interrupt routines in Forth. 23:14:00 then what about its performance? 23:14:21 have you ever got a x86 version for camparing to gforth or others 23:16:19 Not yet. I have not done any real benchmarking, but a lot of tweaking to improve speed. Branches are inlined when compared to stock eForth (e.g. no calls to QBRAN, BRAN, DONEXT, etc.). So an empty FOR-NEXT loop actually runs at half the pipeline speed (on the PIC32 anyway). 23:17:27 Subroutine threading provides copious opportunities to inline and optimize since machine code can be interspersed with Forth calls. 23:18:01 are there any non-ansi forth which focus on speed? 23:18:57 I really don't know. I am not a Forth expert. I am just using it for embedded applications. 23:19:33 But I love tweaking and optimizing, so I couldn't help but tackle the low-hanging fruit. 23:21:44 Hopefully soon I will have an indirect-threaded, portable Unix / C kernel. I will be very interested to test its performance against other Forths on the PC. But I'm not writing that one for performance... 23:22:40 (I am using the term "kernel" in the sense of eForth; it comprises a minimal set of Forth words, implemented in assembly, upon which the rest of the VM is built.) 23:26:35 Implemented in assembly or anything else, actually. 23:59:59 --- log: ended forth/15.03.25