00:00:00 --- log: started retro/09.11.28 03:49:27 --- join: cavva (i=5895f651@gateway/web/freenode/x-spakycsmkmglcqlw) joined #retro 03:49:40 hi 03:50:22 I'm looking the retro ources 03:50:28 sources 03:50:42 hi cavva 03:50:49 hi 03:51:33 * crc is about to head to work, but should be back online within ~1 hour 03:51:37 i don't understand what does the code at line 105 of meta.retro 03:52:04 The : ; ;, ;; [ ? 03:52:09 yes 03:52:17 that redefines ; for the target compiler 03:52:55 ;. lays down code for a return instruction 03:53:03 ;; exits the word 03:53:10 [ turns the compiler off 03:53:28 great thanks 03:53:34 no problem 03:55:06 what i wasn't understand is why this word doesn't end with ; 03:55:12 ;; does the job 03:55:16 right? 04:00:27 * cavva going to lunch 04:42:31 --- join: crcx (i=d8012b82@gateway/web/freenode/x-jfuorsmtfnitjosp) joined #retro 04:43:55 at work now :) 04:47:15 cavva: ; is the same as ;; [ 04:57:20 yiyus: http://retroforth.org/go.diff - patch to read in the standard image files 04:57:32 yiyus: this is looking very nice :) 05:00:45 applied :) 05:02:08 i'm really having a lot of fun doing this 05:05:54 --- join: retro-commit (n=retro-co@sh1-ext.rs.github.com) joined #retro 05:05:54 ngaro: 3Charles Childers 07master0 SHA1-d72e9d5 05:05:54 golang: various fixes, load standard images 05:05:54 http://bit.ly/8sGjJI 05:05:55 --- part: retro-commit left #retro 05:06:46 * crc us starting to get a feel for go :) 05:54:05 --- quit: cavva ("Page closed") 08:16:00 yiyus: it's addictive, isn't it? :) 08:47:19 addictive = good :) 08:56:07 :) 09:03:08 one hour until I'm done work, then I get to spend a few hours with one of my kids :) 09:33:31 --- quit: crcx ("Page closed") 10:31:41 docl: yes, really addictive :) 12:41:51 --- quit: docl (Read error: 110 (Connection timed out)) 13:33:55 --- join: docl (n=luke@74-36-1-108.dr01.hmdl.id.frontiernet.net) joined #retro 14:25:10 home now :) 14:27:45 --- join: retro-commit (n=retro-co@sh1-ext.rs.github.com) joined #retro 14:27:45 retro10: 3Charles Childers 07master0 SHA1-e1f8162 14:27:45 image size for metacompiler now a constant 14:27:45 http://bit.ly/8LPrJO 14:27:46 --- part: retro-commit left #retro 14:32:03 --- join: retro-commit (n=retro-co@sh1-ext.rs.github.com) joined #retro 14:32:03 ngaro: 3Charles Childers 07master0 SHA1-a523c5e 14:32:03 sync against hg repo for golang 14:32:03 http://bit.ly/51AXYz 14:32:04 --- part: retro-commit left #retro 14:41:44 welcome back crc 14:42:13 I'm trying out your various other git repos. 14:43:05 I had to make some changes to rune to get it to compile. 14:43:58 what is the best way to publish my changes? should I make my own fork? 14:44:40 yes 14:44:52 then send me a pull request or patches if you want me to merge changes back 14:45:07 * crc is picky with the retro10 repo, but the others are all much looser :) 14:45:56 ok. how does a pull request work? 14:46:55 on github, there's a button that will send me an alert that you want me to pull from your fork 14:47:31 * crc has used this quite successfully with Mat2's extended-ngaro repository 14:53:13 there we go 14:55:40 the trouble with sh.c was that getline was defined elsewhere. I tried changing the name of getline to mygetline, but then I figured out that I could use a #define to change it for the included file instead of altering the source. 14:56:08 docl: changes merged 14:56:15 :) 14:58:31 I'm mostly developing on a mac now, so I sometimes miss some compatibility things :( 14:58:53 (except in retro, where I build/test on a good variety of systems on a regular basis) 15:00:42 rune is something I've dabbled with for a couple of years 15:02:24 * crc used to have it as a barebones bootable linux, 2.2 kernel + old retro + predecessor to rune as userland 15:09:49 I'm looking for a small kernel to use 15:11:12 nice. does it do its own file handling? 15:11:41 I've not found a suitable kernel for use with rune yet :( 15:12:27 hmm 15:15:44 xv6 looks interesting, but license status is unclear 15:16:47 the copy I have is missing a LICENSE document/statement :( 15:19:18 hmm, found a newer version 15:19:47 MIT-type license :) 15:20:13 * crc will have to get this into a git repo :) 15:20:28 nice 15:22:26 http://github.com/crcx/xv6/ 15:23:54 2 15:23:55 doh 15:28:39 nice! 15:33:56 * docl is getting a copy of qemu to try this out with 15:43:33 ok, I have it building with bochs and qemu, but it says there is no bootable device. 15:46:40 * crc is having problems with git on his linux box :( 15:47:52 darn 15:47:57 does it work on windows? 15:48:02 mac, I mean 15:48:28 I'd have to build a cross compiler with ELF support for the mac 15:48:42 ah ok 15:49:04 does qemu run on mac? 15:49:14 yes 15:54:19 fixed by upgrading ssh on the linux box :) 15:54:37 cool 15:58:49 I'm getting errors building the apps 15:58:58 but the kernel appears to build ok 16:04:02 if I run make xv6.img twice after running make fs.img it finishes. 16:05:23 the img files produced do not seem to be bootable though 16:26:13 I finally got it to build an image, getting ready to test 16:31:12 won't boot for me either 16:32:23 darn 16:32:44 try: "make qemu" 16:33:22 after make clean: 16:33:23 ./sign.pl bootblock 16:33:23 boot block too large: 580 bytes (max 510) 16:33:53 I wonder if there is a purpose for that 510 byte limitation. 16:34:17 512 bytes = 1 sector, 2 bytes are reserved for boot sector signature 16:34:42 ah 16:34:54 can we go to double the size or something? 16:35:05 nope 16:35:13 it's a physical limitation of x86 hardware 16:40:27 the BIOS will only load one sector from a disk; that sector has to load either the kernel, or a larger second stage loader 16:42:38 is this a 1-stage kernel? 16:42:46 I'm not sure :( 16:45:17 http://retroforth.org/fs.img 16:45:25 http://retroforth.org/xv6.img 16:45:50 qemu -m64 -std-vga -localtime -parallel stdio -hdb fs.img -hda xv6.img 16:45:56 that works for me 16:49:08 cool 16:50:37 is that from an earlier version? 16:51:27 no, trial and error on the qemu command line 17:03:57 how did you get the images to compile? 17:04:42 make; make qemu 17:07:36 do you get the error about the boot block being too large when you run make? 17:07:54 no 17:08:21 ah, that explains it. I wonder why I am getting it. 17:08:41 bug in binutils or gas probably 17:13:20 maybe it's because I don't have gdb installed 17:21:35 gas is part of gcc right? 17:21:57 I think so 17:22:32 * crc pushed an update to move the apps out of the main directory 17:40:54 --- quit: docl (Read error: 110 (Connection timed out)) 18:04:48 --- join: docl (n=luke@74-36-211-47.dr01.hmdl.id.frontiernet.net) joined #retro 18:06:52 wb 18:12:56 I'm trying to pull the update into my fork repo 18:13:43 git pull git://github.com/crcx/xv6.git 18:14:16 git pull git://github.com/crcx/xv6.git master 18:14:37 one of those *should* work 18:17:11 I think I got it :) 18:17:31 looks like it :) 18:19:39 I used the web interface 18:22:02 nothing wrong with that 18:33:13 --- join: retro-commit (n=retro-co@sh1-ext.rs.github.com) joined #retro 18:33:13 ngaro: 3Charles Childers 07master0 SHA1-020508a 18:33:13 update MIDP initial image 18:33:13 http://bit.ly/6Sdmq6 18:33:14 --- part: retro-commit left #retro 18:55:20 docl: I pushed another set of updates moving the userspace libraries to a separate directory and keeping the makefile from deleting the .img files with a "make clean" 19:01:19 cool 19:01:22 got them 19:02:11 good 20:07:14 --- quit: docl (Connection timed out) 20:48:37 --- join: docl (n=luke@67-60-48-82.cpe.cableone.net) joined #retro 20:48:56 finally got it to compile properly 20:54:46 I had to compile for the specific archetecture. perhaps it is because I am running on 64 bit hardware. 21:53:09 in retro, how would I script keystrokes? 21:56:19 --- quit: docl ("Lost terminal") 22:00:19 --- join: docl (n=luke@67-60-48-82.cpe.cableone.net) joined #retro 23:59:59 --- log: ended retro/09.11.28