| 01:33:32 | scanline | is away: food-like thingies... |
| 01:59:16 | Sammy | hi all |
| 02:18:19 | scanline | is back (gone 00:44:47) |
| 02:35:59 | | andersee was last seen on #tuxscreen 7 hours, 29 minutes and 51 seconds ago, saying: is away: I'm busy [Wed Mar 20 19:06:08 2002] |
| 02:35:59 | jacques | ibot: seen andersee |
| 02:36:04 | jacques | :-( |
| 03:22:26 | | andersee was last seen on #tuxscreen 8 hours, 16 minutes and 18 seconds ago, saying: is away: I'm busy [Wed Mar 20 19:06:08 2002] |
| 03:22:26 | TomW_ | ibot: seen andersee |
| 03:44:38 | scanline | is away: meeting |
| 04:10:22 | scanline | is back (gone 00:25:44) |
| 04:10:36 | lebouf | whats tuxscreen? |
| 04:11:04 | | hmmm... tuxscreen is http://TuxScreen.net/ order at http://tuxscreen.net/wiki/view/Users/TimRiker and http://tuxscreen.net/scrapbook/images.html or #TuxScreen |
| 04:11:04 | sjm | ibot: tuxscreen |
| 06:57:56 | andersee | is back (gone 11:51:47) |
| 10:28:15 | andersee | uClibc 0.9.10 released.... |
| 10:33:38 | mallum | cool |
| 10:33:52 | andersee | I just updated buildroot-tux to use it |
| 10:38:35 | andersee | K, http://uclibc.org/ now has the announcement. |
| 15:15:05 | scanline | is away: food or something |
| 15:40:23 | Morn | is away: foraging for food... |
| 16:09:59 | RedBaron | Tangent: Howdy |
| 16:30:19 | andersee | CosmicPenguin: you around? |
| 16:34:30 | CosmicPenguin | andersee: I am here |
| 16:34:40 | CosmicPenguin | is fighting with the IR UART on the Geode |
| 16:41:45 | andersee | CosmicPenguin: The latest buildroot now uses uClibc 0.9.10, which provides a g++ wrapper and supports ctors/dtors |
| 16:42:06 | andersee | CosmicPenguin: Does your C++ stuff need libstdc++? |
| 16:43:49 | CosmicPenguin | andersee: yep |
| 16:43:54 | CosmicPenguin | andersee: it also needs pthreads, though |
| 16:44:28 | CosmicPenguin | I am going to try the new uclibc here in a few hours on a buildroot I am working on at work.... |
| 16:44:29 | andersee | Guess we should enable uClibc's libpthread then... |
| 16:45:05 | andersee | For libstdc++, you'll need the full gcc toolchain |
| 16:45:29 | CosmicPenguin | Doh! |
| 16:45:46 | andersee | Or, you could try to use stlport. I've had one report of someone using it sucessfully now that I have ctors/dtors |
| 16:46:04 | CosmicPenguin | I'm not sure if our stuff is ready to move to gcc 3.00+ |
| 16:46:10 | andersee | http://www.stlport.org/ |
| 16:47:18 | CosmicPenguin | you know, now that I think about it, I never use the STL, but libstdc++ still needs to be compiled in... whassup with that? |
| 16:47:49 | andersee | iostreams? |
| 16:48:03 | CosmicPenguin | nope |
| 16:48:06 | CosmicPenguin | not me |
| 16:48:16 | CosmicPenguin | maybe ViewML uses them, though |
| 16:48:18 | andersee | By default, g++ w/glibc's _always_ links libstdc++, needed or not |
| 16:48:37 | andersee | Try just using the uClibc wrapper then. Might work... |
| 16:48:39 | CosmicPenguin | so uclibc will ignore it if I don't need it |
| 16:48:47 | CosmicPenguin | Woo-hoo! |
| 16:49:11 | andersee | BTW, that stlport lib is opensource and provides a libstdc++ implementation... |
| 16:49:27 | CosmicPenguin | is downloading it now |
| 16:51:52 | CosmicPenguin | man, it is downright beautiful outside |
| 17:04:12 | Aelhaz | hmm, buildroot not workin |
| 17:04:15 | Aelhaz | arm-linux-gcc -Wall -Os -fstrict-aliasing -fno-builtin -fPIC -c initfini.c -S -o initfini.s |
| 17:04:15 | Aelhaz | initfini.c:39: stdlib.h: No such file or directory |
| 17:04:15 | Aelhaz | make[5]: *** [initfini.s] Error 1 |
| 17:14:51 | andersee | Aelhaz: that is very odd. |
| 17:15:22 | andersee | Aelhaz: regardless, just edit libc/sysdeps/linux/common/initfini.c and remove the line '#include <stdlib.h>' |
| 17:15:32 | Aelhaz | ok |
| 17:15:33 | andersee | it really isn't needed |
| 17:15:46 | andersee | Odd though. It worked for me... |
| 17:16:20 | Aelhaz | yeah |
| 17:16:54 | Aelhaz | i have a stdlib.h in /usr/include and in build/uClibc-0.9.10/include |
| 17:16:58 | Aelhaz | *shrug* |
| 17:17:37 | andersee | Where did your arm-linux-gcc come from? |
| 17:18:00 | Aelhaz | emdebian |
| 17:18:08 | Aelhaz | running on woody |
| 17:19:04 | andersee | me too.... |
| 17:19:20 | andersee | but it worked for me... |
| 17:19:26 | Aelhaz | *shrug* |
| 17:20:05 | Aelhaz | oh, off hand do you know if anyone's built openssl & openssh against uclibc? |
| 17:21:51 | Aelhaz | tho, I don't honestly believe they'd fit on the tux, anyway . . . I'm turning a tux into a firewall, an ssh daemon would be handy |
| 17:22:31 | Aelhaz | but not critical |
| 17:24:55 | andersee | I know someone built lsh |
| 17:25:04 | andersee | http://www.net.lut.ac.uk/psst/ |
| 17:26:28 | Aelhaz | i thought that was just a client |
| 17:28:23 | andersee | perhaps so. dunno, never tried it |
| 17:28:49 | Aelhaz | *nod* |
| 17:30:18 | Aelhaz | I'm hoping to replace my noisy heat spewing dec alpha firewall. figure bare minimum i need ipchains, boa. sshd would be desirable, socks5 would be nice, tho recent openssh can pretend to be socks5 |
| 17:31:19 | RedBeard | andersee: did you get the parts and pieces for the 5v hack from RatShack? |
| 17:31:42 | RedBeard | what are the specs on the diods you used? |
| 17:31:55 | Aelhaz | were there parts and pieces? looked like three chunks of wire and a bank of dip switches |
| 17:32:11 | Aelhaz | and the switches were only there to manually switch back to 3.3v operation |
| 17:32:32 | RedBeard | so it's switches or diods but not both? |
| 17:32:36 | andersee | RedBeard: didn't use diodes |
| 17:32:43 | andersee | RedBeard: just a switch |
| 17:32:47 | RedBeard | is probably confused between the two sets of pix |
| 17:32:55 | andersee | See the pics from codepoet.org |
| 17:33:07 | RedBeard | goes to look again |
| 17:33:09 | andersee | though they are a bit fuzzy... |
| 17:33:49 | RedBeard | the other instructions showed diods being used |
| 17:34:01 | Aelhaz | could use jumpers instead of switches, and just hot glue them to the board |
| 17:34:18 | andersee | Nah, using the diodes is silly, IMHO |
| 17:34:30 | andersee | Someone needs to update those instructions |
| 17:34:43 | RedBeard | andersee: I'm hoping to do that today/tomorrow |
| 17:34:50 | RedBeard | that's why I'm asking questions right now |
| 17:35:04 | RedBeard | alhaz and I were talking about doing it |
| 17:35:07 | jacques | hi andersee |
| 17:37:57 | andersee | hey |
| 17:39:54 | Aelhaz | guh, boa doesn't like to be configured to cross compile . . . |
| 17:40:04 | Aelhaz | checking whether setvbuf arguments are reversed... configure: error: can not run test program while cross compiling |
| 17:40:12 | Aelhaz | and ./configure stops dead right there |
| 17:40:15 | jacques | oh i love those |
| 17:40:16 | Aelhaz | anyone have a guess how to get around that? |
| 17:40:24 | jacques | lots of programs are like that |
| 17:40:52 | jacques | i have been known to manually hack the config.cache file |
| 17:41:02 | jacques | but you have to know what to put in there |
| 17:41:20 | jacques | usually I just copy or modify what's in there when configured for x86 |
| 17:42:14 | jacques | at least for the parts that want to run a test program |
| 17:42:57 | Aelhaz | I'm cheating, ran ./configure with CC=gcc and edited the makefile, maybe it'll work |
| 17:43:12 | Aelhaz | Sorta looks like it worked |
| 17:44:02 | Aelhaz | and the stripped binary is 64k. Neat. |
| 17:47:32 | mallum | Aelhaz: you'll can hack the configure.in file and remove the test |
| 17:48:46 | Aelhaz | mallum: I'll do that if my binary doesn't work |
| 17:48:52 | RedBeard | prpplague: linkup pix for you right here: http://www.xmission.com/~redbeard/linkup/index.html |
| 17:49:30 | mallum | Aelhaz: dont spose you have any exp with powerpc and Linux ? |
| 17:50:24 | CosmicPenguin | andersee: does uclibc support getpt() and grantpt()? |
| 17:51:46 | Aelhaz | mallum: 'course i do |
| 17:52:23 | mallum | Aelhaz: I got a G4 tibook yesterday |
| 17:52:31 | mallum | Aelhaz: the thing is a bitch |
| 17:53:15 | Aelhaz | Neat |
| 17:53:26 | Aelhaz | the PPC stuff I've used is primarily embedded, tho |
| 17:53:35 | mallum | Aelhaz: yeah I guessed it might be |
| 17:53:56 | RedBeard | mallum: I've heard good things about yellowdog linux, but I have no idea how current it is |
| 17:54:02 | mallum | Aelhaz: so I've got this gorgeous looking expensive machine with no X and no pcmcia |
| 17:54:12 | Aelhaz | mallum: ah |
| 17:54:17 | Aelhaz | mallum: and only one mouse button |
| 17:54:20 | Aelhaz | :) |
| 17:54:24 | mallum | Aelhaz: yep |
| 17:54:33 | Aelhaz | that would drive me insane |
| 17:54:35 | RedBeard | who needs more than one button? any more and you'd confuse the user |
| 17:54:38 | Aelhaz | I use all three regularly |
| 17:55:01 | mallum | Aelhaz: I have X up on top of the framebuffer, but the display is about an inch to the left |
| 17:55:07 | Aelhaz | hmm |
| 17:55:09 | RedBeard | I wish I could get the fourth button to do something |
| 17:55:16 | mallum | Aelhaz: Im now building X 4.2 and praying |
| 17:55:38 | Aelhaz | *nod* |
| 17:57:01 | jacques | i thought the tibooks had 3 buttons??? |
| 17:57:16 | Aelhaz | openssl appears to be happily xcompiling |
| 17:58:01 | Aelhaz | I think the trick with openssh is to make it link openssl statically, so as to avoid putting like 10 megs of every crypto known to man on the target . . |
| 17:58:13 | Aelhaz | ack! |
| 17:58:16 | andersee | CosmicPenguin: yes |
| 17:58:44 | andersee | CosmicPenguin: The behavior depends a bit on UNIX98PTY_ONLY and ASSUME_DEVPTS |
| 17:58:57 | Aelhaz | ok, the compile bombed out, i might have overlooked something tho . . |
| 18:00:47 | Aelhaz | nope, wasn't that. talk about a crash&burn. |
| 18:01:06 | Aelhaz | oh, i think i'm missing some stuff it has to link against . . or something . . |
| 18:01:44 | Aelhaz | hmm. |
| 18:02:29 | CosmicPenguin | andersee: ok, thanks... |
| 18:25:21 | CosmicPenguin | Anyone willing to admit that they are an X expert? |
| 18:25:59 | CosmicPenguin | I've got a quick question |
| 18:33:33 | jacques | i admit that I'm not |
| 18:33:53 | Aelhaz | what sort of question |
| 18:39:57 | CosmicPenguin | I want to know if it is possible to inject data from an incoming pty (for example, a remote control) into X11 for use by other applications |
| 18:41:03 | CosmicPenguin | Or do I have to write a kernel module for that purpose |
| 18:48:39 | Aelhaz | you can send x events to any window |
| 18:48:55 | Aelhaz | see the irxevent util in lirc for an example |
| 18:56:23 | Aelhaz | Neat, boa works |
| 18:57:25 | CosmicPenguin | Aelhaz: but for irxevent, you need to know a specific window name right? |
| 18:57:54 | Aelhaz | CosmicPenguin: True |
| 18:58:16 | Aelhaz | so you're saying you want something more like a keyboard driver |
| 18:58:33 | CosmicPenguin | Yeah, but I really want to be like lirc and keep it all in user space. |
| 18:58:45 | Aelhaz | hmm |
| 18:58:56 | Aelhaz | maybe you could write an xinput driver, or maybe something like irpty would work |
| 18:59:03 | Aelhaz | I don't know that much about X |
| 19:00:09 | CosmicPenguin | yeah, I like irpty, but that works by forking the app and slamming code into stdin.... |
| 19:00:32 | Aelhaz | thought so |
| 19:00:51 | CosmicPenguin | I guess I'll do the kernel module.... |
| 19:01:49 | Aelhaz | i think that's what i've generally seen used |
| 19:01:57 | Aelhaz | tho xinput should work |
| 19:05:09 | CosmicPenguin | Hehe.... Just to make things tougher, I want to use this with TinyX, so I don't know how well xinput will work... :) |
| 19:05:21 | CosmicPenguin | Anyway, I'll do the kernel module... thanks for your help, though... |
| 19:06:26 | Aelhaz | ah |
| 19:06:52 | Aelhaz | I'm just trying to goad anyone into writing an xinput driver for anything out of the ordinary, because there don't seem to be many |
| 19:07:34 | andersee | goes to get some sleep |
| 19:07:39 | Aelhaz | in another channel I've got the guy who wrote spellguard, an as-you-type spell checker that predates microsoft's patent by about 4 years |
| 19:07:39 | andersee | is away: I'm busy |
| 19:08:18 | Aelhaz | he wrote an os2 version and then a windows version, and says he'd write a linux version but he doesn't have a way to attach to keyboard input in X |
| 19:08:41 | Aelhaz | but i couldn't convince him to look at the standard xinput keyboard driver and see if he could just hack that up |
| 20:09:03 | scanline | is back (gone 04:53:58) |
| 20:11:43 | scanline | is away: trying once again to go postal |
| 20:27:47 | collord | so who wanted C++ in uclibc and what magic are they doing with it now? |
| 20:28:38 | jacques | i want it |
| 20:28:42 | jacques | but it's not ready yet |
| 20:28:48 | jacques | sorry |
| 20:28:55 | jacques | i mean the toolchain with libstdc++ is not ready yet |
| 20:29:02 | collord | ah |
| 20:29:13 | jacques | can't do too much without it |
| 20:29:25 | jacques | almost everything uses libstdc++ |
| 20:30:18 | collord | you building a gui app then? |
| 20:30:34 | jacques | there are several things i wanna build |
| 20:30:38 | jacques | such as fltk |
| 20:54:33 | mallum | CosmicPenguin: see XTest for faking events, much better than XSendEvent |
| 20:54:46 | mallum | Aelhaz: got X up on the tibook :) |
| 21:09:55 | CosmicPenguin | mallum: Great, now that I have the whole kernel module written.... :) |
| 21:11:13 | CosmicPenguin | mallum: hmmm... that seems to be what I want |
| 21:14:08 | CosmicPenguin | mallum: as long as I don't map a window, it won't pop up on screen, right? |
| 21:25:10 | mallum | CosmicPenguin: yep |
| 21:25:28 | mallum | CosmicPenguin: sorry I didn't catch you que earlier |
| 21:26:19 | mallum | CosmicPenguin: one thing to watch with Xtest is you have to remember to send the opposite event ( eg always send an up and a down click ), or poor little X will get all confused |
| 21:39:52 | akula | hello. |
| 21:39:57 | akula | I have a question |
| 21:40:03 | Aelhaz | ok |
| 21:40:32 | akula | has anyone experienced a problem with the tux where it appears that you've successfully flashed a new bootloader, but it doesn't work |
| 21:40:37 | akula | I have done it before with success. |
| 21:40:51 | akula | but I tried to re-inferno the thing and it appears that it was a bad image. |
| 21:41:03 | akula | can the flash get so badly hosed that it won't boot anymore. |
| 21:41:04 | akula | ? |
| 21:41:11 | akula | this is with the JTAG |
| 21:41:26 | Aelhaz | I'd make sure that you properly unlocked the flash blocks |
| 21:41:35 | akula | they were unlocked. |
| 21:42:00 | akula | no way to tell now, as it won't boot - but I don't think it would re-lock itself. |
| 21:42:01 | Aelhaz | well, the *contents of the flash can get completely hosed, but even in that state you can reload everything from jtag |
| 21:42:07 | akula | hmm. |
| 21:42:12 | Aelhaz | needs 12v to lock itself, so it can't |
| 21:42:16 | akula | I know it's a good bootloader (blob) image. |
| 21:42:36 | Aelhaz | maybe your parallel port isn't appropriate or isn't set up properly |
| 21:42:44 | akula | and the jflash process reports back that all is good. |
| 21:43:07 | akula | I have two tuxs. I've reflashed both -just one doesn't want to boot anymore. |
| 21:43:33 | akula | could a corrupt inferno image have damaged the flash? |
| 21:43:44 | Aelhaz | i don't really know, but i doubt it |
| 21:43:57 | akula | that's what I think too, but this has me stumped. |
| 21:49:49 | Aelhaz | bbl, gotta power down to fiddle with hardware |
| 21:50:00 | akula | thanks for the help |
| 22:15:08 | LoneTech | looks like the shipment estimates are a bit optimistic.. |
| 22:15:38 | LoneTech | my tuxscreens are in sweden now, but not yet delivered. maybe next week.. will phone the transport company tomorrow. |
| 22:22:18 | RedBeard | is anyone here, that's done the 5v pcmcia hack? |
| 22:45:34 | prpplague | well did any of you lineo gurus get the axe? |
| 22:57:23 | scanline | is back (gone 02:45:40) |
| 22:59:05 | prpplague | RedBeard: well have ya still got a job at lineo? |
| 23:01:44 | RedBeard | prpplague: naah, I took those pix as I was packing my things |
| 23:03:16 | prpplague | RedBeard: lol |