00:00:06 | Niddix | Of course. Why can't mount tell the difference between iso9660 and udf automatically... |
00:00:48 | BZFlag | Niddix: most discs are hybrids. they are _both_ iso9660 and udf |
00:01:07 | Niddix | I know.. it just makes life difficult. |
00:01:27 | Niddix | Or at least if you try and mount udf it drops back to is9660 if it can't. |
00:02:06 | BZFlag | hmm. there must be a way to get that. have you tried building both fs as modules and changing the mod load order? |
00:03:19 | Niddix | Both are modules. |
00:03:35 | Niddix | But my fstab entry lists udf |
00:04:09 | BZFlag | mod probe udf and rmmod iso, put auto in the fstab. what happens? |
00:04:28 | Niddix | Ah well.. hrmmm.. auto.. didn't think of that |
00:05:42 | GoRK | russ,andersee,bzflag: ok whatever i did, it works now... i think it was just me being stupid with devfs |
00:06:52 | Niddix | BZFlag: Still mounting it iso |
00:09:16 | Niddix | I think I'm in love with my Mr Flash CF. |
00:11:00 | GoRK | nobody would listen to my warning about mister flash |
00:11:25 | GoRK | mister flash will try to seduce you. you must resist |
00:11:31 | Niddix | ;) |
00:11:34 | Niddix | It was $25.. |
00:11:49 | Niddix | Nearly 1/3 the price of the one I bought for my camera. |
00:14:37 | Niddix | I am powerless to its guile. |
00:14:50 | Niddix | s/to/against |
00:14:58 | MicroChip32 | Niddix: where, and what size ? |
00:15:25 | Niddix | 64meg |
00:15:40 | Niddix | From http://newegg.com |
00:15:54 | MicroChip32 | hmmm cant use that in the tux can you ? |
00:16:18 | Niddix | CF? Sure. It behaves like any disk. |
00:16:51 | Niddix | MicroChip32: Need a PCMCIA->CF adapter too. |
00:17:05 | MicroChip32 | ahhhhhhh, that explains it |
00:17:15 | MicroChip32 | i was thinking the memory card in the bottom of the tux |
00:17:31 | Niddix | MicroChip32: No.. not a SODIMM. CompactFlash. |
00:17:42 | MicroChip32 | yep doh, dont mind me, long day |
00:18:25 | Niddix | MicroChip32: I think I paid $60 for the other CF card. Of course its a SanDisk. |
00:18:46 | MicroChip32 | $25 for 64m sounds really good |
00:19:13 | Niddix | MicroChip32: It'll probably give me cancer or something. |
00:19:18 | MicroChip32 | lol |
00:19:37 | MicroChip32 | yea but you wont know it for another 20 years |
00:19:56 | Niddix | I'll be 50 then and won't care all that much I think. |
00:20:24 | MicroChip32 | hehehe, we'll both be 50 and have more important things to think about |
00:20:36 | Niddix | spills a full cup of water that he was sure was empty all over his desk. |
00:20:44 | MicroChip32 | ewwwww |
00:21:10 | MicroChip32 | better than dumping a glass of lemonade into your pc |
00:21:19 | Niddix | Anyone water test on of these tux's yet? |
00:21:41 | Niddix | MicroChip32: Thats why I only drink water near my computers. |
00:21:54 | Niddix | MicroChip32: I've de-coke'd too many keyboards. |
00:21:57 | BZFlag | home |
00:22:05 | MicroChip32 | Niddix: unfortunately it wasnt my lemonade ... was my neice *mutter* |
00:22:24 | bhimaji | that's why you don't let kids into your house. |
00:22:52 | Niddix | MicroChip32: Thats not too bad. My daughter topped my trucks gas tank off with the garden hose last weekend. |
00:22:52 | MicroChip32 | took me a whole day and a half a box of q-tips to de-gunk the motherboad, and 3 pci cards |
00:23:08 | bhimaji | mc32: it worked after that? |
00:23:16 | bhimaji | was it turned on? |
00:23:30 | MicroChip32 | bhimaji: yes it was on at the time, and yes it survived, im using it |
00:23:38 | bhimaji | is impressed. |
00:23:48 | Niddix | Oddly enough the screen is tinted yellow now. |
00:24:25 | MicroChip32 | Niddix: lol, nope, that only happens when i forget to visit mother nature regularly, then the screen and keyboard assume a yellow glow |
00:24:43 | bhimaji | is going to go home now and attempt to make a DIN enclsoure for his tuxscreen. |
00:24:49 | bhimaji | (ie: car DIN outlet) |
00:26:26 | MicroChip32 | thinks he's gonna go to bed early and sleep for 12 hours or so |
00:26:41 | bhimaji | I should do that. |
00:26:41 | MicroChip32 | falls asleep on the channel floor |
00:26:50 | Niddix | This is just a general question to anyone who might know. Can the IR port on the shannon understand signals from TV remotes? |
00:27:14 | MicroChip32 | hmmmmmm, interesting idea |
00:27:43 | Niddix | I've got a multifunction remote with 4 (functions) that I don't use. |
00:27:57 | Tangent | Niddix : It should be possible to do that.. yes |
00:28:22 | MicroChip32 | now if there was an IR-tx on the tux ... |
00:28:42 | Niddix | Tangent: Its just a matter of a driver that recognizes the signals, yes? |
00:29:03 | MicroChip32 | hey, the ir uart should have both rx/tx lines right Tangent ?? whats the tx line being used for ? |
00:29:05 | Aelhaz | MicroChip32: you could do that on a gpio pin tho |
00:29:17 | Aelhaz | hmm, yeah |
00:29:23 | MicroChip32 | Aelhaz: i was under the impression there wasnt any free gpio lines |
00:29:51 | Aelhaz | there's at least one on j602, isn't there? |
00:29:55 | Tangent | MicroChip32 : AFAIK, the uart tx lins is not used |
00:30:11 | MicroChip32 | Tangent: then we should be able to hook that up to an ir-tx no ? |
00:30:20 | Tangent | MicroChip32 : In theory.. yes |
00:30:37 | Tangent | MicroChip32 : We could replace the IR receiver with an IR transceiver |
00:30:38 | MicroChip32 | could implement a complete entertainment system control box, with programmable display ;) |
00:31:03 | Tangent | MicroChip32 : Kinda like a beefy version of the Phillips Pronto :) |
00:31:10 | MicroChip32 | Tangent: no, would want the transmitter somewhere besides in the front |
00:31:23 | Niddix | MC32: So move it. |
00:31:33 | Tangent | MicroChip32 : You can always rotate the X dispaly |
00:31:46 | MicroChip32 | Tangent: it wont rotate 360 |
00:31:48 | MicroChip32 | or even 180 |
00:31:54 | Niddix | MC32: Instead of the serialport / shanip mod that Tangent did. Make serial 2 another IR |
00:32:22 | Tangent | MicroChip32 : 'xrandr 2' I believe does 180 degree rotate |
00:33:01 | MicroChip32 | Niddix: why, if the tx of the ir uart is already there. id just make the tx diode separate from the existing rx sensor, so i could point the tx at my tv/etc |
00:33:12 | MicroChip32 | Tangent: my display rotates about 40 degrees total |
00:33:17 | MicroChip32 | left/right anyway |
00:33:29 | Tangent | MicroChip32 : The X server can rotate the X display too tho |
00:33:43 | Niddix | MicroChip32: You can turn it around if you take it apart. |
00:34:04 | Tangent | Mounting the TX in the back sounds easier :) |
00:34:10 | MicroChip32 | much |
00:34:24 | Niddix | Sure |
00:34:53 | MicroChip32 | actually Tangent: i think a small port with a "remote" unit which could be attached wherever the user wanted it, like velcro to the top of the display or something |
00:34:59 | Niddix | types flash ramdisk and crosses his fingers. |
00:35:11 | Tangent | MicroChip32 : Serial port dongle |
00:35:40 | MicroChip32 | Tangent: naaa, wanna leave the existing and expansion serials free |
00:36:34 | MicroChip32 | come to think of it tho, i dunno if a serial ir will work for tv/stereo/etc control ... will it ? |
00:37:01 | Niddix | You can buy some that will |
00:37:19 | Tangent | MicroChip32 : Here's the magic you'd want.. It'd take a little PCB.. http://www.epanorama.net/circuits/ir_send.html |
00:37:25 | Niddix | I can change channels with my laptop. |
00:38:45 | Sammy | hello all |
00:38:59 | Niddix | Sammy! |
00:39:40 | Tangent | Sammy : Morning |
00:39:41 | MicroChip32 | Tangent: hmmmmmmmm |
00:40:23 | MicroChip32 | well guys (and gals if any) im off to bed for the duration |
00:41:23 | Tangent | MicroChip32 : night |
00:41:33 | Niddix | Night MicroChip32 |
00:41:46 | Niddix | I used to be able to read schematics. |
00:41:50 | Niddix | Been too long. |
00:42:00 | MicroChip32 | takes off his shoes and leaves em in the channel ... g'nite <grin> |
00:42:26 | Tangent | moves the shoes to the windowsill and opens the window... |
00:42:27 | Niddix | Oh my... linux on my tux. |
00:42:44 | Tangent | Niddix : Well done.. Congrats :) |
00:43:03 | Niddix | Although it appears my touchscreen isn't working. |
00:45:50 | Niddix | Guess I'll load tuxscreen-image instead of microwin-image. |
00:47:02 | GoRK | iirc none of the current binary images online contain the patch that allows 100% of touchscreens to work. the picogui image does but does not include a calibrator.. no images have been released in quite a while |
00:48:01 | Niddix | If I grab the buildroot from CVS should it be ok? |
00:48:55 | Tangent | Niddix : it works for me |
00:49:36 | Niddix | Weird.. minicom is really chewing up the text. when I do a ls -all |
00:50:02 | Niddix | Does the kernel support virtual consols? |
00:50:21 | Niddix | Appears not |
00:50:26 | GoRK | russ (and other kernel hackx0rs): the ucb1x00 driver is failing because the mcp fails to init properly |
00:50:42 | GoRK | in 2.4.13 |
00:51:20 | GoRK | mcp_get takes a big poopie |
00:51:50 | Tangent | GoRK : core dumped.. so to speak :) |
00:52:08 | GoRK | well not really more like mcp_if never gets set for some reason |
00:52:46 | GoRK | ahhhhhhhhhhhhhh |
00:52:49 | GoRK | doh |
00:53:30 | GoRK | might have it here |
00:55:24 | GoRK | ucb1x00 ID found: 1004 ok that is a good sign then |
00:55:28 | Niddix | Muhahaha.. I can read my CF card. |
00:56:44 | GoRK | ok it detects the device and hard freezes when you try to use it now :) |
01:01:52 | GoRK | audio :) |
01:02:00 | GoRK | whoot |
01:02:13 | GoRK | the sample rates are pretty screwy |
01:02:18 | GoRK | but it doesnt skip |
01:02:24 | GoRK | output: sample frequency 44100 Hz not available; closest 46800 Hz |
01:02:26 | scanline | GoRK: ucb1200 on a new kernel working?! |
01:02:37 | GoRK | scanline: yes.. im now gonna check touchscreen |
01:02:43 | scanline | :-D |
01:02:48 | GoRK | can you relay a message to russ for me? |
01:02:53 | scanline | put it in the topic |
01:03:10 | GoRK | oh yeah |
01:04:10 | Niddix | Any idea why /dev/hda1 /home default 0 0 wouldn't work in my fstab? |
01:04:23 | Niddix | I can mount -t vfat /dev/hda1 /home on the command line. |
01:04:34 | GoRK | Russ: mcp-sa1100.c needs a machine_is_shannon() in the if clause of mcp_sa_1100_init |
01:05:06 | GoRK | touchscreen still locks it but audio works |
01:05:11 | GoRK | ok i gotta go |
01:05:45 | scanline | bye |
01:09:52 | Aelhaz | iirc there's a bug in the vfat driver which makes it difficult if not impossible to give regular users write access to vfat partitions |
01:10:22 | scanline | Niddix: shouldn't it be 'vfat' in fstab instead of 'default' ? |
01:12:11 | Aelhaz | if you don't have to read the card on a windows box or in a camera, just reformat it to minixfs or something |
01:13:12 | Niddix | Aelhaz Well I figured it would be a cheap way to to allow me to transfer files onto my tux until I get a nic workingf. |
01:13:23 | Aelhaz | ah |
01:13:37 | Niddix | However I'll do that when I get the nick working. |
01:13:42 | Niddix | Till then I'll just be root. |
01:13:47 | Aelhaz | anyway, if you want it automounted the proper place is in the pcmcia scripts |
01:13:49 | Niddix | And I can mount it by hand. |
01:13:56 | Aelhaz | not in fstab |
01:14:07 | Aelhaz | ide.opts i think |
01:14:09 | Niddix | Aelhaz: True. |
01:15:04 | Niddix | Does the kernel have ipx support compiled in? |
01:15:25 | Aelhaz | eww, ipx . . . no, I don't think so |
01:15:53 | Aelhaz | can't think of a reason why it wouldn't work, tho, if you want to reconfigure and recompile |
01:16:08 | Niddix | Nod |
01:16:48 | Niddix | BTW.. CF in.. with vfat and you boot.. its ugly.. |
01:16:59 | Aelhaz | ? |
01:17:02 | Niddix | Lots of Waiting for chip to read, status = 4 messages. |
01:17:16 | Niddix | can't find jffs2 filesystem.. |
01:17:54 | Niddix | can't find ext2 filesystem.. |
01:18:12 | Niddix | it does boot however.. |
01:20:24 | Aelhaz | oh, that's normal |
01:21:25 | Niddix | Ah. |
01:21:53 | Niddix | Ok last stupid question for at least 5 minutes. |
01:23:41 | Niddix | nevermind. |
01:23:46 | Niddix | It worked this time. |
01:26:16 | Niddix | Cool.. First spare PCMCIA nic I tried and it works. |
01:28:02 | Aelhaz | congrats |
01:28:08 | Aelhaz | make/model? |
01:28:14 | Niddix | Well.. PCMCIA likes it anyway. |
01:29:12 | Niddix | It is an old MegaHertz of unknown model number. |
01:33:01 | Niddix | It does not however like my EM1144-T |
01:34:38 | Niddix | It does like my APA-1460 Slim SCSI Adapter. |
01:34:58 | Aelhaz | "cardctl ident" should tell you what model it is |
01:37:33 | Niddix | Aelhaz: product info "Megahertz", "ETHERNET ADAPTOR", "00206D100F68", "" |
01:38:11 | Niddix | My Mr Flash PCMCIA->CF adapter shows up as " ", " ", " " |
01:38:29 | Niddix | but it works so I won't complain. |
01:43:09 | Aelhaz | weird |
01:53:00 | Niddix | ouch that was dumb. |
01:54:40 | prpplague | anyone read the really stupid statements from gates/ballmer on newsforge? |
01:55:10 | prpplague | gates - microsoft is responsible for open source development |
01:55:27 | prpplague | ya right |
01:59:51 | Tangent | prpplague : Maybe if we all ignore them, then they'll go away |
02:00:16 | prpplague | Tangent: only if were lucky |
02:03:05 | Niddix | Aelhaz: Nevermind.. my nic doesn't seem to be working. |
02:04:55 | Aelhaz | Niddix: identifies itself but doesn't work? |
02:04:59 | Aelhaz | some nics do that |
02:05:07 | Aelhaz | the driver loaded & everything? |
02:05:14 | Niddix | I'm not getting any sort of message in dmesg. |
02:05:27 | Niddix | I'd say no.. just cardctl is identifying it. |
02:05:33 | Aelhaz | hmm |
02:05:38 | Aelhaz | maybe you're just missing the driver |
02:06:17 | Niddix | Most of the other Megahertz seem to be using smc91c92_cs |
02:06:29 | kram | sup guys |
02:06:31 | Aelhaz | what's ifconfig -a say? |
02:06:34 | Niddix | So I tried making an entry in config for the card. But it doesn't seem to care. |
02:06:41 | Tangent | Niddix : That's exactly what my 5V NIC does at 3.3V |
02:06:43 | Aelhaz | well yeah if the driver isn't there |
02:07:08 | Aelhaz | 5v hacks are easy these days if you're handy with a soldering iron |
02:07:13 | Niddix | Aelhaz: Just the loopback. |
02:07:26 | Aelhaz | hmm |
02:07:46 | Niddix | I tried running /etc/pcmcia/network start eth0 |
02:07:53 | Aelhaz | got anything else with pcmcia in it? |
02:07:58 | Niddix | But it complained about something missing. |
02:08:01 | prpplague | has been working for 12 hours 2 minutes |
02:08:17 | Niddix | I've got a CompactFlash adapter inserted right now. |
02:08:36 | Aelhaz | no i mean, another linux box to test it in |
02:08:38 | Tangent | gives prpplague a medal |
02:08:42 | Niddix | Ah... |
02:08:49 | Niddix | Yeah I can try my laptop. |
02:09:53 | prpplague | Tangent: can i trade the medal for a nice stout? |
02:10:09 | Tangent | prpplague : There's no such thing |
02:11:38 | prpplague | Tangent: oh come on, your from the uk, ya gotta like stouts... |
02:14:06 | Tangent | prpplague : Yeah.. I'm in the UK, but I can't stand bitter etc... |
02:14:25 | Tangent | prpplague : I like a good german or czech lager myself |
02:15:06 | prpplague | Tangent: lagers, pilsners? |
02:15:14 | Tangent | prpplague : Yep |
02:16:34 | prpplague | Tangent: as long as it not american lagers or pilsners |
02:16:59 | Tangent | prpplague : I believe the old joke is... American beer is like making love in a canoe |
02:17:24 | prpplague | whats the punch line? |
02:17:45 | Tangent | fsking close to water |
02:17:50 | prpplague | lol |
02:18:02 | prpplague | yep |
02:20:42 | prpplague | Tangent: http://www.ratebeer.com/ViewUser.asp?UserID=1210 |
02:20:46 | Tangent | Hmm.. I need to find some good fonts for the TuxScreen... anti-aliased fonts and sub-pixel rendering don't really work too well on a 256 colour display |
02:21:27 | Tangent | prpplague : Cool... actually Sam Adams isn't too bad as beer goes |
02:21:37 | Tangent | prpplague : The summer ale was a little to sweet tho' |
02:22:10 | GPSFan | prpplague: if you like good beer you need to try and find some: http://www.stonebrew.com/ |
02:22:41 | prpplague | GPSFan: yep got about 6 cases of all stonebrew coming in next week |
02:22:55 | prpplague | GPSFan: good beer in texas is hard to find |
02:23:01 | prpplague | GPSFan: i have to special order |
02:24:03 | GPSFan | prpplague: check out: http://www.stonebrew.com/cool/album/anniv2/index.html I took the pic's with blue borders. |
02:24:04 | Tangent | prpplague : This is my favorite lager... which you can't get in the US at all... http://www.budvar.cz/ |
02:26:02 | GPSFan | Tangent: budvar is GOOD beer. |
02:26:19 | Tangent | GPSFan : Certainly is :) |
02:30:21 | Niddix | Aelhaz: It appears that maybe the card is a dead stick. |
02:30:31 | Niddix | Aelhaz: Its not even working under windows. |
02:32:58 | Aelhaz | Niddix: weird |
02:33:20 | Tangent | hits the hay.. Night all |
02:33:29 | Aelhaz | Niddix: is it by any chance a thing with a big dongle on it that's got both rj45 and bnc sticking out the side of it? |
02:33:30 | GPSFan | Tangent: nite~ |
02:34:03 | Aelhaz | which would be a CE-2, which was end-of-lifed a long time ago and there's no linux driver |
02:34:47 | prpplague | Tangent: later |
02:35:18 | Aelhaz | Niddix: there's a list of working 3.3v nics on the wiki - you could just get any nic with a linux driver in the kernel and do a 5v mod if it doesn't work at 3.3v tho |
02:35:50 | prpplague | Tangent: i've not had budvar, looks good though |
02:36:11 | prpplague | GPSFan: you bum, i have to wait 45 days for a six pack of stone |
02:38:26 | GPSFan | prpplague: haven't had a stone in quite a while, we moved to Colorado, and there are no dist's near, although we are working on it. ;>) |
02:39:34 | prpplague | GPSFan: i've convinced a local gas station chain to do a "beer of the month" |
02:39:37 | GPSFan | prpplague: we almost drove back for Stone 5, but Sept 11 happened, and well ..... |
02:39:49 | prpplague | bummer |
02:40:15 | GPSFan | prpplague: it really was, we were going to bring back about 25 cases. |
02:40:46 | prpplague | GPSFan: one of the good things about my job in get to travel about once a month, and most of the time its in the caribean so i get to try lots of unique brews |
02:43:30 | GPSFan | prpplague: I see you like fullers, I've got a couple of bottles in the fridge, waiting for a good time to celebrate.. SD was pretty good for unique beer availability, CO isn't as good, but the brewing weather is better, if only I could find time to get back to it again. |
02:44:37 | GPSFan | prpplague: does the gas station give a free brew with fillup? ;>)) |
02:48:26 | prpplague | GPSFan: i wish |
02:49:26 | prpplague | GPSFan: ya fullers london porter is really smooth, a real pleasure to drink |
02:49:50 | GPSFan | prpplague: I'm looking forward to enjoying it!! |
02:50:39 | prpplague | GPSFan: you had any mac-n-jack? |
02:51:07 | GPSFan | prpplague: can't say that I have, where is it brewed? |
02:51:14 | prpplague | seattle |
02:51:24 | prpplague | only available on draft |
02:51:52 | GPSFan | ah, seattle, there are some really good microbreweries there. |
02:52:32 | prpplague | GPSFan: there are not 1, not 2, but 3 microbrewiers in the seatac airport |
02:53:11 | GPSFan | grants, ballard, thomas kemper, pyramid, and more since I was last ther. |
02:53:25 | GPSFan | s/ther/there |
02:54:02 | Russ | GoRK: devfs wasn't your problem, commenting out that line is fstab is what fixed it |
02:54:48 | prpplague | argh, i'm ready to go home, i wish this guy would call and let me know that his router is up |
02:58:00 | __jason | late hours again, eh prpplague? |
02:58:42 | prpplague | that you lefler? |
02:59:10 | __jason | yep...thought I'd check in on ya |
03:00:03 | prpplague | __jason: ya got a couple of installers out there setting up routers |
03:34:08 | prpplague | GPSFan: later |
03:34:10 | prpplague | night all |
05:06:06 | jacques | anyone know the resolution of the tux touchscreen? |
05:31:33 | Russ | 640x480 |
06:11:22 | GoRK | Russ: yeah i know i discovered that. i forgot to take it out :-/ did you get my message about the mcp ? |
06:12:09 | GoRK | ah i have finally discovered a good way to organize all my resistors! |
06:12:43 | GoRK | i found 3x5 ziplock bags. i can put an index card in them and resistors and stick them in a card file :) |
06:17:31 | jacques | whoo! I just got a free 25GB laptop drive |
06:20:21 | jacques | Russ: heh, I was referring to the resolution of the touch sensor - is it really pixel resolution? |
06:21:42 | GoRK | jacques: it is actually probably better than that if you take multiple readings. the driver in 2.4.7 returns up to 1000x1000 |
06:21:55 | GoRK | jacques: i dont know yet how the new driver works |
06:23:12 | jacques | GoRK: WOW |
06:24:47 | GoRK | jacques compared to rf tablets that's nothing. i have a calcomp 12x12 tablet that gets like 2500 points/in |
06:25:36 | GoRK | BZFlag: got audio. no touchscreen yet :-/ |
06:26:38 | jacques | GoRK: again I sa yWOW |
06:27:23 | Russ | GoRK: ya |
06:28:40 | GoRK | Russ: anyway ts is hard locking. could it be the backlight stuff? did you apply any big changes to that driver? |
06:44:34 | jacques | GoRK: you testing a new kernel version? |
06:45:31 | GoRK | yeah :) russ is doing the hard stuff though. i learned how to code from a klutz book |
06:46:10 | GoRK | it came with a free compiler |
06:50:29 | jacques | :-) |
07:41:30 | GoRK | lalalala sorting resistors is funnnnnnnnnnnnn |
07:42:07 | jacques | no it's not |
07:42:47 | rfbrown | I hope they're not like the ones on the Tux PCBs... |
07:43:04 | jacques | yeah - room temperature superconductors |
07:43:42 | GoRK | what? smt devices? |
07:43:56 | jacques | 0 Ohm resistors |
07:44:09 | GoRK | i have some 1 ohm resistors here too |
07:44:11 | jacques | (I know I know - this joke has gone on too long) |
07:48:01 | GoRK | if only it was at all funny :) |
07:48:25 | jacques | :-\ |
07:48:30 | jacques | i thought it was funny the first time |
07:48:38 | jacques | th othis it like the 8th time... |
07:48:43 | GoRK | hih im joking |
07:48:53 | jacques | but ppl say I'm easily amused |
07:52:31 | BZFlag | just think of all the 0 ohm resistors between my computer and yours. |
07:55:00 | GoRK | any of thousands could explode at any moment and render this communication impossible |
07:56:14 | jacques | hi BZFlag |
07:56:30 | jacques | got up too late to get my order in today, but I will get it in this weeknd |
07:56:38 | jacques | just trying to decide how many to get :-D |
07:59:29 | rfbrown | < BZFlag has pleasant daydreams about all the things he can do with his basement when these $@#%@ phones are gone...> |
08:00:38 | jacques | heh |
08:01:51 | BZFlag | heh. just finished packing up tonights phones. |
08:02:10 | scanline | 'night everyone |
08:02:10 | BZFlag | GoRK: that must be what happened to riker.dsl.inconnect.com |
08:02:25 | GoRK | BZFlag: heeheh |
08:02:27 | scanline | is away: ZzZzZzZz |
08:02:49 | jacques | BZFlag: biggest box you ship is 2 phones, right? |
08:02:54 | GoRK | ahh this is much nicer with all these resistors neatly organized... |
08:03:12 | jacques | GoRK: whatever makes you happy... :-) |
08:04:07 | BZFlag | jacques: I've been linking 2 2packs for people lately. 34'2" |
08:04:50 | jacques | BZFlag: oooooh |
08:04:56 | jacques | cool |
08:05:23 | jacques | so I can use the 34'2" value to calculate shipping cost for those? |
08:06:09 | BZFlag | can't link singles cause they are single boxed, but the doubles have 2 white boxes in one brown box. I can connect them together top to bottom. |
08:06:23 | BZFlag | jacques: yep. and include that in the comments. |
08:06:31 | jacques | BZFlag: right-o |
08:08:11 | rfbrown | Forgive my ignorance of imperial measures, but isn't 34'2" like thirty-four feet and two inches? What kind of letterbox takes that? |
08:08:29 | jacques | rfbrown: I have a really big door |
08:08:35 | jacques | really big |
08:09:52 | GoRK | rfbrown: it's 34 cubic feet i think |
08:10:10 | jacques | more like 34 pounds, 2 ounces |
08:10:18 | jacques | shipping weight |
08:10:22 | GoRK | oh heh yeah |
08:10:26 | jacques | :-D |
08:11:27 | rfbrown | Thanks... I knew I had it wrong somewhere, but the mental image was funny! |
08:13:58 | rfbrown | GoRK: I've dismantled my Tux for easy hardware hacking, and mounted the PCBs side by side on an aluminium sheet. Would it be useful for me to remove the RF shield from the DSP? ( As in, is there anything under there that I'll want to fiddle with?) |
08:16:56 | jacques | heading home |
08:18:17 | GoRK | rfbrown: i dont think so. the leads for the mongoose board that is under that are accessible on the underside of the board |
08:25:04 | rfbrown | I'm thinking that once I've caught up with what's going on, I'll offer to help <whoever it is> with the schematics... but I'll leave the shield on for now. |
08:35:22 | GoRK | it's ken mcguire (gpsfan) |
08:35:39 | GoRK | afaik he hasnt even booted his tuxscreen yet :) |
08:36:51 | rfbrown | thanks |
08:45:13 | BZFlag | nite |
09:23:49 | GoRK | hello |
09:23:54 | andersee | howdy |
09:24:10 | andersee | Just finished up the must do things for the evening... |
09:24:23 | GoRK | this audio driver doesnt skip at all :) |
09:24:29 | andersee | hows it going for you? |
09:24:41 | andersee | whats changed? |
09:24:46 | GoRK | the whole thing |
09:24:49 | GoRK | 2.4.13 |
09:24:54 | andersee | no more shanip2? |
09:25:05 | GoRK | yes still need shanip2 |
09:25:12 | andersee | oh. new kernel. gotcha. |
09:25:15 | GoRK | though the new driver does support the telecom codec |
09:25:40 | GoRK | so we can send and recieve audio from the line (i think) without shanip2 |
09:25:56 | andersee | hmm. that sounds very promising indeed. |
09:26:24 | andersee | shanip2 is still needed to play audio out the speakerphone speaker though, right? |
09:27:33 | GoRK | well... i suppose a few things are possible: 1) the ucb1200 telecom codec acts like a phone extension and could be used to play and record audio from the telephone handset or speakerphone... but it probably only will work if there is an actual phoneline hooked up. |
09:28:05 | GoRK | or 2) either the phoneline can be attached to ucb1200 or to the dsp ... so we could have sort of an answering machine situation |
09:28:24 | GoRK | ie, i dont think it will eliminate the need for shanip2 |
09:29:05 | GoRK | but it might |
09:29:06 | andersee | Any changes to the PCB recently? The one I saw from you the other day looked pretty sweet |
09:29:12 | GoRK | version 8? |
09:29:27 | GoRK | err 7 |
09:29:34 | GoRK | the one with stereo jacks and the transformer? |
09:29:40 | GoRK | http://www.blurbco.com/~gork/tuxscreen/shanip2-gork7.gif |
09:30:15 | andersee | nope, thats a newer one. :) |
09:30:19 | GoRK | i ordered some jacks and thumbwheel pots from mouser. they arrived today. i need some press n peel and i am giong to make one of these for testing the layout etc |
09:31:23 | andersee | cool. |
09:31:55 | andersee | Let me know how it goes. I'm looking forward to having a bit of time to make myself a couple of 'em. |
09:33:39 | GoRK | well if it works and the design is good, i can just put in an order for a ton of them... with solder masks and silkscreen they will be only about $5 i think |
09:34:00 | GoRK | the jacks and pot from mouser cost about $5 |
09:34:16 | andersee | I''ll buy a couple of em for sure. |
09:35:05 | andersee | bother. uclibc readdir64 segfaults. |
09:35:14 | andersee | I recently added large file support.... |
09:35:40 | GoRK | any love for libm yet? :) |
09:35:47 | andersee | nope. |
09:36:02 | GoRK | does diet-libc include a libm? |
09:36:16 | andersee | not much of one. c89 only. |
09:36:59 | andersee | uClibc can't steal from dietlibc since its GPL.... (cept ideas).... |
09:37:09 | GoRK | oh |
09:37:18 | andersee | But there are several other libs we can steal from. |
09:37:19 | andersee | :) |
09:37:22 | GoRK | since ucblibc is gpl or dietlibc/ |
09:37:26 | GoRK | ? |
09:37:35 | andersee | diet is GPL. |
09:37:43 | andersee | uclibc is LGPL |
09:38:12 | GoRK | that is so braindead to license a libc gpl. blah |
09:38:21 | GoRK | i didnt realize that |
09:38:25 | andersee | yup. Pretty freaking stupid |
09:38:52 | andersee | otherwise I would tried to merge the two |
09:40:25 | andersee | Might want to take the "ShanIP-2" bit and move it to the left hand side. Right now it is upside down rel to the "mic_input" text... |
09:40:49 | GoRK | i know.. mic_input is just for my own reference. it's not on the board itself |
09:40:58 | GoRK | shanip-2 gets screened onto the board |
09:41:47 | andersee | oh, ok. |
09:42:00 | GoRK | this board is pretty small... like 1.5" x 2.5" |
09:44:42 | andersee | heh. I just realized I have like 180 Gig connected to my box. |
09:45:19 | andersee | I wonder how many more test drives I can connect.... |
09:45:30 | GoRK | scsi? |
09:45:58 | andersee | scsi and ide |
09:46:17 | GoRK | i counted up all the storage i owned the other day... it's about a terabyte |
09:46:49 | andersee | I have about 10 more test drives in the basement. big ones too, like 80 Gig firewire drives and such |
09:47:09 | GoRK | sweet |
09:47:11 | andersee | Of couse, when I finish this project I have to send em all back. |
09:47:16 | andersee | :( |
09:47:42 | Squart | damn, thats to bad |
09:48:11 | andersee | But I'm expecting another "big box" of stuff this week. More firewire and usb stuff |
09:48:46 | andersee | So it'll be a while before I have to start returning stuff. |
09:49:24 | andersee | I'll be glad to return the tape drives. They are _way_ loud. |
09:50:44 | rfbrown | andersee: can the drives perhaps suffer unavoidable damage in the course of your project? You could generously offer to "dispose" of the worthless hardware ;-) |
09:51:07 | andersee | heh. That would be nice. |
09:51:28 | andersee | I hope this laptop suffers similar damage.... |
09:51:33 | GoRK | and if they want it back, just destroy it and ship it to them! |
09:52:02 | rfbrown | 'night all. |
09:52:05 | andersee | They sent me a laptop just like they use when they are in the field.... |
09:52:23 | andersee | It is nice for testing with. And other things. :) |
09:52:28 | GoRK | i feel bad staying up later than people in NEW ZEALAND |
09:52:35 | andersee | heh |
09:52:49 | andersee | Its only 4am. |
09:52:59 | GoRK | 5 here |
09:53:11 | andersee | I'll probably be asleep by then.... |
09:53:13 | andersee | :) |
09:53:27 | andersee | <not that I have been all week> |
09:56:33 | Squart | gees, it is 12:00 pm here |
09:57:38 | andersee | Sounds early. I don't get up till 1pm. ;-) |
09:58:33 | Squart | damn it, I wish I could do that... alarm rings at 7:15am |
09:59:15 | andersee | The joy of being a consultant. I don't even need to get out of bed to go to work... :) |
10:00:48 | Squart | Sigh... I did freelance work before. Although I liked working at home, you still end up with the feeling you aren't a part of the project |
10:01:08 | GoRK | gotta sleep. nite |
10:01:20 | Squart | bye |
10:01:49 | andersee | night |
10:02:29 | andersee | In this case I am the project. Both of them. |
10:04:49 | andersee | Though it would be nice to see other geeks on occasion. |
10:05:51 | Squart | exactly |
10:13:23 | Squart | Do you guys have a way to communicate new achievements on the Tux other than this irc channel? De website is rather summier... |
10:15:41 | andersee | the wiki and the mailing list |
10:15:52 | Squart | ok |
10:20:41 | Squart | can't seem to find the mailinglist. Can you tell me where it is? |
10:25:45 | Tangent | wakes up |
10:26:33 | Squart | 'good morning |
10:29:44 | Tangent | Squart : Mornin' |
10:30:27 | Tangent | Oooh.. is there a mailing list? |
10:32:10 | Squart | ;-) |
10:32:27 | Squart | andersee told me so... it seems it is on the sourceforge server |
10:32:41 | Squart | http://sourceforge.net/mail/?group_id=20228 |
10:34:17 | Tangent | I wonder if there's archives... Oh yes.. there is.. good... means I don't have to subscrbe |
10:35:33 | Squart | Would be nice if people would be more aware of it and that all the hacking results end up on a central place... like your second serial port hack |
10:39:34 | Tangent | Squart : I still find it very difficult searching down Tux info and documents on the site |
10:39:51 | Tangent | Squart : I'd spotted the mailing list before, but assumed it was CVS commit logs |
10:40:10 | Squart | There is one for that as well |
10:42:04 | Tangent | Now.. somehow... supposedly, I have write access, and can put piccies etc up on the site.. |
10:42:37 | Squart | I am new to this whole project and am still quite confused of how everything fits together |
10:45:17 | Tangent | Squart : Nod.. I've been here since pretty much the start, but I'm still confused too.. |
10:45:33 | Tangent | Squart : Maybe that's just natural stupidity tho' ;) |
10:45:53 | Squart | Tangent: ;-) so it is not only me then.. |
10:47:38 | Tangent | thinks sourceforge is misleading... Activity Percentile (last week): 0% |
10:48:04 | Squart | Tangent: I really like to order a Tux, but my PayPal credit account still isn't verified.. |
10:48:32 | Tangent | Squart : Yeah.. took me a few days for paypal stuff... I nearly didn't order becasue paypal pissed me off so much |
10:49:43 | Squart | ;-) hope there will be Tux'es left when it does verify.. |
10:49:56 | Squart | kee, gone for lunch |
11:29:54 | Tangent | adds a couple of wiki pages... http://www.tuxscreen.net/wiki/view/Users/JamesConner |
11:30:24 | Squart | Tangent: Cool.. |
11:31:45 | Tangent | I still didn't work out how to upload pictures to tuxscreen.net tho' |
11:32:31 | Squart | Tangent: Don't ask me either ;) |
11:34:10 | Squart | Tangent: You know I bould a secondhand LaserJet IIIP yesterday with 2 NEW toners for about 35$ . That's almost stealing! |
11:34:26 | Squart | bould==bought ;) |
11:34:36 | Tangent | Squart : That's a pretty amazing deal.. and those LaserJet IIIs run forever |
11:35:13 | Squart | Great huh? I thought I'd buy it for making PCB's |
11:35:40 | prh | sq: for 35$ you lucky sod! :O) |
11:35:57 | Squart | ;) |
11:35:58 | Tangent | Squart : It should be good for that... as long as it can print a decent quality black without missing dots etc |
11:36:11 | Tangent | My laser cost me around $1500 |
11:36:34 | Squart | auch.. |
11:37:05 | prh | tangent; what have you got? |
11:37:08 | Tangent | Squart : It does do 20ppm.. and it's colour too |
11:37:20 | Tangent | prh : Minolta Magicolour 2200 |
11:37:21 | prh | I was just going to ask if its colour :O) |
11:37:24 | prh | ooh nice |
11:37:36 | prh | wanna sell it for 35 quid? :O) |
11:37:43 | Tangent | s/colour/color |
11:37:53 | prh | no you were right the first time |
11:37:55 | Tangent | prh : Erm.. let me think... erm... nah |
11:38:14 | Squart | The only thing is: is has 2 outputs, front and top. When I use the top one, it wrinkles the bottomhalf of the page |
11:38:15 | prh | did you see http://www.ananova.com/news/story/sm_444729.html |
11:38:20 | Tangent | prh : Sadly I was wrong the first time... The writing on the side of the printer says Color (whatever that means) |
11:38:33 | prh | good grief |
11:38:42 | Tangent | prh : Maybe it's some special dithering algorithm to make black dots look like Colour |
11:38:52 | prh | that would be very cool |
11:39:17 | Squart | hehe |
11:39:21 | Tangent | prh : I vaguely remember filter screens for B&W TVs that were supposed to make it look like colour |
11:39:59 | prh | but all they did was turn white into a sort of yello? |
11:40:02 | prh | +w |
11:40:17 | Squart | Tangent: I see pictures now.. http://www.tuxscreen.net/wiki/view/Users/ShanIP232 |
11:40:35 | Tangent | ack.. keyboard death... it's permanently hammering the down arrow for me... fortunately that doesn't affect my IRC client |
11:40:46 | Tangent | Squart : The pictures are hosted on my box still tho' |
11:41:02 | Squart | ow.. I see ;0) |
11:43:04 | Tangent | Time to power cycle this cruddy Windoze box |
11:43:25 | prh | and put linux on it |
11:43:37 | prh | you know you want a real operating system on it |
11:44:10 | TanentTux | prh : You don't know how much I want a real OS on it... |
11:44:30 | TanentTux | prh : But then there's network multiplayer games on a thursday evening to think about |
11:44:48 | prh | hmm - what games do you plaY? |
11:45:06 | TanentTux | Squart : Having the piccies hosted on my box isn't too bad.. it's co-located with 10Mbps bandwidth |
11:45:44 | TanentTux | prh : Lately it's been Balders Gate 2... but we've done Diablo2, HalfLife, Star Trak Armade & Away Team etc |
11:46:13 | TanentTux | prh : Mainly strategy style stuff... me and my mates aren't really into shooters |
11:46:15 | prh | wishes he had time for that kind of stuff ;O) |
11:46:20 | Squart | TanentTux: That isn't bad at all. Are you irc'ing from your Tux now? |
11:46:47 | TanentTux | Squart: Yep.. Tux is doing IRC, and running my office automation stuff |
11:47:04 | Squart | Rule! |
11:47:33 | Squart | how is it connected to the internet? |
11:48:06 | TanentTux | Squart : 10BaseT NIC, plugged into my Hub/Firewall etc |
11:48:51 | Tangent | notes that the alpha state of the Tux is more reliable than the release version of W2K |
11:49:10 | Tangent | Actually it's unfair to blame these problems on W2k.. I think it's just a dodgy keyboard |
11:49:26 | prh | you're not actually surprised are you? |
11:49:37 | prh | w2k has a dodgy keyboard driver |
11:49:55 | Squart | TanentTux: But the Tux doesn't have 10BaseT Ethernet, does it route via serial? |
11:49:56 | Tangent | prh : Yeah.. well maybe.. but only on one of my machines... |
11:50:03 | Tangent | prh : The other one hasn't been rebooted for months |
11:50:08 | prh | centrica have been rolling 2k out to their call centres and they are finding that on _all_ the boxes the keyboards lock up about 3 times a week |
11:50:15 | Tangent | Squart : The Tux has 2xPCMCIA sockets |
11:50:27 | Tangent | Squart : I have 1GB microdrive in one of them, and NIC in the other |
11:50:28 | Squart | ah, f course |
11:50:42 | Squart | yeah that's great |
11:51:08 | Squart | Tangent: Without the microdrive you can't do much, or can you? |
11:51:39 | Tangent | Squart : Well, you can still NFS mount stuff.. but you can't do a huge amount with just the 4MB built in flash |
11:53:12 | Squart | Tangent: Writing to the flash wouldn't do good for its lifespan as well |
11:53:24 | prh | when I get mine I'll be hoping to do everything off NFS - ie the the flash to boot nfs |
11:53:27 | Squart | that is: writing a lot ;-) |
11:53:47 | Tangent | Squart : You'd never notice.. The flash is good for a million erases... so that's 60 full reflashes a day for 5 years |
11:54:59 | Squart | Yeah, I was thinking in the lines of it being a writable fs and linux writing on the same spot alot |
11:55:34 | Tangent | Squart : Nod.. but jffs2 is very efficient with it's erases... |
11:55:50 | Tangent | Squart : Don't forget.. it's not a million writes we get.. it's a million erases... |
11:56:11 | Tangent | Squart : An eraseblock (128K on the Tux I expect) will only be erased once it's filled up |
11:56:17 | prh | you only get 1000 writes ;O) |
11:57:08 | Squart | Tangent: Aha.. Does your Tux boot from flash or microdrive? |
11:57:30 | Tangent | Squart : And the journalling nature of jffs2, and built-in wear levelling means that we don't do erases too often anyhow... i.e. you can edit and save an 8k file 16 times before you need to do another erase |
11:57:48 | Tangent | Squart : It boots from flash, but the /linxrc does a pivot_root to the microdrive |
11:59:14 | Squart | is it possible to create a basic root on the flash, which won't be written to, and do everything else from nfs? |
11:59:33 | Tangent | Squart : Yes... in fact it's pretty much done by default.. |
12:00:12 | Tangent | Squart : By default, the /linuxrc will a.) look for a hard disk and pivot_root to there if it's found, or b.) Look for an NFS mount, and pivot_root to there or c.) boot from flash |
12:00:31 | Tangent | Squart : The flash is still writeable in b.) but there's no processes trying to write to it |
12:02:22 | Squart | Tangent: So in the a) and b) the flash basicly only holds the kernel and /linuxrc |
12:03:38 | Tangent | Squart : The linuxrc is a shell script.. it's the first thing that gets executed after the kernel starts... so the flash hold a full bootable filesystem, but it'll never get as far as anything that uses it (apart from the /linuxrc) unless we fail to find HDD and NIC |
12:04:01 | Tangent | s/unless/if/ |
12:04:28 | MicroChip32 | s/if/unless |
12:05:08 | Squart | Tangent: aha.. Is it just as easy to setup a network over serial and connect nfs through it? |
12:05:40 | Tangent | Squart : Well I haven't tried on the Tux, but I have done it on the iPAQ over PPP |
12:05:53 | Tangent | Squart : I wouldn't recommend it particularly.. it's somewhat slow |
12:06:02 | uChip | daggone isp |
12:06:26 | fontenot | nfs over serial is worst case scenario |
12:06:32 | uChip | i swear my isp watches me for the most inopportune time to discon me |
12:06:47 | Squart | Tangent: True... but it is nice for first, easy, cheep testing |
12:06:58 | Tangent | uChip : They do.. |
12:07:20 | Tangent | sets uChip's network traffic as discard elligible |
12:07:24 | uChip | Tangent: they like to discon me at the end of a long download |
12:07:48 | MicroChip32 | <grin> well at least my nick autochanger is working again |
12:08:17 | Tangent | Hmm.. maybe we could get ibot doing that |
12:08:31 | Tangent | that's a very handy thing to have |
12:08:55 | MicroChip32 | yea it is. mine waits for my 'ghost' to leave and auto changes my nick back to my main |
12:08:59 | | i haven't a clue, fontenot |
12:08:59 | fontenot | ibot: who am I? |
12:09:41 | | i haven't a clue, microchip32 |
12:09:41 | MicroChip32 | ibot: what is the purpose of life |
12:09:50 | MicroChip32 | daggone clueless bots ;) |
12:10:19 | MicroChip32 | dcc's ibot a clue |
12:10:36 | | no idea, tangent |
12:10:36 | Tangent | ibot : the purpose of life is <reply> What??? you think there's a purpose to all this?? |
12:10:51 | MicroChip32 | lol |
12:10:51 | | tangent: wish i knew |
12:10:51 | Tangent | ibot : the purpose of life is <reply> What... you think there's a purpose to all this\?\? |
12:10:53 | prh | pants. I've tried doing the registration for paypal using my barclays debit card... the barclays web thing doesnt show the paypal confirmation code. GRrrrrr |
12:11:00 | | OK, Tangent. |
12:11:00 | Tangent | ibot : the purpose of life is <reply> What... you think there's a purpose to all this |
12:11:18 | prh | tangent: of course there is a purpose. to spod! |
12:11:35 | | I'm just an idiot. :-( |
12:11:35 | fontenot | ibot: what is your major malfunction? |
12:11:40 | MicroChip32 | lol |
12:11:41 | SteveD_atwork | thinks - no point trying to imitate ibot. He's just too cool as it is |
12:11:49 | prh | but seriously - I believe we were made for a purpose and Geneisis describes that |
12:12:55 | Tangent | I seriously believe almost the exact opposite.. |
12:13:10 | MicroChip32 | Tangent: how can you change ibots answer for something? |
12:13:17 | SteveD_atwork | prh: you mght just have to wait for your statement? the paypal business for non US is a pain |
12:13:24 | Tangent | MicroChip32 : ibot forget xxx |
12:13:35 | Squart | the paypal verification takes awhile... |
12:13:35 | MicroChip32 | agrees with prh (mostly) |
12:13:40 | MicroChip32 | tan, k |
12:13:47 | Tangent | SteveD : There's always phoning the bank :) |
12:14:03 | prh | tangent: you believe in creative accidents? |
12:14:04 | SteveD_atwork | Tangent: that what you did? |
12:14:24 | prh | is tempted to phone the bonking bank - but they are crap |
12:14:45 | fontenot | thinks he might be on the wrong channel. |
12:14:47 | MicroChip32 | Tangent: ask ibot for "the answer" |
12:14:58 | SteveD_atwork | prh: not familiar with that sort of bank |
12:15:04 | Tangent | prh : Nah.. I believe lots of weird stuff.. kind of a superset of the concepts of good and evil from different religions.. with a bit of interconnectedness of all things... and self aware universe type of thing... |
12:15:14 | fontenot | hi |
12:15:28 | SteveD_atwork | renames Tangent to Dirk Gently. |
12:15:31 | Tangent | prh : It works for me |
12:15:33 | MicroChip32 | changes the topic to theology |
12:15:38 | prh | riighhhtt... :O) |
12:15:57 | | the answer is 42 |
12:15:57 | Squart | ibot: What is the answer? |
12:16:02 | Squart | hehe |
12:16:11 | | i guess the question is who'll get there first... I should have gone to bed about two hours ago.. but I'm playing on Tangent's stowaway keyboard now |
12:16:11 | SteveD_atwork | ibot, what is the question |
12:16:16 | | prh: bugger all, i dunno |
12:16:16 | prh | ibot: what was the question? |
12:16:43 | | fontenot: San Jose, San Jose International Airport, CA, United States; (KSJC) 37-21-33N 121-55-27W 25M; last updated: Nov 09, 2001 07:53 AM EST; Dew Point: 43.0 F (6.1 C); Pressure (altimeter): 30.06 in. Hg (1017 hPa); Pressure tendency: 0.00 inches (0.0 hPa) lower than three hours ago; Relative Humidity: 73%; Sky conditions: mostly clear; Temperature: 51.1 F (10.6 C); Visibility: 10 mile(s); Weather: 37 F (3 C); Wind: Calm; Windchill: 26 F (-3 C) |
12:16:43 | fontenot | ibot: weather ksjc |
12:17:14 | Tangent | wonders when ibot has been playing with my keyboard??? |
12:17:33 | | Tangent: what? |
12:17:33 | Tangent | ibot doesn't even live in my house... ibot lives about 30 miles form here |
12:17:34 | Squart | Who loaded it up with all those answers? |
12:17:47 | | Since Mon Nov 5 15:33:28 2001, there have been 88 modifications and 216 questions. I have been awake for 3 days, 20 hours, 44 minutes, 18 seconds this session, and currently reference 3029 factoids. Addressing is in require mode. |
12:17:47 | fontenot | ibot: status |
12:17:48 | prh | who owns ibot then? |
12:17:51 | | Tangent: what? |
12:17:51 | Tangent | ibot : I was talking about you.. not to you... be quite |
12:18:01 | fontenot | quite what? |
12:18:34 | | Slashdot - Updated 2001-11-09 12:52:42 | Pedal Your Way Through Quake | USNA "Budget" Satellite Launched and Functioning | KDE Wins 3 awards | Ballooning into Space |
12:18:34 | fontenot | ibot: slashdot |
12:18:39 | Tangent | prh : I set ibot up on one of my co-located boxes... but BZFlag offered to do all the maintenance/patches etc... which I'm very grateful for |
12:19:07 | | Tangent is nothing but an elf-skinned mound of toad-spotted cold sores. |
12:19:07 | fontenot | ibot: insult Tangent |
12:19:13 | fontenot | ewww |
12:19:16 | Tangent | It's true |
12:19:34 | MicroChip32 | ewww |
12:19:34 | prh | is the source open? |
12:19:37 | | fontenot: huh? |
12:19:37 | fontenot | ibot: nickometer |
12:19:44 | | 'fontenot' is 0% lame, fontenot |
12:19:44 | fontenot | ibot: nickometer fontenot |
12:19:45 | Tangent | You wouldn't believe how hard it is to get elf skin these days |
12:20:13 | fontenot | yep, they are an endangered species |
12:20:25 | fontenot | you have to get grandfathered elf skin |
12:21:06 | Tangent | fontenot : Takes ages to iron the wrinkles out :) |
12:21:40 | fontenot | Tangent: yep, and really easy to scorch |
12:22:19 | Tangent | wishes he's never mixed Lord of the Rings, Silence of the Lambs, LSD, PCP and Crack |
12:22:38 | fontenot | ROFLMAO |
12:22:45 | MicroChip32 | Tangent: the easy way to iron them is take em to an ironworks factory and have em run it thru one of the iron rollers |
12:23:04 | fontenot | Tangent: that's classic |
12:23:29 | MicroChip32 | egads, the first two would be bad enough |
12:23:41 | Tangent | Maybe we should start a page for classic quotes... I was reading the quotes for #debian a week or so ago... some of them are very funny |
12:24:09 | fontenot | definitely |
12:24:23 | fontenot | Tangent: where did you find the quotes for #debian? |
12:24:38 | Tangent | fontenot : I wish I could remember.. i'm just hunting around now |
12:25:03 | Tangent | http://www.red-bean.com/~decklin/quotes/debian.txt |
12:25:08 | Squart | This is all to funny, I don't get any work done here.. |
12:25:38 | Tangent | remembered about DMAU... google narrowed it down to one page... DMAU is 'Debian Mantainers against Users' |
12:26:12 | MicroChip32 | anyone here feel inclined to help me debug some segfaulting c ? |
12:27:06 | MicroChip32 | heheheh boy that shut everyone up *lol* |
12:27:15 | fontenot | those wacky debian guys |
12:27:44 | Squart | I am going to a nice employee and earn my money.. later guys |
12:28:12 | Tangent | MicroChip32 : Sure.. I'll help... I beleive the 'rm' command normally stops most things from segfaulting... but seriously.. I doubt I'd be much help |
12:28:30 | MicroChip32 | heheh youre so helpful Tangent :) |
12:28:48 | Tangent | read BOFH from an early age |
12:29:13 | MicroChip32 | bofh? **thinking he should know that** |
12:29:39 | fontenot | Tangent: i read that at a very impressionable age |
12:30:16 | Tangent | MicroChip32 : Bastard Operater From Hell.. find the early stuff.. very funny.. later stuff isn't as good |
12:30:30 | fontenot | "ok, unplug the monitor, now stick your tongue on the prongs of the eleftrical cord" |
12:31:36 | fontenot | agrees with Tangent. |
12:31:47 | Tangent | MicroChip32 : http://bofh.ntk.net/Bastard.html |
12:31:57 | MicroChip32 | hmmm never heard of it, doesnt sound like my type |
12:32:54 | MicroChip32 | reads tom clancy, and clive cussler |
12:33:23 | MicroChip32 | also coonts, crichton, and cook |
12:33:29 | Tangent | MicroChip32 : Me too.. .BOFH is just one page every couple of weeks |
12:33:41 | MicroChip32 | Tangent: hmmm |
12:34:05 | fontenot | reads only authors whose last name does not start with the letter C. |
12:34:18 | MicroChip32 | lol fontenot |
12:35:11 | fontenot | that's a lie - I have read just about everything crichton has published |
12:35:28 | fontenot | also greg bear |
12:35:32 | fontenot | vernor vinge |
12:35:58 | fontenot | of course, stephenson |
12:36:12 | MicroChip32 | bbias, gotta have breakfast |
12:39:01 | fontenot | mmmm, 1/60th minute eggs |
12:39:33 | prh | agrees with tangent - the early bofh was inspired. recent stuff is worth reading but not as good as simons first writing |
12:40:48 | MicroChip32 | fontenot: nope, pop tarts |
12:40:53 | fontenot | "can I have more space on my account?" "what account?" |
12:41:23 | MicroChip32 | that sounds like a dilbert |
12:41:32 | fontenot | MicroChip32: which ones? I am embarassed to say the ones I like. |
12:42:06 | MicroChip32 | fontenot: lol, frosted brown sugar cinamon. cant find my favorite any more, frosted dutch apple |
12:42:10 | Tangent | User: "I need more space" , BOFH: "Move to Texas" |
12:42:28 | MicroChip32 | no no, move to ISS |
12:43:33 | fontenot | MicroChip32: oh ... my ... god |
12:43:49 | MicroChip32 | wonders what he said |
12:44:23 | fontenot | frosted dutch apple are my all time favorite |
12:44:28 | MicroChip32 | lol |
12:44:31 | fontenot | havent seen those in years |
12:44:46 | fontenot | brown sugar are my third fave |
12:44:53 | MicroChip32 | me either, daggone marketers push the new crap |
12:45:07 | fontenot | these days I generally eat the s'mores ones - that's why I'm embarassed |
12:45:23 | fontenot | with the graham cracker crust |
12:45:28 | fontenot | :-D |
12:45:39 | MicroChip32 | hehehe, never tried that 'strange' stuff prefer the 'simple' stuff |
12:45:52 | fontenot | yeah, i didnt think id like them either |
12:46:15 | fontenot | but the combination of artificial ingredients is in perfect harmony |
12:46:30 | MicroChip32 | looks around, wondering why everyone is staring at him and fontenot |
12:47:16 | fontenot | heh |
12:47:48 | fontenot | at work we get free snacks |
12:48:49 | fontenot | cokes, bottles of coffee, juice, pop tarts, cookies, chips... |
12:49:03 | MicroChip32 | lucky |
12:49:14 | fontenot | yeah, I think so |
12:49:23 | prh | thats nice |
12:49:27 | prh | we get... nothing |
12:49:38 | prh | and our bar is being closed end of year |
12:49:42 | fontenot | :-( |
12:49:57 | MicroChip32 | stares at his source code trying to figure out why it segfaults at the end, but commenting out either of two lines of code in the middle of the execution cycle fixes it |
12:50:12 | MicroChip32 | not cool prh |
12:51:18 | fontenot | maybe it's leaving the stack in a bad state |
12:51:46 | MicroChip32 | i dunno how, you feel like having a look ? |
12:52:36 | fontenot | MicroChip32: im afraid i would not be much help - i havent actually written C in over 6 years |
12:53:04 | fontenot | MicroChip32: what does the code do? |
12:53:09 | MicroChip32 | fontenot: then you have 4 years on me, i aint done c since college in 92 |
12:53:21 | fontenot | is it pretty self-contained? |
12:54:06 | MicroChip32 | fontenot: yes, and the code in question simply grabs a numeric value from a fixed-size field in a string and returns it as an int |
12:54:10 | fontenot | MicroChip32: do you get a core file? have symbols compiled in? debugger tell you anything? |
12:54:25 | MicroChip32 | core file yes, dunno nuffin about doing debugging stuff |
12:54:32 | fontenot | using gcc? |
12:54:34 | MicroChip32 | yes |
12:54:52 | fontenot | compiling with -g ? |
12:54:57 | MicroChip32 | no |
12:55:24 | MicroChip32 | just did, same result |
12:55:31 | fontenot | you could use gdb to 1) load the core file and ask it where the segfault occurred, 2) step through the code |
12:55:32 | MicroChip32 | i assume that adds debug info to the core ? |
12:55:46 | fontenot | MicroChip32: -g just compiles in symbols - makes for easier debugging |
12:55:51 | fontenot | yes |
12:56:16 | fontenot | you get actual function and variable names instead of memory locations |
12:59:26 | MicroChip32 | gdb map core (map is prog name) after compile with -g |
12:59:37 | MicroChip32 | entered 'run' in gdb |
12:59:43 | MicroChip32 | get normal output except for ... |
13:00:01 | MicroChip32 | Program received signal SIGSEGV, Segmentation fault. |
13:00:01 | MicroChip32 | 0x4004012f in ?? () from /lib/libc.so.6 |
13:00:35 | MicroChip32 | how do i start it so i can step thru it totally ? |
13:03:49 | fontenot | i think you just type step |
13:03:58 | fontenot | you can examine variables etc |
13:04:13 | fontenot | there are some good graphical front ends for gdb |
13:04:28 | fontenot | because it's a pain to use from the command line |
13:04:29 | MicroChip32 | cept no x on my server |
13:04:45 | fontenot | i always have to read the help when I use gdb |
13:04:57 | MicroChip32 | yep, have man up in 2nd terminal ;) |
13:05:10 | fontenot | if you load the binary and type step, what happens? |
13:05:17 | fontenot | you can also set breakpoints |
13:05:29 | fontenot | that makes stepping a lot more convenient |
13:05:39 | MicroChip32 | yea i set a break on main so it stopped at the beginning and am stepping thru no |
13:05:40 | MicroChip32 | now |
13:05:52 | fontenot | cool |
13:06:09 | fontenot | try to examine the variables as you go when you think it makes sense |
13:06:33 | fontenot | i think you're corrupting something - probably a bad pointer or a non-terminated string |
13:06:53 | Niddix | Why would a PCMCIA nic driver loading lock up my tux? |
13:07:10 | MicroChip32 | yea thats what i figure too, the wierd thing is that this worked PERFECTLY a couple days ago, and i havent changed anything in the function in question |
13:07:31 | fontenot | MicroChip32: riiiiight. :-D |
13:07:47 | fontenot | MicroChip32: this is a delayed segfault |
13:07:59 | fontenot | MicroChip32: there may not even be a problem in that function |
13:08:27 | fontenot | MicroChip32: sometimes memory corruption is not detected until several funtions later |
13:08:38 | MicroChip32 | fontenot: it seg's at the end of the prog. if i comment out 2 lines of code it doesnt segfault. so the problem is definitly in those two lines, both which call this function |
13:08:53 | fontenot | MicroChip32: the fact that it's crashing in the libc only re-enforces that theory |
13:09:31 | fontenot | MicroChip32: that's likely true, but not 100% true |
13:09:49 | MicroChip32 | also intersting is that a 3rd line which is essentially identical to the first does NOT cause a segfault |
13:09:59 | fontenot | MicroChip32: the calling and running of that function may cause manipulations of the stack and memory which expose the corruption |
13:10:30 | MicroChip32 | hehe you sound like a politition talking about corruption like that ;) |
13:10:43 | fontenot | MicroChip32: I would check your pointer usage and string termination in the whole program |
13:10:57 | fontenot | MicroChip32: tho I dont know how big it is |
13:11:06 | MicroChip32 | small |
13:11:16 | fontenot | MicroChip32: are you using any other function libraries other than libc? |
13:11:43 | MicroChip32 | stdlib.h and math.h compiled with -lm |
13:12:14 | MicroChip32 | stdio.h i mean |
13:12:31 | MicroChip32 | string.h too |
13:13:40 | MicroChip32 | ok, took out math.h and -lm (which i dont need currently) no effect |
13:14:25 | Tangent | MicroChip32 : Are you sure that you're using atol correctly? |
13:14:33 | fontenot | the fact that you are only using proven libraries and very few points to an actual problem in your code |
13:14:46 | Tangent | MicroChip32 : If I'm understanding correctly, you're passing it just one char at a time |
13:14:57 | fontenot | heh, are you using it on a non-null-terminated string perhaps? |
13:15:20 | Tangent | MicroChip32 : but atol seems to handle whole strings... |
13:15:31 | fontenot | but hates non-terminated ones |
13:15:46 | fontenot | as do most of the libc functions |
13:16:00 | Tangent | MicroChip32 : I changed atol to just deduct 'A' instead (Convenient Ascii to Hex), and it doesn't segfault anymore in the get_int_field function |
13:16:07 | MicroChip32 | yes string is terminated, and the function returns the proper numeric value as shown by prints in my code. |
13:16:40 | fontenot | Tangent: isn't there a version of libc he can link to which has the symbols built in? |
13:16:45 | MicroChip32 | Tangent: change what where ? |
13:17:13 | Tangent | MicroChip32 : in functions.h... where you're calling atol |
13:17:14 | fontenot | in redhat there is a rpm called glibc-debug |
13:17:20 | MicroChip32 | Tangent: and no not passing atol a character by character |
13:17:46 | fontenot | Tangent: where are you seeing the code? |
13:17:55 | fontenot | did i miss a link? |
13:18:05 | Tangent | fontenot : MicroChip32 posted me a .zip |
13:18:10 | MicroChip32 | fontenot: yes you did http://home.mctech.org/pub/map.zip |
13:18:29 | fontenot | thanks |
13:18:55 | MicroChip32 | Tangent: the function puts a temporary null at the end of the fixed field and passes atol a pointer to the start of the fixed field |
13:19:26 | fontenot | ARg no unzip on this laptop |
13:19:31 | fontenot | crap redhat POS |
13:19:42 | fontenot | installs 100s of MB of crap i dont need |
13:19:46 | fontenot | and does not in stall unzip |
13:19:50 | MicroChip32 | hehehe |
13:20:15 | Niddix | Try gunzip |
13:20:32 | MicroChip32 | id tar.gz it but not familiar with tar/gz except for UN taring stuff |
13:20:59 | MicroChip32 | Tangent: back to your change, what did you change? from what to what ? |
13:22:03 | fontenot | localhost% gunzip map.zip |
13:22:03 | fontenot | gunzip: map.zip: unknown suffix -- ignored |
13:22:03 | fontenot | l |
13:22:09 | Tangent | MicroChip32 : I wans't entirely understanding what the code did.. but I changed the atol line... it won't do what it should now.. but it doesn't segfault in get_int_field anymore |
13:22:32 | MicroChip32 | lol, well that helps ;) i can do that by commenting them out ;) |
13:22:51 | MicroChip32 | and yea, i dont comment well do i ;) |
13:23:15 | Niddix | fontenot rename it to Z |
13:24:09 | Tangent | MicroChip32 : Comments are OK... but I've had to start from not knowing what the code is supposed to do... and work my way to understanding that you're reading stuff from a file, and splitting it up... (what's wrong with awk? ;) |
13:24:23 | fontenot | doesn't that only work on single files zipped? |
13:24:27 | fontenot | trying... |
13:24:52 | fontenot | localhost% gunzip map.Z |
13:24:52 | fontenot | gunzip: map.Z has more than one entry -- unchanged |
13:25:06 | fontenot | i need unzip |
13:25:09 | MicroChip32 | Tangent: not familiar with it, and would it work with fixed size and position fields within a string? |
13:25:11 | Niddix | hrmmm.. |
13:25:21 | fontenot | this sucks |
13:25:26 | fontenot | no cdrom drive |
13:25:35 | Niddix | Not sure why you didn't get it. |
13:25:37 | MicroChip32 | how would i gz the file guys and i'll to that |
13:25:42 | MicroChip32 | tgz |
13:25:45 | MicroChip32 | do |
13:25:46 | Tangent | MicroChip32 : It would work.. but it wasn't a serious suggestion |
13:25:55 | Niddix | Find unzip-5.41-3.rpm |
13:26:23 | fontenot | it's not a big deal |
13:26:26 | Niddix | MicroChip32: tar zcvf files.tar.gz blah |
13:26:27 | fontenot | :-\ |
13:27:23 | Niddix | MicroChip32: tar z(zip)c(create)v(verbose)f(file) |
13:28:23 | MicroChip32 | http://home.mctech.org/pub/map.tar.gz |
13:29:47 | fontenot | MicroChip32: getting file not found |
13:29:50 | MicroChip32 | oops, sec |
13:30:16 | MicroChip32 | try now |
13:30:46 | fontenot | got it |
13:32:23 | MicroChip32 | looks like two ppl grabbed it |
13:32:29 | fontenot | i see one problem right off the bat |
13:32:36 | MicroChip32 | which is ? |
13:32:55 | fontenot | you're doing an fgets(256 to a 256 element array |
13:33:04 | fontenot | where's the termination? |
13:33:40 | fontenot | might not be *the* problem, but that's exactly the sort of thing that gets one into this sort of trouble |
13:33:42 | MicroChip32 | hmmmmm i was gonna say that string is never printed, but it is for debugging |
13:34:02 | fontenot | you're printing it on the next line |
13:34:34 | MicroChip32 | changed fgets to 255, still segfaults |
13:34:50 | MicroChip32 | fontenot: yea, for debugging, making sure its reading the file |
13:36:49 | fontenot | MicroChip32: i just null terminated is and it still segfaults so that's not the current problem |
13:37:12 | fontenot | BTW you should compile with -Wall and see if you understand all the warnings |
13:40:37 | MicroChip32 | fixed two implicit declarations by including ctype.h and stdlib.h |
13:41:27 | fontenot | good |
13:41:52 | MicroChip32 | here's where i think the problem is, but not sure exactly how ... |
13:42:17 | MicroChip32 | in the types.h the 3 fields in question are defined as short, long long, and short; respectively |
13:42:40 | MicroChip32 | the get_int function returns a long (currently) |
13:43:02 | MicroChip32 | ive played with all the above types and always get the segfault |
13:43:20 | Tangent | MicroChip32 : I assume you tried atoll too |
13:43:20 | MicroChip32 | the values remain printed properly in my debug prints however |
13:43:43 | MacAtack | says Microchip32 is becoming a C god |
13:43:54 | MicroChip32 | ummm no, didnt know of it, but that would only be relevant to the 2nd field |
13:43:59 | MicroChip32 | MacAtack: working on it *lol* |
13:44:21 | MicroChip32 | helps to have 15+ years programming in other languages ;) |
13:44:34 | MacAtack | yeah |
13:44:47 | Niddix | MicroChip32: I doubt thats the problem. C is pretty good at automatically casting. |
13:44:56 | Niddix | Have you tried to recast the return value? |
13:45:21 | MicroChip32 | Niddix: thats what i figured. should i have the function return the biggest type possible and cast the return value as needed ? |
13:45:24 | MicroChip32 | or vice versa ? |
13:45:59 | MicroChip32 | aka use atoll, return a long long, and cast to short as needed ? |
13:46:19 | Niddix | MicroChip32: That would be preferable I'd think. |
13:46:34 | Niddix | MicroChip32: Otherwise you might lose something. |
13:46:45 | MicroChip32 | didnt work |
13:47:18 | MicroChip32 | besides which, its never seg'd on the 3rd get_int execution which is essentially identical to the first, which does seg |
13:47:32 | MicroChip32 | SPLIT! |
13:47:38 | Niddix | Let me reboot into Linux. |
13:48:17 | MacAtack | that was fun |
13:48:35 | MicroChip32 | they should sell tickets for those rides |
13:49:15 | MicroChip32 | argh, that daggone red worm is STILL probing me |
13:52:11 | MicroChip32 | ok i have it down to NO -Wall warnings |
13:52:12 | MacAtack | werkin your internet connecton to death to huh? heheheheh |
13:52:26 | MicroChip32 | that doesnt take much here *lol* |
13:52:49 | MicroChip32 | crap my modem balked at sending my zipped code to two people at once *lol* |
13:53:02 | scanline | is away: class |
13:53:02 | pattieja | is back (gone 14:06:45) |
13:53:23 | MacAtack | MicroChip32: send it to me and I'll post it on an oc-3 ftp site |
13:53:47 | MicroChip32 | hehe no biggie MacAtack, its not that big |
13:53:55 | MacAtack | ko |
13:54:02 | fontenot | MicroChip32: what dist are you using? |
13:54:15 | MacAtack | i know how much you like not eating up all of your b/w |
13:54:35 | MicroChip32 | rh6.2 kernel 2.2.19 |
13:54:55 | MicroChip32 | MacAtack: what bandwidth |
13:55:01 | fontenot | MicroChip32: rpm -q glibc-debug |
13:55:06 | MicroChip32 | MacAtack: we use dixie cups out here |
13:55:18 | MacAtack | ehee |
13:55:53 | MicroChip32 | fontenot: not installed |
13:56:09 | fontenot | MicroChip32: yeh, me neither :-\ |
13:57:38 | MicroChip32 | MacAtack: or on the good days we use bird calls |
13:58:04 | MicroChip32 | on the bad days its smoke signals |
13:59:13 | MicroChip32 | wants his own T1 |
14:00:26 | MacAtack | sayz sprint has a pretty sweet looking deal going on now.. Free Router with a 1 year commitment for the t-1 :) |
14:03:51 | prpplague | MacAtack: be sure and read the fine print |
14:04:15 | MicroChip32 | gets out his magnifrying glass |
14:04:38 | MicroChip32 | hmmmm says here that "we own you, your house, your wife, and your dog when you sign this paper" |
14:06:08 | Niddix | Not the dog. Thats going to far! |
14:06:27 | MicroChip32 | oh, i forgot the kids too |
14:06:30 | MicroChip32 | and the car |
14:10:40 | MacAtack | prpplague: yeah i already have it seems pretty straight forward... but they want people to forget they left thier customers high and dry for 11 days when thier network melted down a few years ago |
14:12:06 | prpplague | MacAtack: well this is one of my favorite site since i have to deal with verizon so often - http://www.verizonreallysucks.com |
14:13:10 | MacAtack | yeah most def... since were on the subject of companies that suck ever heard of Quicknet ? they've attempted to repair two of thier devices that i have bought and both times they have failed...so they want me to ship it back again for another round of repair attempts..at my expense (shipping) gettin annoyed |
14:13:38 | prpplague | MacAtack: what is it? router? |
14:13:49 | MacAtack | VoIP telephony devices |
14:14:26 | MacAtack | they have a PCMICA version that should work in the tux (so were not completely off topic here) but I highly doubt I will ever purchase one now |
14:16:25 | MacAtack | prpplague: that verizon site is pretty funny... i like the pic of the van |
14:17:10 | prpplague | MacAtack: it my understanding they vadilized that van and no one noticed for a month |
14:17:58 | prpplague | MacAtack: ann voip stuff |
14:19:01 | Niddix | MicroChip32: Ever figure out your problem? |
14:19:21 | MicroChip32 | nope, ive got rid of all -Wall warnings and still seg's |
14:20:02 | Niddix | Ok.. Your problem is in the line data = get_record_t1( line ); |
14:20:06 | MicroChip32 | Niddix: actually i know my problem ... its c |
14:20:11 | MicroChip32 | Niddix: how so ? |
14:20:12 | Niddix | Well its actually before that. |
14:20:30 | Niddix | Your dumping your RT1 struture into thin air. You need to allocate memory for it. |
14:21:23 | Niddix | Add the line... data = ( RT1 * ) malloc( sizeof(RT1) ); right before that line. |
14:21:40 | MacAtack | prpplague: ROTFLAMO that is sooo funny |
14:21:44 | MicroChip32 | better idea yet ... |
14:22:05 | MicroChip32 | Niddix: in the funtion get_record_rt1 ... |
14:22:42 | MicroChip32 | Niddix: change the def of rec to RT1 rec, instead of RT1 *rec, then return &rec instead of return rec; |
14:23:54 | MicroChip32 | but still, good catch |
14:24:20 | Niddix | Or you could just put the RT1 *data; inside main(); |
14:25:11 | Niddix | MicroChip32: Nevermind.. that still wouldn't work. |
14:25:21 | MicroChip32 | what good would that do w/o one of the other things |
14:25:44 | MicroChip32 | but youre right, it probably should go inside main since its not necessary to be global |
14:26:11 | MicroChip32 | but note that most of main() is testing stuff right now |
14:27:02 | Niddix | MicroChip32: I'm not positive, but I think if you do what you were saying you might lose your memory structure due to going out of scope. |
14:27:03 | MicroChip32 | hmmm -Wall doesnt like it when my function returns the address of its local variable. is that a big deal ? |
14:27:24 | CosmicPenguin | MicroChip32: Probably..... |
14:27:28 | MicroChip32 | hmmmmmmm |
14:28:04 | CosmicPenguin | MicroChip32: Did you alloate the memory, or is it just a regular local variable on the stack? |
14:28:17 | MicroChip32 | regular variable, not pointer/malloc |
14:28:28 | CosmicPenguin | MicroChip32: Definately don't do that.... :) |
14:28:36 | MicroChip32 | CosmicPenguin: ok, how about this ... |
14:28:36 | Niddix | MicroChip32: Yeah then it will poof when you leave the function. |
14:28:58 | Niddix | MicroChip32: You might actually get away with it for a little while.. |
14:29:02 | MicroChip32 | in the func define the var as pointer, malloc, and return the var/pointer ... will that poof ? |
14:29:15 | CosmicPenguin | MicroChip32: That won't poof.... |
14:29:20 | Niddix | MicroChip32: That'll workl |
14:29:25 | MicroChip32 | cool, that will suit my purpose then ;) |
14:29:32 | CosmicPenguin | Niddix: The fun thing about those bugs, is they they don't poof until some undetermined time later.... |
14:29:42 | Niddix | CosmicPenguin: No kidding. |
14:29:46 | MicroChip32 | CosmicPenguin: yea, we noticed |
14:29:55 | Niddix | CosmicPenguin: Then are are fun to track down. |
14:30:00 | MicroChip32 | Niddix and all those who helped ... many many thanks |
14:30:18 | MicroChip32 | hates c |
14:30:38 | MicroChip32 | is remembering why |
14:30:41 | Niddix | MicroChip32: C gives you enough rope to hang yourself. |
14:30:49 | pattieja | morning |
14:30:56 | MicroChip32 | Niddix: and helps you tie the knot too |
14:30:56 | Niddix | MicroChip32: It'll even tie the knot for you. |
14:31:01 | MicroChip32 | rotfl gmta |
14:31:52 | pattieja | anybody know anything about the Compaq iPAQ Home Internet Appliance? |
14:31:55 | CosmicPenguin | Niddix, MicroChip32: C does all this while at the same time assuring you that everything is going to be fine.... |
14:32:00 | Niddix | Now if I could just figure out why my tux locks when initializing the nic.... |
14:32:08 | pattieja | TigerDirect is selling them for $99.99 |
14:32:23 | CosmicPenguin | pattieja: Strong arm based? |
14:32:31 | pattieja | how should I know? :) |
14:32:35 | prh | no |
14:32:41 | prh | tangent was looking at those yesterday |
14:32:45 | Tangent | AMD K6 266 based |
14:32:46 | MicroChip32 | Niddix: malloc( rec, sizeof(rec) ); right ? |
14:32:52 | prh | ahh there he is ;O) |
14:33:03 | pattieja | It says it has support for Java VM, Windows Media Player, and ActiveX controls, so I would think it's running some form of Windows |
14:33:15 | Tangent | prh : Yeah..just lurking.. doing the jobserve thing |
14:33:20 | CosmicPenguin | pattieja: Of course.... Compaq is very Windoze friendly.... |
14:33:22 | prh | shudders at the thought of running windows |
14:33:31 | pattieja | does this mean that it could be hacked to run Linux? |
14:33:34 | prh | tang: any sign of the right thing? |
14:33:48 | CosmicPenguin | pattieja: Of course.... given the right bits of informatoin.... |
14:34:08 | Tangent | pattieja : Very probably... but there's no touchscreen, and only one CF slot, so it's a bit more limited than the Tux, inspite of the better screen and extra memory |
14:34:30 | MicroChip32 | Niddix: actually rec = malloc( sizeof(rec) ); |
14:34:34 | CosmicPenguin | Tangent: USB? Sounds like a pretty good thin client to me..... |
14:34:41 | pattieja | Tangent: right, but I just want a thin-client for my sister (or any of her friends that come over and always use my computer :) |
14:34:52 | Tangent | CosmicPenguin : Yeah..I forgot about USB actually |
14:35:16 | prh | no touchscreen |
14:35:26 | pattieja | the keyboard is wireless and has a builtin pointer. ooooohhhhh! |
14:35:29 | CosmicPenguin | Tangent: Awesome... Toss rdesktop on there, and you've got yourself a pretty cool system that your sister can use too...... |
14:35:40 | prh | hmm |
14:35:42 | pattieja | rdesktop? |
14:35:46 | pattieja | I don't have WTSE |
14:35:52 | pattieja | or Citrix |
14:36:00 | prh | can't you serve X as a WTSE session? |
14:36:01 | MicroChip32 | ha! got it ! |
14:36:01 | pattieja | nor do I have any desire to mess with them |
14:36:07 | Tangent | CosmicPenguin :I have no shortage of computers for my sister to use.. just a shortage of seating |
14:36:19 | CosmicPenguin | pattieja: Then use rdesktop........ |
14:36:26 | pattieja | huh? |
14:36:27 | CosmicPenguin | Most of your newer PCs will serve RDP |
14:36:41 | CosmicPenguin | s/PC/Windoze/ |
14:36:45 | pattieja | I pretty much only use Linux at home |
14:37:02 | Tangent | CosmicPenguin : Only if you use Server versions of the stuff |
14:37:05 | pattieja | I thought rdesktop was a terminal server connection protocol client |
14:37:08 | CosmicPenguin | Tangent: True, true.... |
14:37:23 | pattieja | why not go free and use VNC? |
14:37:30 | CosmicPenguin | pattieja: Thats good too..... |
14:37:32 | pattieja | ooh. that rhymed |
14:37:40 | pattieja | (however you spell that) |
14:37:44 | pattieja | T-H-A-T |
14:37:45 | pattieja | :) |
14:37:46 | pattieja | hehe |
14:38:17 | pattieja | If I get one of these, I just don't want it to turn out like the Audrey's I bought |
14:38:21 | Tangent | pattieja : You really don't want to get me started with poetry |
14:38:32 | CosmicPenguin | pattieja: No luck on hacking the Audrey then? |
14:38:34 | pattieja | Tangent: would that get you off on a different tangent? |
14:38:47 | pattieja | CosmicPenguin: not as yet. Haven't messed with it for quite some time. |
14:39:08 | prh | whats an audrey? |
14:39:09 | CosmicPenguin | pattieja: Do you know anyone that works at 3Com that can give you some pointers? |
14:39:12 | | audrey is http://www.audreyhacking.com/ |
14:39:12 | CosmicPenguin | ibot:audrey |
14:39:18 | pattieja | Maybe I can get my uncle to look at it. He could probably isolate the onboard flash to give me access to flashing it (possibly) |
14:39:31 | pattieja | CosmicPenguin: not a one |
14:39:39 | prpplague | pattieja: we've order a couple of the ipaq units to see if we can put out tcs software on it |
14:39:45 | MicroChip32 | woohoo, it works again |
14:39:51 | pattieja | prpplague: sweet |
14:39:52 | MicroChip32 | runs circles around his desk |
14:39:56 | CosmicPenguin | prpplague: Is it CF based or flash based? |
14:40:03 | pattieja | MicroChip32: curly? |
14:40:10 | Tangent | pattieja : I've done quite a few like this now... http://www.handhelds.org/pipermail/intimate/2001-July/000074.html |
14:40:12 | pattieja | prpplague: have you gotten any in yet? |
14:40:22 | prpplague | CosmicPenguin: don't know haven't got them yet |
14:40:23 | MicroChip32 | pattieja: eh? |
14:41:21 | Tangent | pattieja : Here's another... http://www.handhelds.org/pipermail/ipaq/2001-June/007453.html |
14:41:53 | pattieja | Tangent: that's cool |
14:42:24 | pattieja | MicroChip32: 3 stooges. Curly. He did the funniest thing like running around on the floor. It was hilareous. |
14:42:38 | prpplague | Tangent: cute |
14:42:44 | MicroChip32 | pattieja: ahhhhhh |
14:42:45 | Tangent | prpplague : Thanks |
14:42:52 | Tangent | spse : Afternoon |
14:43:08 | Tangent | spse : Before you ask... The usual.. and yes... and yes |
14:43:25 | MicroChip32 | pattieja: im just happy cuz Niddix and others helped me find a c bug thats been 'bug'ing me for 2 days |
14:43:46 | pattieja | MicroChip32: I know somewhat how you feel. :) |
14:45:31 | prpplague | MicroChip32: same here |
14:48:28 | Niddix | Any PCMCIA guru's on? |
14:49:27 | pattieja | dah! all compaq's links for specifications, upgrades & options, software, warranty are broken on their Home Internet Appliances IA-1 site |
14:51:12 | prpplague | Niddix: i'm a pcmcia wannabie guru does that count? |
14:51:45 | prpplague | pattieja: i searched for about 2 hours yesterday an didn't find anything |
14:52:04 | Niddix | prpplague: If you can tell me what how to debug my tux locking up when I stick my nic in, I'll crown you. |
14:54:38 | pattieja | prpplague: when do you think your shipment of iPAQ's will arrive? and which model did you order? |
14:54:46 | pattieja | and from where? |
14:55:07 | prpplague | wed |
14:56:47 | prpplague | howdy mallum |
14:57:35 | prpplague | mallum: saw a couple of your posts on the linux-arm-kernel, said to myself, "i know that guy!" |
14:58:04 | mallum | prpplague: heh |
14:58:20 | mallum | prpplague: posts like 'Hi, I really dont know what Im doing :) ' |
14:59:32 | prpplague | mallum: you know when i see your posts, in my minds eye i see that planet of the apes drawing on your web page |
15:00:08 | pattieja | prpplague: have you seen jailbait? |
15:00:16 | pattieja | it's a distro of Linux that fits in 16MB |
15:00:30 | pattieja | includes Netscape 4.73 and ishmail for imap mail and vncviewer |
15:00:34 | mallum | prpplague: heh. was it the posts about the CATS board ? |
15:00:52 | mallum | prpplague: I like the way russ king is so aggressive on the list |
15:01:14 | prpplague | pattieja: ya i tried it but didn't have much luck, the buildroot system that the tux uses is a far easier and configurable system for build an easy rootfs |
15:01:52 | prpplague | mallum: ya i think that was what it was, i remember rmk being a little condesending |
15:01:54 | pattieja | well, I saw it on the Audrey website and thought about it for the IA-1 |
15:04:00 | prpplague | pattieja: it just wasn't felixble enough |
15:04:14 | pattieja | prpplague: k |
15:04:16 | mallum | prpplague: this was arm-newbie, I dont think rmk is on that list |
15:04:20 | MicroChip32 | g'mornin kernelwork |
15:04:40 | pattieja | prpplague: I found a host of bbs entries on linux-hacker.net for the IA-1 iPAQ |
15:04:59 | pattieja | They're under the MSN Companion Other I-Appliances links. |
15:06:43 | CosmicPenguin | pattieja: Whats the news? |
15:08:47 | pattieja | CosmicPenguin: you can overclock the IA-1 to 320MHz |
15:08:50 | pattieja | from 266MHz |
15:08:52 | pattieja | hehe |
15:09:59 | pattieja | sweet! |
15:10:08 | pattieja | people have been booting the thing off the CF card slot |
15:13:03 | pattieja | prpplague: I'm seeing entries that people have gotten Jailbait to work on the IA-1 no problem (and then others were having problems, but...) |
15:13:54 | mallum | pattieja: what is jailbait ? |
15:14:04 | pattieja | an embedded distro of linux |
15:14:10 | pattieja | jailbait.sourceforge.net |
15:14:45 | kernelwork | Really? I though jailbait was something else... |
15:14:58 | kernelwork | has a friend going out with some jailbait |
15:15:16 | mallum | kernelwork: hehe , I did wanna say ..... |
15:15:41 | kernelwork | Is it linux for young people? |
15:16:07 | kernelwork | They could call it JB Linux |
15:17:02 | mallum | wonders how they fit netscape navigator into 16megs |
15:17:05 | AlHaz | jailbait is an old, poorly done embedded linux distro |
15:17:17 | AlHaz | which hasn't been updated in over a year |
15:17:34 | pattieja | netscape by itself will fit in 7MB |
15:18:10 | mallum | AlHaz: yeah I noticed they are using full X too |
15:18:54 | AlHaz | mallum: and 2.4.0-pre10, tho that should be easy enough to replace |
15:19:16 | prh | erm - whoever was talking about rdesktop - how do I tell my w2k box to serve it's desktop? |
15:19:32 | mallum | AlHaz: I guess it X86 only too , if its using netscape and XF86_SVGA |
15:21:07 | CosmicPenguin | prh: I guess you need a w2k server..... thought I thought w2k had RDP built in.... |
15:21:16 | CosmicPenguin | <-- Happy to not be a Windoze hacker... |
15:21:22 | prh | but how do you turn it on? (me too :O) |
15:21:48 | CosmicPenguin | AlHaz: <poorly done embedded linux distro> -- Sounds like the happy words of an Embeddix man... :) |
15:22:18 | AlHaz | CosmicPenguin: well, sure, I work there and all . . . |
15:22:41 | Tangent | CosmicPenguin : RDP is only on the server versions.. it's true |
15:22:47 | AlHaz | I installed JB just once. and regretted it. |
15:22:49 | CosmicPenguin | AlHaz: I'm excited to see the Sharp, even though they use our competitors software.... |
15:23:17 | AlHaz | CosmicPenguin: I wish i were more excited to see it ship, i think we could have done better if the japan office had cooperated with lindon |
15:23:49 | CosmicPenguin | Tangent: I can never tell the difference between w2k server and client. After all, I am still getting nimda hits... I can't believe that all those machines are servers.... |
15:23:50 | AlHaz | it's still nice hardware tho, and I'm still getting over the shock of seeing my company's name mentioned in someone elses product docs |
15:23:55 | mallum | CosmicPenguin: you've got your hands on one of them sharps ? |
15:24:24 | CosmicPenguin | mallum: Nope... We're direct competitors of QT, so I doubt we will be getting any free samples... :) |
15:24:25 | Tangent | CosmicPenguin : I think the differences are.. .a.) The splashscreen. b.) RDP. c.) Price |
15:24:26 | MicroChip32 | CosmicPenguin: im still seeing nimda hits too, altho far fewer than originally |
15:25:15 | CosmicPenguin | Tangent: Changing those splashscreens are expensive..... :) |
15:25:43 | Tangent | CosmicPenguin : Yep.. floodfill on the background uses a lot of expensive CPU cycles.. |
15:26:23 | CosmicPenguin | Tangent: Yeah, but wouldn't a company as sophisticated as M$ use advanced assembly to write the blit routines? |
15:26:36 | MicroChip32 | CosmicPenguin: hahaha yea right |
15:26:39 | CosmicPenguin | Tangent: Oh, wait... they just want to you upgrade your machine..... |
15:26:41 | MicroChip32 | they prolly use vb to do it |
15:26:45 | Tangent | CosmicPenguin : s/assembly/visual basic/ |
15:26:50 | CosmicPenguin | Tangent: lol |
15:26:56 | MicroChip32 | shakes his head... gmta |
15:27:18 | AlHaz | it's probably in ring 0 anyhow . . . |
15:28:50 | AlHaz | CosmicPenguin: So, are you prepared for the awsome power of Embedded XP? |
15:29:11 | CosmicPenguin | AlHaz: Yep...... All 150MB of it.... |
15:29:36 | AlHaz | hehe |
15:29:37 | Tangent | Embedded XP? Seriously? |
15:29:42 | AlHaz | Tangent: Seriously |
15:29:46 | Tangent | in concrete? |
15:29:52 | Tangent | ;) |
15:29:52 | AlHaz | they want it to replace CE, don't they? |
15:30:19 | CosmicPenguin | Tangent: I'll probably be holding a copy in my hands on Sunday |
15:30:37 | CosmicPenguin | And I kid you not.... the smallest they have gotten it down to is 150MB |
15:30:44 | Tangent | When will M$learn to take their immense fortune, and just go away |
15:31:06 | mallum | Really, the reason |
15:31:07 | mallum | you see open source there at all is because we came in and said there should |
15:31:08 | mallum | be a platform that's identical with millions and millions of machines," - bill gates |
15:31:15 | Tangent | 150MB doesn't really cut it... that's the sort of think I'd piece together for an OS ;) |
15:31:28 | Tangent | s/thik/thing |
15:32:12 | Tangent | mallum : Hrm... yes... what was that file format on the Amiga... ah yes.. Interchangable File Format.. .worked on all differnt apps, systems... etc... except Windows |
15:32:54 | mallum | Tangent: heh, there was a big thread of stuff like that on the gllug list ( where that was posted ) |
15:34:06 | Tangent | starts a list... every time this windows box pisses me off I'll put it on the list.. and when the list is 20 lines long, I'm gonna format it and drop linux on |
15:34:47 | Tangent | The sad thing is that I'll still count as a windows user for their statistics, as I have a opened copy of windows that I can't legally sell (and fortunately didn't have to buy either) |
15:34:59 | spse | Tangent: so you'll have linux on it by 6pm then ;) |
15:35:54 | Tangent | spse : Sadly not.. the application containing the shit list just crashed and lost all it's files |
15:35:58 | Tangent | ;D |
15:36:18 | mallum | Tangent: I gave my girlfriend my old viao picturebook and put KDE on it |
15:36:51 | mallum | Tangent: she thinks its fantastic, cos the cards in ksolitare animate much better than windows solitaire ( and she can change there colors ) |
15:37:45 | Tangent | mallum : My problems with linuxing this machine are... a.) Will my soundcard work... _properly_ with surround etc... 2.) Will my WinTV Card work... 3.) Will my raid card work.. 4.) Will I loose my 160GB collection of movies and mp3s? 5.) Will I ever get it working right under linux or will I just break it every two days with a new kernel |
15:38:35 | AlHaz | the wintv card will work, at least |
15:38:35 | spse | i'll all work ok, plus you can watch dvd's on the second head of the matrox ;) |
15:38:43 | Tangent | mallum : Sigh.. if only _I_ was that easily pleased :) |
15:38:51 | CosmicPenguin | spse: mmm... dual heads..... |
15:39:01 | AlHaz | the raid card depends . . . most of the ide raid cards aren't really raid by a longshot |
15:39:15 | Tangent | AlHaz : Nah.. it's not real raid.. it's a promise card |
15:39:20 | AlHaz | ah |
15:39:25 | spse | the promise is, cos its software raid |
15:39:34 | Tangent | AlHaz : But I need a kernel that supports it, on a bootable CD so that I can reformat my drives |
15:40:08 | Tangent | spse : Didn't you tell me the other day that every piece of DVD playing software you found failed to work on linux |
15:40:19 | Tangent | Hmm.. and that's the other thing.. how about DVD ripping? |
15:40:32 | spse | video lan client works fine |
15:40:48 | AlHaz | MPlayer works pretty well |
15:41:02 | Tangent | spse : How could I live without graphedit? |
15:41:06 | spse | yes, but for dvd playing the interface is still being worked on |
15:41:48 | mallum | Tangent: I got a kt7 mother board with raid and ata100 working on my desktop at home |
15:41:49 | spse | plus I looked at the mplayer code and it looks like it has been written by 13yr old demo coders, it works well but i dont think they have heard of indenting ;) |
15:42:02 | CosmicPenguin | xine rules! |
15:42:12 | CosmicPenguin | For everything except Disney.... |
15:43:09 | prpplague | CosmicPenguin: we don't do disney at my house... |
15:43:15 | mallum | Tangent: havn't you got debian on your laptop ? |
15:43:42 | prpplague | taught his 4 year old son to say "disney and m$ suck!" |
15:43:54 | Tangent | mallum : Yep.. got debian on laptop.. tuxscreen, ipaq, firewall and dev box.. +VMware copies on Windows boxes |
15:44:07 | AlHaz | I can't stand xine . . . |
15:44:22 | mallum | Tangent: well do you really need to install Linux on your big box then ? |
15:44:24 | AlHaz | for one it seems to confuse chapters with .vob files |
15:44:25 | Tangent | prpplague : Well done :) |
15:44:26 | spse | i found the dvd nav on xine a bit broken |
15:44:31 | CosmicPenguin | prpplague: The GF loves disney.... We always take the laptop on trips, and she tried to load up Tarzan, but Xine wouldn't handle it |
15:44:32 | AlHaz | which isn't always the case |
15:44:56 | prpplague | Tangent: he love to play xbill |
15:45:07 | Tangent | mallum : It's not that I _need_ to.. It's just that M$ are pissing me off with their press releases, and I really want to have nothing to do with them |
15:45:10 | mallum | has Tron on DVD |
15:45:18 | CosmicPenguin | spse: Yeah, I can't play my simpsons DVDs on Xine because of the navigation |
15:45:38 | mallum | Tangent: I know !, sell the box and buy a big box, stcik OSX on it :) |
15:45:50 | mallum | s/big box/big mac/ |
15:46:08 | CosmicPenguin | mallum: I would prefer Yellowdog personally..... |
15:46:14 | Tangent | mallum : Certainly not... not enough mouse buttons ;) |
15:46:26 | mallum | CosmicPenguin: debian runs on ppc too |
15:46:46 | mallum | Tangent: I have an ms mouse on my mac under osx working fine with more buttons |
15:46:53 | mallum | Tangent: though I am selling my mac |
15:46:57 | AlHaz | well, at least with desktop macs you can throw away the little mouse they come with and plug in a normal usb mouse |
15:47:17 | mallum | Tangent: cos I never use it and its not really powerful enough for OSX and the keyboard sucks |
15:47:22 | AlHaz | powerbooks would look real nice to me if they had another mouse button. oh, and a trackpoint instead of a greasepad. |
15:47:33 | mallum | AlHaz: yeah Id like a power book |
15:47:41 | Tangent | mallum : You want to leave that last sentence out of the advert :) |
15:47:53 | CosmicPenguin | Aelhaz: Embeddix won an award from Linux Journal... :) |
15:48:03 | CosmicPenguin | mallum: I am debian free? |
15:48:04 | AlHaz | CosmicPenguin: That's nice |
15:48:20 | AlHaz | CosmicPenguin: this just recently? |
15:48:28 | mallum | CosmicPenguin: why the '?' ? |
15:48:28 | AlHaz | I know they liked the SDK, which was good to hear |
15:48:35 | CosmicPenguin | AlHaz: Just read it on LinuxDevices.... |
15:48:38 | AlHaz | ah |
15:48:41 | AlHaz | I'll go look later |
15:48:46 | CosmicPenguin | mallum: s/?/!/ |
15:49:00 | CosmicPenguin | Which I read on my Ipaq via bluetooth..... :-) |
15:49:01 | mallum | CosmicPenguin: whys that then ? |
15:49:21 | CosmicPenguin | mallum: I dunno... mainly to piss off kernelwork (who is a debian fanatic) |
15:49:27 | AlHaz | CosmicPenguin: Sorta surprising, since LD spends so much time toadying up to MontaVista |
15:50:27 | CosmicPenguin | AlHaz: No comment... I know another certain company that seems to get more then their fair share of press from Rick Leherbaum...... :) |
15:51:39 | Aelhaz | CosmicPenguin: heh |
15:52:01 | Aelhaz | Well, montavista does do one thing that lineo doesn't that gets them some recognition. |
15:52:12 | mallum | AlHaz: is embeddix like a superchrged buildroot ? |
15:52:17 | Aelhaz | you can actually go to an ftp site and download a binary distribution that will actually boot to a prompt and do stuff |
15:52:19 | CosmicPenguin | Aelhaz: Real Time? |
15:52:25 | CosmicPenguin | Aelhaz: Ahh.... |
15:53:06 | Aelhaz | CosmicPenguin: hey now, we actually ship RTAI with some sdks, which is actually hard real time. MV just has that preemptable kernel. |
15:53:21 | CosmicPenguin | Aelhaz: I haddn't heard..... |
15:53:39 | Aelhaz | doesn't matter, hardly anyone actually needs hard real time |
15:54:18 | Aelhaz | but lots of people think they do, so you have to offer it |
15:55:03 | CosmicPenguin | Aelhaz: Hey, when that pr0n server responds, I don't want be be screwing around with some other process,right? |
15:55:13 | Aelhaz | hehe |
15:55:35 | Aelhaz | Well, with RTAI, we can do in software, in linux, things that are usually done in ASICs |
15:55:57 | Aelhaz | it's up to the customer to decide if that's a good thing or not . . . |
15:56:58 | prpplague | CosmicPenguin: you guys know rick very well? |
15:57:09 | CosmicPenguin | prpplague: Our director of marketing does...... |
15:59:46 | prpplague | CosmicPenguin: rick is supposed to get john lombardo to do a review of our tcsx-1 unit |
16:01:57 | mallum | What exactly does a premptive kernel do ? in what situations is it useful ? |
16:02:50 | Aelhaz | mallum: anyway, yeah, the embedix sdk is similar to buildroot, but more versatile, with a gui, and a hundred some packages floating around the base installation |
16:03:01 | Aelhaz | well, MV has this kernel patch that makes linux more preemptable |
16:03:23 | Aelhaz | that means that whena high priority task needs something done right now, it can preempt whatever the kernel is doing and get it done, and then the kernel can pick up where it left off |
16:03:25 | Niddix_away | Anyone want to place any bets that this Netgear FA411 I just bought will work? |
16:03:46 | Aelhaz | Niddix_away: yes, it oughta work |
16:03:51 | Aelhaz | didn't someone already try one? |
16:04:02 | mallum | Aelhaz: ok I get you , so its sort of like an RT kernel ...? |
16:04:07 | Niddix_away | Aelhax: Yeah.. however.. |
16:04:23 | Aelhaz | mallum: not really |
16:04:38 | Aelhaz | mallum: in RT, the right task at the wrong time is the wrong task |
16:04:54 | mallum | Aelhaz: ok |
16:05:05 | Aelhaz | mallum: so with hard real time, you're looking to get something done at a particular time, with microsecond accuracy |
16:05:16 | Niddix_away | Wow.. it worked. |
16:05:19 | Niddix_away | I'm amazed. |
16:05:54 | CosmicPenguin | Niddix: Netgear? Ack! |
16:06:01 | Aelhaz | mallum: being able to preempt the kernel resolves a lot of latency issues, but it still doesn't guarantee that something gets done at a particular time |
16:06:12 | Niddix | CosmicPenguin: Whats wrong with Netgear? |
16:06:25 | Aelhaz | preemptability and HRT are actually complementary |
16:07:11 | Aelhaz | the way most (all?) linux real-time solutions work is to use a sort of microkernel between the system and the linux kernel |
16:07:21 | Aelhaz | so the kernel actually runs as a client task of the real time system |
16:07:35 | Aelhaz | thus victor yodaiken's hilarious patent |
16:07:37 | mallum | Aelhaz: yep |
16:07:58 | Niddix | Of course, now I'm not getting a login. |
16:08:04 | Aelhaz | So, in that case, the kernel gets blocked whether it likes it or not, and it's generally OK with that, but it would be nice if it were built to be preempted |
16:08:48 | CosmicPenguin | Niddix: That NSC chip isn't the best... or rather, the driver for that NSC chip |
16:09:24 | Niddix | CosmicPenguin: Will it work semi-reliably? |
16:09:37 | Niddix | CosmicPenguin: Its not like I'm planning to host a website on my tux. |
16:10:52 | mallum | Aelhaz: thanks for explaining that all to me |
16:11:47 | Aelhaz | OS/2's kernel is fully preemptable, and IBM used to sell an almost-hard-real-time addon for it. I think if there was something like that for linux it could do a lot of things for people, but the marketing buzz around "real time" has everyone convinced they need hard real time |
16:12:03 | mallum | Aelhaz: what sort of systems/apps would work better with a preemptive kernel ? |
16:12:04 | Aelhaz | when in most cases they just need low latency, or would be better off with an asic or dsp anyhow |
16:12:35 | Aelhaz | mallum: audio editing & mixing is a common use for it |
16:12:55 | Aelhaz | 'course, in MacOS, they just clobber the rest of the OS to get everything done |
16:13:07 | Aelhaz | you can get away with that on cooperative multitasking systems . . . |
16:13:12 | mallum | Aelhaz: cool, yeah of course - it would be cool in audio |
16:13:37 | mallum | Aelhaz: getting tracks synced up and the like |
16:14:00 | Niddix | How up to date is the buildroot in downloads/v0.4? |
16:14:13 | Aelhaz | I found that in os/2, unless something wedged the input queue (which was a major problem, it had a synchronous input queue), one app generally couldn't make the rest of the system unusable, which was nice |
16:14:34 | CosmicPenguin | Niddix: It *should* work ok, but if it freaks on you, you'll know why.... |
16:14:54 | Niddix | nod.. well its freaking right now. |
16:15:02 | Aelhaz | in hard real time systems you actually end up with a problem they call priority inversion |
16:15:09 | Niddix | Either that or its taking forever to timeout. |
16:15:22 | Aelhaz | where a high priority task is waiting on a low priority task to get something done, which means a middle-priority task never happens |
16:15:46 | Aelhaz | even though you'd rather have it than the low priority task |
16:16:30 | mallum | Aelhaz:I gotta shoot - hometime - thx again for the explination |
16:16:32 | mallum | laters all |
16:17:10 | scanline | is back (gone 08:14:44) |
16:20:28 | Niddix | ls |
16:20:36 | Niddix | blushes |
16:21:51 | prh | O) |
16:22:26 | prpplague | Niddix: hey didn't get a chance to answer you about the pcmcia, i haven't got that far yet i'm still working on learning the code for pcmcia controlers |
16:22:56 | Niddix | prpplague: Thats alright. I went out and bought a card thats known to work. |
16:23:12 | Niddix | prpplague: The LinkSys came out of my wifes laptop, was just trying it. |
16:24:15 | Niddix | Of course my known to work card isn't working.. but I'm sure its just me. |
16:24:43 | Niddix | The drivers init.. but I doesn't seem to be configuring right. |
16:55:00 | CosmicPenguin | Has anyone ever tried to compile SSH on uclibc? |
17:09:19 | pattieja | prpplague: gonna go for an IA-1 :) |
17:11:05 | MicroChip32 | kernelwork: are you around ? |
17:15:54 | kernelwork | Yeah, but I'm working |
17:16:34 | MicroChip32 | heheheh well just wanted to comment on my progress ... i convinced perl to write the code for my typedefs and the line parsing functions for me |
17:18:21 | kernelwork | cool |
17:18:38 | MicroChip32 | save a crapload of typing ;) |
17:18:44 | kernelwork | I'm told I may be waiting until tuesday at the soonest to get DSL again |
17:18:50 | MicroChip32 | ewwww |
17:19:13 | kernelwork | This is going from a 4 day outage to a 2 week outage at the rate Qwest is going. |
17:19:37 | MicroChip32 | sounds like my deal this summer |
17:19:40 | kernelwork | I consider Qwest as responsible for the very fact that they just can't switch my providers in a timely fassion |
17:20:08 | kernelwork | They are providing line, but they aren't connecting service. |
17:27:13 | prpplague | pattieja: whats that? your going to buy a IA-1? |
17:34:11 | prpplague | anyone know what a java .obdl file is? |
17:44:47 | pattieja | prpplague: yep. Just ordered over the phone. Hung up in the credit deparment... Waiting... |
17:45:29 | pattieja | couldn't order online for some reason. Tiger's pages kept crashing both Mozilla and Netscape (under Linux) |
17:55:53 | prpplague | pattieja: well will see how well ltsp works on it |
18:44:14 | pattieja | prpplague_lunch: yep |
18:52:46 | as_ | Evnin all. |
18:53:37 | Tangent | as_ : Hi... You're up early |
18:54:02 | prpplague | as_: had a cup of joe yet? |
18:54:36 | as_ | I am up early ;] |
18:54:48 | as_ | I'll be going to sleep early as well :/ |
18:55:27 | as_ | Had a new sat system installed, guy turned up 4 hours after I went to sleep, screwed my nigh day cycle. |
18:56:59 | as_ | Damnit why dont they turn up when there supposed too. |
18:57:36 | Tangent | as_ : They must make you suffer because they have a cruddy life drilling holes in the outside of peoples houses in the pouring rain |
18:58:04 | as_ | He didnt drill any holes, he ripped the old system out and used the same holes. :/ |
18:58:09 | as_ | Grrr. |
19:59:27 | TimRiker | heh, dsl is up.. bbiab |
20:48:07 | Russ|werk | where are the latest wheaties/buzby drivers? |
20:48:29 | GoRK | Russ: the latest wheaties has timing problems. the delays arent long enough |
20:49:14 | GoRK | Russ: the version that im using doesnt even have the serial flash support though... so i dont know what the best version is right now... the latest have some other good changes... |
20:50:05 | GoRK | Russ: have any idea what is making opening the touchscreen hard lock? it looks like lock_kernel() is called and the ucb1x00 thread never returns |
20:50:19 | Russ|werk | hmm... |
20:50:43 | Russ|werk | did you check for more machine_is_ stuff |
20:50:45 | Russ|werk | ? |
20:52:02 | GoRK | yep its only stuff in ucb1x00-core... you have a backlight machine_is in ucb1x00-ts.c but it's not the problem |
20:52:16 | SteveD | Tangent: notice you didn't put audio in/out sockets etc on your shanip232 board? |
20:53:43 | GoRK | his board is the straight shanip2 circut. mine (which has sockets) is at http://www.blurbco.com/~gork/touchscreen/shanip2-gork7.gif |
20:54:13 | GoRK | if you are only ever going to be doing voip stuff you're not going to need the full board. the sockets, volume control, etc will also raise the cost about 10$ |
20:55:50 | SteveD | GoRK: my existing Tux seems to have a broken Mongoose so I will probably just connect ucb1200 speaker/mic straight to speaker/mic. I've got another coming and I expect I'll just make up the simple passive cct on some plan perfboard |
20:55:51 | Russ|werk | GoRK: did you have one of those tuxscreens that used to lock on ts opening? |
20:56:12 | GoRK | Russ: no, mine always worked |
20:56:28 | GoRK | Russ: should i comment out the lines about resetting the codec and try? |
20:57:08 | Russ|werk | try this really quick |
20:57:19 | Russ|werk | ucb1x00-core.c |
20:57:24 | Russ|werk | - else if (machine_is_shannon()) { |
20:57:32 | Russ|werk | + if (machine_is_shannon()) { |
20:57:43 | GoRK | i had already done that |
20:58:01 | GoRK | whoops forgot to tell you that one... it didnt seem to do anything but i did make that change |
20:58:06 | Russ|werk | maybe put a sleep in after the reset of 100ms |
20:58:40 | GoRK | after the codec reset? in the other ifdef? |
20:58:56 | Russ|werk | /* reset the codec */ |
20:59:00 | Russ|werk | after that bit of code |
20:59:15 | Russ|werk | (inside the machine_is_shannon() clause of course) |
21:01:44 | GoRK | that's a udelay(100) ? |
21:03:07 | Russ|werk | think so |
21:03:33 | Russ|werk | #include <linux/delay.h> |
21:04:12 | Russ|werk | Using udelay() for intervals greater than a few milliseconds can |
21:04:12 | Russ|werk | * risk overflow for high loops_per_jiffy (high bogomips) machines. |
21:04:28 | Russ|werk | The |
21:04:29 | Russ|werk | * mdelay() provides a wrapper to prevent this. For delays greater |
21:04:29 | Russ|werk | * than MAX_UDELAY_MS milliseconds, the wrapper is used. |
21:04:42 | Russ|werk | oh, so you don't need to worry about taht |
21:05:29 | GoRK | ok im trying this now :) |
21:06:24 | GoRK | 2.4.13-ac8-rmk1 is out.. nothing in this driver though |
21:06:47 | Russ|werk | how big is the interdiff |
21:08:02 | SpaceCoaster | GoRK: hi |
21:08:55 | GoRK | um where is that located? the difference in size between this patch and the last one is fairly significant.. ~35k gzipped |
21:09:14 | GoRK | SpaceCoaster: greetings. i cant seem to find a suitable 'good' version of wheaties |
21:09:35 | GoRK | the ones old enough to work have no support for serial flash etc |
21:09:51 | SpaceCoaster | GoRK: I would suggest not releasing the serial_flash or 32M support, they aren't ready for prime time at all. |
21:10:41 | SpaceCoaster | GoRK: the original wheaties driver is the most robust. |
21:10:59 | GoRK | SpaceCoaster: heh :) you should talk to russ then... im trying to track down why the new ucb1x00 ts driver causes a hard lock |
21:11:29 | SpaceCoaster | Russ|werk: did you catch that? |
21:12:01 | GoRK | Russ: the delay didnt help anything. i cant use sysrq. |
21:12:24 | Russ|werk | ya |
21:12:39 | Russ|werk | double check the irg gpio's+ |
21:12:41 | Russ|werk | ? |
21:13:02 | GoRK | they are good it seems. i checked them the other day. audio uses a lot of them and it works fine ... |
21:15:11 | SpaceCoaster | Russ|werk: Can you back out the 32M and serial_flash changes from this release and get a wheaties.c and wheaties.h from GoRK that he is happiest with? |
21:15:35 | GoRK | wait it sets the backlight stuff on open. im gonna comment that all out and see |
21:16:15 | SpaceCoaster | GoRK: is your ucb1200 reset early enough? |
21:16:34 | GoRK | SpaceCoaster: donno. should be |
21:18:28 | GoRK | Russ: is there some reason you were setting initial brightness/contrast on ts device open? |
21:18:41 | GoRK | Russ: since you couldn't do it when the driver was init'd? |
21:21:23 | GoRK | hmm nope |
21:25:45 | Russ|werk | because the default brightness/contrast is bad |
21:26:58 | Russ|werk | I'll take whatever driver GoRK gives me |
21:27:29 | Russ|werk | the backlight stuff is being moved into a seperate driver that will register with sa1100fb.c |
21:27:35 | SpaceCoaster | Russ|werk: thanks |
21:27:42 | Russ|werk | so it can be compiled out, or as a module |
21:29:34 | Russ|werk | hey andersee, any bb mount -a insight/ |
21:29:34 | Russ|werk | ? |
21:29:45 | andersee | I didn |
21:29:48 | andersee | I didn' |
21:29:48 | GoRK | Russ: there is no problem. |
21:29:57 | andersee | argh. fingers are off. |
21:30:00 | GoRK | Russ: i hadnt taken out devfs |
21:30:09 | prpplague | anyway done any work with NET+ARM processor? |
21:30:29 | andersee | I didn't get to it last night, but as soon as I take a shower and get some breakfast (its only 3:30pm) I'll take a look |
21:30:40 | GoRK | Russ: err acutally that line was still commented out... hang on |
21:31:49 | GoRK | ok yeah mount -a is still foo |
21:34:12 | GoRK | but i think i know what the problem might be... my proc/mounts with devfs reads /dev/mtdblock/2 / jffs2 rw 0 0 |
21:34:16 | andersee | Russ|werk: just saw -- looks like you are feeding tuxscreen patches to rmk. cool! |
21:34:29 | GoRK | and without devfs it reads /dev/root / jffs2 rw 0 0 |
21:34:49 | GoRK | andersee could /dev/root be messing with the mind of busybox mount -a ? |
21:34:56 | andersee | /dev/root is ok. |
21:35:50 | andersee | I wrote a function to fixup the /dev/root name a long time ago, which is why 'df' and 'mount' behave |
21:36:09 | andersee | So thats not it. |
21:36:15 | GoRK | could it have broken in these recent kernels? *shrug8 |
21:36:24 | andersee | Could be. |
21:36:43 | GoRK | where does it get the real name of /dev/root from? |
21:37:05 | andersee | My next step (after eating) is to brew up a custom busybox with every possible failure mode spewing stuff to stderr... |
21:37:17 | BZFlag | andersee: xmission dsl is up. ;-) DNS is not settled down yet, so I'll wait a bit to start the move back. |
21:38:06 | andersee | GoRK: busybox/libbb/find_root_device.c |
21:38:54 | andersee | I basically scan /dev/ for matching devices, then use that name instead. |
21:39:04 | andersee | BZFlag: cool! |
21:39:31 | andersee | BZFlag: mananged any sneakiness, or is your cisco doing NAT? |
21:40:25 | BZFlag | nat at the moment. I'll play more this weekend. I will stop nat on one of the cisco or my server. ;-) |
21:40:38 | GoRK | andersee: hmm.. it looks like it just returns /dev/root if nothing matches. i will try a symlink |
21:41:56 | andersee | Right. |
21:42:19 | andersee | So its best have actually have a /dev node for each block device you will be mounting. |
21:42:34 | GoRK | nope.. |
21:42:42 | andersee | Do we not have a proper mtdblock dev? |
21:43:05 | GoRK | yes and it matches the root dev in the kernel commandline |
21:43:16 | GoRK | /dev/mtdblock2 |
21:43:37 | GoRK | i also made a /dev/root with the same major and minor and it didnt work |
21:43:54 | andersee | was just looking at the device_table file |
21:44:04 | andersee | doesn't work? |
21:44:16 | andersee | Still prints "/dev/root"? |
21:44:41 | GoRK | proc shows /dev/root ... i didnt make the function find_real_root print anything |
21:45:07 | andersee | Nothing will change the proc entry short of a kernel patch. |
21:45:19 | andersee | Its a chicken and the egg problem in the kernel. |
21:46:10 | andersee | The kernel doesn't have root mounted yet when it mounts the rootfs, so it doesn't know what /dev might want to call / after / is mounted |
21:46:23 | andersee | So it just opts out and uses /dev/root |
21:46:29 | prpplague | anyone used this before - http://www.netsilicon.com/EmbWeb/products/netuClinux.asp |
21:46:40 | andersee | busybox fixed that up for df and 'mount' |
21:48:17 | andersee | GoRK: do 'df' and 'mount' behave (i.e. not show /dev/root)? |
21:49:11 | GoRK | andersee: checking.. |
21:51:36 | GoRK | df behaves yes |
21:51:40 | GoRK | # df |
21:51:40 | GoRK | Filesystem 1k-blocks Used Available Use% Mounted on |
21:51:48 | GoRK | /dev/mtdblock2 3072 1664 1408 54% / |
21:51:53 | GoRK | ramfs 7060 16 7044 0% /var |
21:52:42 | andersee | k |
21:52:55 | andersee | Then thats not it either. |
21:53:46 | Russ|werk | I really don't think this has anything to do with devfs |
21:53:56 | andersee | My theory is that mtd is broken subtly in the kernel itself... |
21:54:09 | Russ|werk | some behavior in the kernel definately changed |
21:54:18 | andersee | Otherwise busybox mount would be failing on other systems. And it isn't |
21:54:27 | andersee | Russ|werk: right. |
21:55:08 | GoRK | it is so bizzare that mount -a fails and mount -o remount,rw /dev/mtdblock2 / works though |
21:55:24 | Russ|werk | what about mount -o remount,rw /dev/root / |
21:58:05 | GoRK | yes that works |
21:58:30 | andersee | Well, time to get breakfast, take a shower, then I'll spend up to an hour on this. |
22:03:31 | BZFlag | andersee: vim syntax for dns? |
22:03:46 | BZFlag | andersee: move SOA to first or second line to enable. |
22:03:53 | BZFlag | it's cool. |
22:04:34 | SpaceCoaster | BZFlag: hi, got a mo to talk memory? |
22:04:57 | BZFlag | sure. whatcha findin? |
22:06:15 | SpaceCoaster | Somethings wrong with the C0300000, I don't think the memory mapping is as general purpose as it tries to be. |
22:06:47 | SpaceCoaster | We are still hacking it so something will drop out soon. |
22:07:33 | SpaceCoaster | Eric Mouw (mr blob) has some ideas that might allow progress. |
22:08:20 | SpaceCoaster | Using the 2M segments means that we need a slightly smaller kernel. |
22:08:35 | SpaceCoaster | And we have problems loading the ramdisk. |
22:09:03 | Russ|werk | I've always had problems loading a ramdisk |
22:09:17 | Russ|werk | some work, some don't |
22:09:40 | SpaceCoaster | Eric proposes putting a 1M at the start, sortof to distract the kernel, and then loading things at a later 2M segment. |
22:10:07 | SpaceCoaster | BZFlag: anycase any ideas, your input appreciated too Russ|werk |
22:11:58 | Russ|werk | SpaceCoaster: maybe make the kernel decompressor understand the segementation |
22:12:15 | Russ|werk | and have the first thing the kernel does be setup the MMU, so hte sections are joined |
22:13:25 | SpaceCoaster | Russ|werk: the compressor part would be straighforward, those are get/put byte macro hacks. The kernel split in two and sewing itself back together, hmmm. |
22:13:59 | BZFlag | SpaceCoaster: hmmm. leaving the 1M at the start might not be bad. |
22:14:30 | BZFlag | that means 1M 2M 2M 2M 1M for the 8M case? |
22:14:32 | Russ|werk | yah, just make sure the code to turn on the MMU is within the first 1M |
22:14:56 | Russ|werk | just do 8 1M sections |
22:15:02 | BZFlag | the kernel sould still be in a later 2M segment, no? |
22:15:19 | BZFlag | Russ|werk: 1M is not enough, and too many segments. |
22:15:26 | Russ|werk | then have the decompressor take a quick look at the ATAG_MEM nodes |
22:15:40 | Russ|werk | BZFlag: does it need to be? |
22:15:49 | BZFlag | to be what? |
22:15:56 | Russ|werk | BZFlag: make the decompressor decompress across segments |
22:16:12 | SpaceCoaster | Russ|werk: that part is easy |
22:16:21 | Russ|werk | then, as soon as the kernel starts, it turns on the MMU, and the segments are tied together |
22:16:51 | SpaceCoaster | Russ|werk: do you have a pointer for some MMU code? |
22:17:03 | Russ|werk | the kernel |
22:17:11 | BZFlag | heh |
22:17:15 | Russ|werk | the kernel turns on the MMU as one of the first things it does |
22:17:50 | BZFlag | before or after it decompresses? |
22:17:59 | BZFlag | after, right? |
22:18:02 | SpaceCoaster | after |
22:18:33 | SpaceCoaster | I didn't see MMU stuff in arch/arm/boot/compressed |
22:18:33 | Russ|werk | so for maximum compatibility, the decompressor would need to parse ATAG_MEM |
22:18:39 | BZFlag | so can the compressed kernel go in the first meg with the decompressed in the next 2M (which alias) ? |
22:18:58 | BZFlag | er with alias... |
22:20:14 | BZFlag | Russ: hmmm. I don't see the point in looking at the ATAG. it won't be able to move anything as the addresses will be fixed. you mean just to determine where to put the plit tail portion that will be joined at mmu init? |
22:20:30 | BZFlag | s/plit/split/ |
22:20:48 | Russ|werk | the first thing it does is map 4M |
22:21:19 | Russ|werk | so if its assuming that there is 4M of contiguous memory at 0xc0000000, you'll have problems |
22:21:42 | BZFlag | does the 4M need to be contiguous? |
22:21:42 | Russ|werk | (looking at arch/arm/kernel/head-armv.S) |
22:22:35 | BZFlag | can we load the compressed image in the last 2M and decompress to c000? then the mmu could link all 8M back together? |
22:23:05 | Russ|werk | I would say try to do it with no "tricks" |
22:23:06 | BZFlag | 1,1,2,2,2 |
22:23:26 | BZFlag | how then? |
22:23:33 | Russ|werk | hove the decompressor look at the map passed by the bootloader, and decompresses arcoss the breaks |
22:23:52 | Russ|werk | then, have the MMU setup in head-armv.S be a bit smarter |
22:24:11 | BZFlag | bit smarter = relocatable? |
22:24:35 | GoRK | heh you can run this new ucb1x00 audio driver at about 62kHz sampling rate :) |
22:25:08 | Russ|werk | /* |
22:25:10 | Russ|werk | * Create identity mapping for first MB of kernel to |
22:25:10 | Russ|werk | * cater for the MMU enable. This identity mapping |
22:25:10 | Russ|werk | * will be removed by paging_init() |
22:25:10 | Russ|werk | */ |
22:25:16 | Russ|werk | krnladr r2, r4, r5 @ start of kernel |
22:25:19 | Russ|werk | add r3, r8, r2 @ flags + kernel base |
22:25:19 | Russ|werk | str r3, [r4, r2, lsr #18] @ identity mapping |
22:25:26 | Russ|werk | /* |
22:25:27 | Russ|werk | * Now setup the pagetables for our kernel direct |
22:25:27 | Russ|werk | * mapped region. We round TEXTADDR down to the |
22:25:27 | Russ|werk | * nearest megabyte boundary. |
22:25:27 | Russ|werk | */ |
22:25:36 | Russ|werk | add r0, r4, #(TEXTADDR & 0xff000000) >> 18 @ start of kernel |
22:25:37 | Russ|werk | bic r2, r3, #0x00f00000 |
22:25:37 | Russ|werk | str r2, [r0] @ PAGE_OFFSET + 0MB |
22:25:37 | Russ|werk | add r0, r0, #(TEXTADDR & 0x00f00000) >> 18 |
22:25:37 | Russ|werk | str r3, [r0], #4 @ KERNEL + 0MB |
22:25:37 | Russ|werk | add r3, r3, #1 << 20 |
22:25:39 | Russ|werk | str r3, [r0], #4 @ KERNEL + 1MB |
22:25:39 | BZFlag | Russ... just past the source name and line number... |
22:25:41 | Russ|werk | add r3, r3, #1 << 20 |
22:25:43 | Russ|werk | str r3, [r0], #4 @ KERNEL + 2MB |
22:25:45 | Russ|werk | add r3, r3, #1 << 20 |
22:25:47 | Russ|werk | str r3, [r0], #4 @ KERNEL + 3MB |
22:26:00 | Russ|werk | I'd need to look stuff up, but it looks like its just blindly assuming 4M contiguous memory |
22:26:24 | Russ|werk | and I think that is the part that would need to jump across the 1M boundries |
22:26:34 | Russ|werk | as well as later in the kernel when it sets up the full page tables |
22:26:36 | BZFlag | I believe that's what it's doing. |
22:26:52 | Russ|werk | but, I think it does that with the ATAG_MEM nodes, so that is set up properly |
22:27:03 | Russ|werk | is just in the head file, its doing it blindly |
22:27:42 | Russ|werk | so the decompressor needs to know about the first 4M, and head-armv.S needs to know about the first 4M |
22:28:06 | BZFlag | then they need to handle 10 memory regions too. |
22:28:19 | Russ|werk | no, the later setup does |
22:28:20 | BZFlag | er 12 actually for the 16M sodimm. |
22:28:27 | Russ|werk | which I don't think is a problem |
22:28:33 | BZFlag | oh, worse! 16 for the 8M |
22:28:44 | Russ|werk | in the initial setup, we only need to worry about the first 4M |
22:28:56 | BZFlag | true. |
22:29:11 | Russ|werk | which is probably where we are running into problems |
22:29:50 | SpaceCoaster | I think it is the 128M alignment issue which is tripping us up. |
22:30:04 | Russ|werk | so maybe if make the bootloader put something somewhere that tells the offsets of the first 4 1M blocks |
22:30:25 | Russ|werk | so the decompressor would know where to decompress the kernel |
22:30:40 | Russ|werk | and head-armv.S knows where to do the initial MMU mapping |
22:31:21 | Russ|werk | and we get to keep the ZTEXTADDR etc, stuff all the same, and don't need to move up to 0xd0000000 |
22:31:23 | SpaceCoaster | Sounds good, do you speak MMU? We could hardwire the mappings to test. |
22:32:03 | SpaceCoaster | Russ|werk: the latests tests were at 0xC0300000 but this has alignment problems. |
22:32:09 | Russ|werk | I don't, but it looks like the above section of code is what is doing it |
22:32:29 | Russ|werk | this way would let us do it at 0xc0000000 |
22:32:52 | SpaceCoaster | I keep on thinking about ordering that ARM architecture manual but don't actually do it :-( |
22:33:29 | Russ|werk | and would assume that there is 1M memory at 0xc0000000, and then three more 1M chunks before 0xd0000000 |
22:33:37 | Russ|werk | (could all be together) |
22:33:53 | Russ|werk | I think the sa1100 manual has enough info |
22:34:15 | Russ|werk | I have it printed out, but not here |
22:35:28 | Russ|werk | I'm fairly certain you can turn up the number of allowed memory regions to 16, but it defaults at 8 irrc |
22:35:31 | Russ|werk | er, iirc |
22:35:47 | SpaceCoaster | ok, thanks that is a good avenue to try. |
22:35:58 | Russ|werk | you could start out with #ifdef's, and then figure out how to pass actual values later |
22:36:14 | Russ|werk | then, one kernel, one blob could boot any tuxscreen |
22:36:31 | BZFlag | it would be nice to be able to avoid the tricks and still have a method that works on other hardware. |
22:36:38 | SpaceCoaster | sounds good |
22:37:15 | Russ|werk | this would help out other arches too |
22:37:24 | SpaceCoaster | BZFlag: why, their method doesn't work on ours ;-) |
22:37:32 | BZFlag | how do you propose fixing the decompressor? setup the mmu, decompress, unset the mmu and continue? |
22:37:57 | SpaceCoaster | BZFlag: the decompressor is really simple, it can step over the spaces. |
22:37:58 | Russ|werk | no, change the mmu to jump across boundries |
22:38:02 | Russ|werk | er |
22:38:10 | Russ|werk | change the decompressor to jump across the boundries |
22:38:18 | BZFlag | ok. easier. |
22:39:14 | SpaceCoaster | BZFlag: the decompressor uses the technique that I hadn't seen before #define macros and include the source to override internal functionality, a real hack. |
22:39:48 | SpaceCoaster | jffs2 and ppp both include zlib.c which is bad news. |
22:40:11 | SpaceCoaster | the boot code includes "inflate.c" |
22:40:39 | Russ|werk | inflate.c is from zlib |
22:41:48 | SpaceCoaster | hmm, the same code gets included 3 times with different macros, well it works. |
22:42:17 | Russ|werk | blob will probably end up using mini_inflate for jffs2 |
22:43:48 | SpaceCoaster | Any ideas about blob and the ramdisk for our 16x1M config? |
22:45:20 | Russ|werk | load it at 0xd0000000 |
22:45:39 | Russ|werk | maybe blob should look for a place for the ramdisk on its own istead of it being hardcoded |
22:46:54 | Russ|werk | I think the MMU mapping is complete by the time the kernel looks for the ramdisk, its just a question of how the address translates |
22:48:04 | SpaceCoaster | Hadn't thought about how the kernel would see it. Just blob at the moment. |
22:53:09 | Russ|werk | I don't know exactly how the kernel locates the ramdisk |
22:55:25 | Russ|werk | unfortunately, I'll be out of town this weekend |
23:02:00 | Russ|werk | away |
23:04:34 | SpaceCoaster | Russ|werk, BZFlag: thanks for all that good stuff |
23:06:08 | SpaceCoaster | GPSFan: look at this RAS/CAS calculator http://appzone.intel.com/hcd/sa1110/memory/ |
23:06:27 | SpaceCoaster | It's for the SA-1110 but it might be useful. |
23:13:37 | GPSFan | SpaceCoaster: that's cool, I'll check it out. ;>) |
23:29:14 | Tangent | waits for something to happen |
23:29:41 | Tangent | discovers the perils of the scrollback buffer... again |
23:29:43 | pattieja | is away: home |
23:34:04 | GoRK | GPSFan: hello |
23:34:44 | GoRK | GPSFan: did you see my latest pcb? with stereo jacks and the audio transformer? |
23:34:46 | GPSFan | GoRK: hi. |
23:35:15 | GPSFan | GoRK: I looked, but the url didn't work. |
23:35:32 | GoRK | GPSFan: there is some ground loop problem plugging stuff into line out that i fixed with a transformer, but i didnt know if there was an easier way to do it |
23:35:36 | GoRK | http://www.blurbco.com/~gork/tuxscreen/shanip2-gork7.gif |
23:35:58 | GoRK | GPSFan: i know if someone plans on using this in their car, they'll probably need a transformer to isolate it anyway ... :-/ |
23:36:46 | GPSFan | GoRK: transformer isolation is always a good thing to do when you have too many ground interactions betweek digital & analog ckts. |
23:39:05 | GPSFan | GoRK: betweek (?) yeah, that's what we have been doing to this stuff, we be-tweekin' it. |
23:43:11 | GoRK | GPSFan: ok well what transformer should i be using for the best results and highest level? i am looking at parts on this page: www.mouser.com/catalog/cat_607/331.pdf |
23:44:06 | GoRK | GPSFan: between the TY series plug in pc audio transformers, which one would owrk best |
23:47:55 | GPSFan | GoRK: well, the impedance of the speaker out is rather low, and the Line out port should be 600 ohms, I don't see anything that would fit exactly, but the TY145P should work. |
23:47:55 | GoRK | ok i didnt know what the impedence was supposed to be at all |
23:48:02 | GPSFan | GoRK: a properly matched one would also tend to step up the voltage which may be undesireable. |
23:48:46 | GoRK | so the ty145p may not work out? |
23:49:22 | GoRK | i can't find a stupid ps/2 female jack from these fools :( |
23:52:54 | GPSFan | GoRK: I'm sure it will be fine, it will repeat the input voltage at the output, and it will isolate the ground systems of the external audio amp, but this is computer generated audio, and unless you spend more for your audio system then you make in a year, it won't matter. |
23:53:18 | GoRK | GPSFan: ok! :) |
23:53:30 | GoRK | GPSFan: if we wanted really nice audio from this we wouldnt be using ucb1200 anyway |
23:53:37 | GPSFan | GoRK: what do you need a ps/2 female for? |
23:53:48 | GPSFan | right ;>) |
23:53:56 | GoRK | GPSFan: i need to put a ps2 port on this motherboard that doesnt have one |
23:54:21 | GoRK | it only has little .050 headers and a pinout. they are selling the stupid thing for like $35 |
23:55:25 | GPSFan | ah, you need to go to somewhere like HSC or alltronics. url's comming... |
23:56:09 | GPSFan | GoRK; http://www.allelectronics.com/ http://www.halted.com/ |
23:56:22 | GoRK | the LART schematics show a dac on the ssp port of sa1100... there are drivers in the new kernel for good stereo audio from it. do you know if the tuxscreen uses the ssp port for anything? im wondering if it'd be possible to use the same circut for the tux.... |
23:57:22 | BZFlag | SpaceCoaster: perhaps blob should look at the size of the ramdisk and locate it at the end of memory wherever that is? |
23:57:34 | SpaceCoaster | GoRK: the ssp port is used for the DSP and the serial flash |
23:58:33 | GoRK | SpaceCoaster: ahh. is it a bus that can be shared or a serial port or something? |
23:58:51 | SpaceCoaster | BZFlag: it was the fragmentation for loading it into the flash that I was considering first. I can always put it at 4M. |
23:59:36 | BZFlag | it should still work on an 8M tux, so that means fragmenting in the the later blocks. |
23:59:43 | GoRK | btw tangent did get his extra serial port working, didnt he? |