| 00:41:07 | file | hery prpplague |
| 00:41:09 | file | er hey |
| 01:00:16 | prpplague | hey file |
| 01:00:22 | file | wazzup? |
| 01:00:42 | file | bah hum bug |
| 01:00:46 | file | the last episode of Roswell is playing |
| 01:00:50 | prpplague | ssdd |
| 01:01:02 | file | is unhappy |
| 01:01:05 | file | all my favorite shows are ending |
| 01:01:31 | prpplague | which ones are those? |
| 01:01:43 | file | ST: Voyager... Roswell... |
| 01:01:46 | file | I forget the rest |
| 01:01:53 | file | is very forgetful |
| 01:02:00 | file | so, any news? |
| 01:02:22 | prpplague | on? |
| 01:02:23 | file | hrm? |
| 01:02:31 | file | any news on anything |
| 01:03:02 | file | few more months... |
| 01:03:10 | prpplague | just getting ready for a trip to barbados in august |
| 01:03:14 | file | ahhhh |
| 01:03:16 | prpplague | and the OLS in june |
| 01:03:22 | file | I'm leaving for Quebec on Wednesday |
| 01:03:29 | file | at 6am... |
| 01:03:41 | prpplague | and working on a CPU board |
| 01:04:01 | file | is watching the last episode of Roswell, drinking pepsi, and eating cold pizza |
| 01:04:44 | prpplague | is drinking a frosty glass of "goose island honkers ale" and testing rhat 7.3 |
| 01:05:00 | file | Redhat? yuck |
| 01:05:25 | prpplague | file: what do you use? |
| 01:06:02 | file | I use a Slackware variant |
| 01:06:54 | prpplague | ahh, i've got a slack box, but i have to work with rhat at work, so it pays to know to use rhat |
| 01:07:01 | file | nods |
| 01:09:16 | prpplague | rhat makes it easy to switch ppl from winblows and proprietary unix |
| 01:09:51 | file | uh huh |
| 01:10:09 | prpplague | i know alot of ppl that don't like rpm's but it sure makes it easy to manager large groups of servers |
| 01:10:18 | file | nods |
| 01:12:09 | file | the look of Linux... is in... your distro |
| 01:12:16 | file | the look... your compile... can't disguise |
| 03:29:06 | pmisfit | is back (gone 73:12:20) |
| 03:29:06 | pmisfit | ' |
| 03:29:16 | pmisfit | file: are you quoting a burt bacharach tune??? |
| 03:29:46 | pmisfit | that's twisted. |
| 03:31:18 | pmisfit | shrugs |
| 03:31:41 | pmisfit | is away: slinking away into the shadows, to emerge again at some other random interval |
| 03:49:12 | prpplague | andersee: howdy howdy |
| 03:49:22 | andersee | hey |
| 03:50:09 | prpplague | andersee: any fun stuff on the horizon? |
| 03:51:01 | andersee | just watching my new little girl |
| 03:51:22 | bmidgley | how new?! (congrats) |
| 03:51:40 | andersee | born monday. thanks. :) |
| 04:28:51 | prpplague | andersee: cool, congrads |
| 05:48:43 | BZFlag | andersee: contrats. pics on the web yet? ;-) |
| 05:49:06 | BZFlag | er congrats. give me a break, one handed typing is hard. |
| 05:49:49 | BZFlag | finds the pics... ;-) |
| 05:50:28 | BZFlag | er.. wrong kid. |
| 05:54:16 | andersee | BZFlag: http://www.codepoet.org/gallery/albums.php |
| 05:54:57 | andersee | BZFlag: Had to let the family see of course. ;) |
| 08:35:17 | Russ | andersee: btw, busybox ping should really return false (non-zero) if the ping fails |
| 09:12:33 | BZFlag | Russ: fixed. |
| 11:51:17 | TomW | yawns |
| 14:27:39 | TomW | GPSFan: morning |
| 14:29:26 | GPSFan | TomW: good morning, what's new & exciting? |
| 14:30:43 | TomW | ah, nothin' |
| 14:30:54 | TomW | rain this morning, cold. |
| 14:31:23 | GPSFan | wish we had some rain here, very dry... |
| 14:31:56 | TomW | got a Creative Labs Banshee video card working for a friend's linux box, actually it is his son's computer. |
| 14:32:13 | TomW | what part of the country are you in? |
| 14:32:19 | file | finishes his work |
| 14:32:22 | GPSFan | Western Colorado.. |
| 14:32:36 | TomW | Yeah, it gets dry there. |
| 14:32:39 | file | I wrote a mIRC Script to randomly play my midi files and handle artists/names etc |
| 14:33:24 | GPSFan | this year we are at about 10% normal. it's going to be a tough summer. |
| 14:33:26 | TomW | file: so, we can play selections from your music collection for you? |
| 14:33:36 | file | ... no |
| 14:33:41 | file | I don't use mIRC on here |
| 14:33:48 | file | and I only have about 43 midi's |
| 14:34:01 | file | from "I Shot the Sherif" to "The Locomotion" |
| 14:34:15 | TomW | GPSFan: that was last year! The walnut tree here in the yard is about had it, too little rain & too many caterpillers. |
| 14:35:30 | file | is in Canada |
| 14:35:40 | file | where it's... not too bad out |
| 14:36:20 | file | Oooo, I love the way you... |
| 14:36:20 | file | Love the way you love me |
| 14:36:20 | file | There's nowhere else I'd rather be |
| 14:36:20 | file | Oooo, to feel the way I... |
| 14:36:20 | file | Feel with your arms around me |
| 14:36:21 | file | I only wish that you could see... |
| 14:36:32 | GPSFan | ranchers around here are starting to sell off their cows, there may be only one cutting of hay, and no water to irrigate the pastures. |
| 14:36:32 | TomW | ?? maybe I didn't uderstand, play the midi stuff on the Tux? |
| 14:36:38 | file | TomW: no no |
| 14:36:42 | file | on my desktop system |
| 14:37:24 | TomW | file: ah, ok, I don't get into the music thing. Whatever is on the FM radio station is fine with me. |
| 14:37:41 | file | I'm not into MP3's, midi's I like though |
| 14:37:46 | file | you can sing with them with your own words |
| 14:38:18 | TomW | GPSFan: what do you do for a living? |
| 14:39:27 | file | now to make my things move around... |
| 14:40:46 | GPSFan | sort of retired (early), was EE, designed computer systems for various folks. now a consultant, raise sheep & horses. |
| 14:44:43 | TomW | GPSFan: ah, yeah, I sorta retired at 42. I had a job over in NJ for about 8 years doing embedded design. Then they wanted me to work 8-5 'cause they screwed up and let some knowledgable people go. They wanted me to pick up the slack, I eventually resigned. I didn't want to go back to work for someone else ... So I now consult. |
| 14:50:10 | GPSFan | TomW: yep, that's sort of what happened here too, my wife & I rebuilt her mom's house in Colorado, & moved there in late 2000, after working in Calif. for 20 years. Nice to be on my own, no hi-tech jobs anywhere around here. but I still get work from clients on the coast. |
| 14:50:41 | GPSFan | TomW: gotta run & feed the animals, bbl |
| 14:50:59 | TomW | GPSFan: later |
| 14:52:43 | file | there |
| 14:52:51 | file | I added a neat effect to my midi playing script |
| 14:52:58 | file | the title/artist moves around on the screen |
| 17:37:57 | LoneTech | "fake.hmtl"? |
| 19:35:14 | lyakh | TomW - you're still there?:-) |
| 19:35:24 | TomW | yup |
| 19:35:38 | lyakh | hi, did you solve the problem with compilation? |
| 19:36:48 | TomW | I am compiling it natively on the Tux. Takes a while to compile glibc, but.. I will have to setup a cross-dev environ for this, I can see that now. |
| 19:37:54 | lyakh | :-)))) gosh... compiling libc natively on strongarm... you want an entry in the book of records?:-) |
| 19:38:22 | lyakh | I just recalled - don't know if it's relevant to the ARM-debian port: |
| 19:38:33 | TomW | heh, it certainly is a heavy test of the IDE drive, eh? |
| 19:39:08 | TomW | I am recompiling it against the 2.4 headers. |
| 19:39:50 | lyakh | to run debian with 2.4 (potato that is) you have to upgrade to some "unstable" packages by adding an entry to some apt-get configuration file and running apt-get upgrade or something like that... |
| 19:41:17 | TomW | not going to replace glibc, just getting a version that knows how to talk to the 2.4 kernel, then I'll statically link the reiserfs tools against that glibc. |
| 19:41:41 | lyakh | You know, linux/Documentation/Changes lists (among other things) minimum requirements for running this version of the kernel - which version of modutils, utils-linux, etc. |
| 19:41:43 | TomW | already thought about that: incompatibilites introduced into existing progs. |
| 19:44:02 | lyakh | So, if you have a look at Debian-ARM - maybe they'll have somethng similar. Alternatively, when on Tuesday I get to work, I can check what versions of libc, etc. we are running (with 2.4.2, 2.4.13, 2.4.18, 2.5.6) |
| 19:44:38 | lyakh | But I did have to recompile a couple of these packages - cross-recompile, of course:-) |
| 19:44:59 | TomW | lyakh: ok. There doesn't seem to be a lot of stuff available for 2.4, must be in cvs or something. |
| 19:45:18 | TomW | lyakh: I am 433K into the swapfile on the Tux. |
| 19:45:40 | TomW | this Tux has been upgraded to 64M of DRAM. |
| 19:46:47 | TomW | If anyone asks "have you loaded anything big into that modified DRAM", I can tell them that I compiled glibc with it. heh |
| 19:47:20 | lyakh | :-)) David Schleef told me 192MB is what he uses on his laptop to compile the kernel "completely in RAM":-) |
| 19:48:21 | TomW | <sigh> I wish I could put more RAM in this thing, but it would means totally revamping the DRAM electronics. Not worth the trouble. |
| 19:50:55 | TomW | this little machine is loaded! idle: 0.9% and 572K into swap.. |
| 19:51:37 | TomW | lyakh: you know, the easy way to do this don't you? |
| 19:51:56 | lyakh | To do what? load a machine?:-) |
| 19:52:24 | TomW | yup, stick the drive in my laptop and create the reiser filesytem in there. |
| 19:52:40 | TomW | no, cannot do that, too easy. |
| 19:52:58 | lyakh | Why not? |
| 19:53:19 | TomW | I don't know, the challenge maybe? |
| 19:53:23 | lyakh | You don't like easy ways?:-) |
| 19:53:44 | TomW | don't learn anything by doing certain things the easy way. |
| 19:53:55 | lyakh | ok, do you want another challenge? |
| 19:54:17 | TomW | what I would like to do is get paid to do linux development. |
| 19:54:52 | lyakh | hm, depends how much you want:-) |
| 19:55:31 | TomW | yeah, I know, you would like to get paid for doing it too. Right? |
| 19:55:32 | lyakh | One could say that I am payed for linux development, in a way... |
| 19:55:35 | TomW | ;-) |
| 19:57:00 | lyakh | writing linux (real-time) drivers, porting the kernel, fixing bugs in others' drivers - that's my job:-) |
| 19:57:26 | TomW | still cannot see how I can speed up the interface, if only the ARM released the CS3 during memory fetches. I could tell it that the bus was 16 bits and to fetch a long from the interface. |
| 20:00:16 | TomW | lyakh: maybe linux will payoff somehow, but, in the meantime, we are having fun with it. |
| 20:00:35 | lyakh | I tried to have a quick look yesterday on all those CSx, and some other registers in the doc, but, obviously, it was not enough to understand them...:-) |
| 20:01:22 | TomW | lyakh: that is what I do for a living, bang bits into registers. The MECR stuff is setup correctly, I verified it on a scope. |
| 20:01:23 | lyakh | oh, yeah, lots of fun, I don't want a job with 10 times my present salary doing... you know what:-) |
| 20:01:43 | TomW | shudders |
| 20:02:25 | TomW | must be 1AM on your side. |
| 20:02:37 | ctr | what do i need to do to fix and issue where gcc isn't picking up the def of a function, yet the include file is included |
| 20:02:43 | lyakh | wrong, only 22 |
| 20:02:49 | ctr | eg. strdup with #include <string.h> |
| 20:03:52 | TomW | ctr, what I usually do is verify that it is reading the #define by mis-spelling it as #dfine and see if gcc issues an error: never assume |
| 20:04:53 | TomW | lyakh: the Tux still hasn't crashed yet, still compiling. |
| 20:05:04 | TomW | heh, this is cool |
| 20:05:58 | lyakh | yep, cool, actually, you might have an idea for my problem... |
| 20:06:36 | TomW | what's that. |
| 20:07:16 | lyakh | latencies. seem to be due to chache misses. |
| 20:08:01 | lyakh | When we ported RTAI to ARM, and ran a latency-measurement program, we were shocked. |
| 20:08:20 | TomW | sorry, as far as I have gone into the kernel "wild" stuff was to trace the scheduler through it's startup. |
| 20:08:55 | lyakh | running an ordinary user-space load increased latency A LOT. |
| 20:09:08 | lyakh | wait, remember, this is not linux - this is RTAI |
| 20:09:20 | TomW | I have been thinking of realtime, but you people talk of realtime in the 50..100ms range with jitter of 10..30ms. That isn't good enough for me. |
| 20:10:13 | lyakh | latency was measured in a real-time task, so, it shouldn't depend on user-space load. |
| 20:10:22 | lyakh | what values do you need? |
| 20:10:23 | TomW | true. |
| 20:10:59 | TomW | I need periodic interrupt every 10ms with jitter limited to < 2ms. |
| 20:11:48 | lyakh | Measuring latency on StrongARM is very simple - I tried it without real-time - just set up a match register, and check how much later you get this interrupt - same dependancy on user-space load... |
| 20:12:02 | TomW | processing of the data would be done in the lower level and the result would be examined via the kernel, rather than being passed up into it. |
| 20:12:30 | lyakh | 10ms is CERTAINLY possible, jitter < 2ms, should be possible too... |
| 20:13:16 | TomW | yes, as long as the underlying process (o/s) has absolute priority over the interrupt and the interrupt is of the highest priority. |
| 20:13:37 | lyakh | how much processing and how fast a CPU? It all depends, you know. |
| 20:13:50 | lyakh | of course - how else?... |
| 20:14:06 | TomW | lyakh: not really. I am not doing an FFT with the thing. :-) |
| 20:14:59 | TomW | All I have to do is debounce some inputs and drive waveforms onto outputs. |
| 20:16:04 | lyakh | what do you mean "debounce"? |
| 20:17:40 | TomW | What I was thinking of doing was this, have the jiffies interrupt be trapped by my underlying routines. they do thier job in < a few milliseconds, then return to the jiffies function. I think that with a minimum of invasion into the kernel, I can induce some realtime for what my application needs to do. I don't think that I need to go into full-blown queues & fifos between the layers. |
| 20:18:24 | TomW | lyakh: when you close a mechanical contact on an electrical switch, it tends to bounce - literally. |
| 20:18:38 | lyakh | aha, so, you want real-time in non-realtime linux?:-) what about FIQs? |
| 20:18:44 | TomW | the result is that you get: on-off-on-on-off-on |
| 20:19:06 | TomW | FIQs? |
| 20:19:28 | lyakh | is it on ARM?... |
| 20:19:44 | TomW | yes, I am looking to do this on ARM. |
| 20:20:30 | lyakh | ARm (at least strongarm) has IRQs and FIQs (fast interrupt requests) - the latter ones have absolute priority over irqs |
| 20:20:38 | TomW | ok |
| 20:21:32 | lyakh | But there's not much support for FIQs in the kernel (because they have to be fast:-)), but some people were successful in using them with linux. |
| 20:22:15 | TomW | I am not putting together a product, yet. This is the reason I have 5 TuxScreens, I took the CPU board out of one of them so that I could use it to research approaches. I have a customer who is interested, I have not yet decided which processor to go with. |
| 20:22:56 | lyakh | some other CPUs have native interrupt priorities, so, you may try that. |
| 20:23:45 | TomW | I have been looking at Cirrus ARM chips, they look pretty interesting. |
| 20:25:17 | TomW | I really like ARM, lots of power and uses very little current. |
| 20:27:40 | lyakh | yep, this is why we use it |
| 20:28:23 | lyakh | Can anybody give me reasonable switches to wget to download docs from tuxscreen?... |
| 20:29:40 | TomW | It has been such a bitch to get started with the things. I have CLPS-7500FE development board that I purchased two years ago, but when I asked about it on linux-ARM list, all they would say was "gee, that is really old, the ARM kernel used to run on that..." |
| 20:30:22 | TomW | What are you trying to get? |
| 20:30:53 | lyakh | docs - installation, descriptions, etc... |
| 20:31:16 | lyakh | trying -r -A+tuxscreen.net at the moment... will see how far it gets me |
| 20:31:39 | TomW | That site uses a wiki, that is a strange beast. |
| 20:31:54 | lyakh | s/+/=/ |
| 20:32:14 | TomW | I have some stuff on installation over on: http://www.openhardware.net/TuxScreen/index.html |
| 20:32:52 | TomW | I waded through all the information & dis-information of the Wiki and digested it on my own pages. |
| 20:33:31 | lyakh | yep, thanks, will check too |
| 20:34:06 | TomW | that site is apache and plain old html. |
| 20:34:40 | TomW | AFAIK, the pages you see on the wiki are generated on the fly. |
| 20:35:09 | TomW | not sure though... |
| 20:36:01 | TomW | hmph, I have too many other things calling for my attention to get too deeply into how the wiki operates. |
| 20:37:07 | lyakh | ok:-) it did get a funny set - almost nothing:-) |
| 20:38:07 | lyakh | ok, getting your site - only had to remove "www":-) |
| 20:40:20 | lyakh | hm, -A works strangely... need something different... |
| 20:42:24 | TomW | wget -r should work ok. |
| 20:43:20 | lyakh | yep, but a bit too much for my connection:-) |
| 20:43:25 | lyakh | ok, got it... |
| 20:45:54 | lyakh | wget -Ahtml,htm,txt -r http://... |
| 20:46:00 | TomW | I never documented how to install inferno. Inferno on a linux 2.4 machine is not documented (well) on the wiki. You have to replace the emu with another one in inferno. |
| 20:46:26 | lyakh | do I need it (inferno)? it is pre-installed, isn't it? |
| 20:47:24 | TomW | you also need to get the jpg - j-pegs. Needed info is in those. |
| 20:48:28 | lyakh | power - is it just 12v? |
| 20:48:40 | TomW | You need to install inferno on the workstation so you can interact with the Tux inferno to erase the flash and install blob for the first time. |
| 20:49:13 | TomW | 3.3v & 5v, comes from the wall supply. 115v @ 60Hz. |
| 20:49:34 | lyakh | aha, ok, hope, it can be done on linux?:-) I think there was something about this on the site... |
| 20:50:22 | TomW | yes, you can do it on linux. |
| 20:50:27 | lyakh | 220/110 adapter, or an alternative power-supply |
| 20:51:28 | TomW | right, adapter or build your own supply: 3.3v @ 1.25Amp & 5v @ 600ma |
| 20:53:25 | lyakh | hm, got all files from wiki, but cross-references are broken... anyway, my internet bill this month will be wild...:-) falling off, cheers |
| 21:06:18 | as_ | Evnin all. |
| 21:06:55 | dc_ | evening |
| 21:46:19 | Russ | prpplague: mame is on the camion with 3 games |