| 00:16:49 | BZFlag | boodle: I've got the lcd version. no expansion slots, just soldered in pieces |
| 00:29:52 | boodle | BZFlag: ? sorry, not following.. this has only 1 cf slot.. didn't know there's other version (well ia-2 but that's crt..) |
| 00:33:05 | Keiser_Soze | anybody home? |
| 00:35:39 | BZFlag | right. I opened the ia-1 no sockets that I recall. |
| 00:37:38 | Keiser_Soze | BZFlag, your offering the tuxscreen's on http://tuxscreen.net/wiki/view/Users/InstallingLinux? |
| 00:39:02 | Keiser_Soze | I saw this line on the page: "As this is a wiki page, you may want to confirm that address on my web site before sending money: http://Rikers.org/resume.htm", but I cannot access the page. |
| 00:40:56 | dirkx | Hmm - what does the 'relay' do - in the LED list. |
| 00:41:17 | scanline | dirkx: It connects the phone line |
| 00:41:53 | scanline | dirkx: Just curious, what are you working on? |
| 00:42:12 | dirkx | scanline: txs - trying to work out why pressing the speakerphone button does not give me an dialtone in tuxphone |
| 00:42:44 | scanline | dirkx: Really... I haven't heard of anyone having problems with the speakerphone button |
| 00:42:53 | dirkx | scanline: just trying to get to a stage where I can a) make it wireless (done) and b) connect it to Vocal - a VoIP stack. I expect that I need some wiring for that. |
| 00:43:11 | scanline | dirkx: Next time you see GoRK, you should ask him about it. He wrote most of tuxphone |
| 00:43:23 | scanline | About VoIP, have you heard of our shanip2 project? |
| 00:43:27 | dirkx | scanline: right now when I press the speakerphone button I get a feedback feep tone. But the phone line is not switch to it. |
| 00:43:35 | dirkx | scanline; no - not heard of shanip2 |
| 00:43:51 | scanline | It's a hardware modification to connect the ucb1200 audio chip to the DSP |
| 00:44:01 | scanline | lets you do VoIP, play MP3s, etc. |
| 00:44:07 | dirkx | As per the data on the lucent site ? |
| 00:44:24 | scanline | Well, we started with the lucent designs, made it work right, and ended up with shanip2 |
| 00:44:30 | scanline | (the lucent design was shanip) |
| 00:44:34 | dirkx | any links ? |
| 00:44:49 | scanline | tuxscreen.net has the schematics |
| 00:45:12 | boodle | BZFlag: so even main 32MB not in a socket? :(( |
| 00:45:39 | scanline | dirkx: wander through http://tuxscreen.net/download/, there's a lot of good stuff in there |
| 00:45:46 | dirkx | gotsha. |
| 00:45:58 | scanline | dirkx: The 'schematics' directory has all of GPSFan's goodies, including a PDF of the shanip2 schematic |
| 00:46:21 | scanline | dirks: GoRK is working on a PCB with an improved shanip2, but i don't have a URL handy |
| 00:46:48 | scanline | dirkx: The 'shanip2-pictures' directory is just my crazy little method of building shanip2 without a PCB :) |
| 00:46:54 | dirkx | Looking at the schematics - you want to just attach an audio output plug ? |
| 00:47:09 | scanline | The audio output on the schematic isn't right |
| 00:47:21 | scanline | GoRK's board has line in/out, microphone in, and headphone out |
| 00:47:47 | scanline | The shanip2 sends audio to/from the DSP, so it can be routed to the handset or speaker |
| 00:48:01 | Keiser_Soze | how long does it take to boot a tuxscreen running nano-X? |
| 00:48:26 | dirkx | Aye - all I am looking for is intercepting/generating what goes to/fro the handset - so right now the A/D-D/A of the DSP goes where ? Just to the speaker ? |
| 00:49:18 | scanline | No... the DSP can route the audio to the handset or the speaker |
| 00:49:36 | scanline | I think that all the stuff you're trying to do has already been done, except for the software :) |
| 00:50:04 | scanline | But anyway... GPSFan is the one to talk to about schematics, and GoRK is the DSP Guru |
| 00:50:11 | dirkx | So the DSP can right now only -route- and not generate/listen-to ? |
| 00:50:57 | scanline | The DSP has no A/D or D/A that's accessable to the CPU, but the ucb1200 audio chip has D/A and A/D. The shanip2 connects the ucb1200 and the DSP |
| 00:51:20 | scanline | ucb1200 handles audio and touchscreen |
| 00:52:06 | scanline | The shanip2 is all the hardware you'll need for VoIP |
| 00:52:26 | boodle | BZFlag: from http://www.linux-hacker.net/ipaq/ looks like cpu could be replaced |
| 00:52:28 | dirkx | ok - so the impression i got from the lucent page was that i needed to conenct the ucb1200 to routing control of the DSP. |
| 00:52:46 | scanline | Right. But the schematics on the lucent page are wrong |
| 00:53:00 | dirkx | So the shanip2 does -two- things - it routes audio to an extra plug -and- it gets you a/d/a to the audio ? |
| 00:54:15 | scanline | Well, the shanip2 as-is, it just routes between ucb1200 and DSP. But, it's easy to connect additional jacks so you can plug in headphones, a stereo, an external microphone... |
| 00:54:56 | dirkx | Ah - so that one does not get's its own data channel ? from the spec's at agere.com it would seem likely that there is a third free audio path available ? |
| 00:55:58 | scanline | The DSP has only one input and output. The ucb1200 has two sets of input and output: |
| 00:56:16 | scanline | the ucb1200 audio is not used, and the shanip2 connects it to the DSP |
| 00:56:31 | boodle | BZFlag: interesting general thread on linuxhacker to use k6-2+ mobile chips (up to 500Mhz) |
| 00:56:33 | scanline | the usb1200 telecom channel is already connected to the phone line, but we don't have drivers for it yet |
| 00:57:03 | scanline | So with the shanip2, we have one channel going to the handset/speaker/line jacks, plus one channel going to the phone line |
| 00:57:20 | dirkx | Hmm- ok so this DSP is not a normal 1600 template which has the two in/out and one out ? |
| 00:57:28 | scanline | That phone line channel is designed to be a modem, but we could also use it for an answering machine type program |
| 00:57:57 | dirkx | Ok - I get it now. I was confused - as I assumed that the DSP did the answering stuff like with the DSP1620's. |
| 00:58:06 | scanline | dirkx: I don't know what the capabilities of the DSP chip itself has, and it doesn't really matter... We can't reprogram it |
| 01:00:13 | GPSFan | scanline: & dirkx: take a look at: http://tuxscreen.net/download/schematics/cave_drawings.jpg it's a bit dark, but the DSP section is pretty accouate as to what's inside. |
| 01:00:47 | GPSFan | scanline: there is however no flash inside. |
| 01:00:59 | scanline | Ah, i was wondering what that diagram was :) |
| 01:01:25 | Keiser_Soze | ** can anyone hear me? ** |
| 01:01:50 | scanline | Keiser_Soze: yes |
| 01:01:56 | Keiser_Soze | I must be acting like a boring newby. Sorry. |
| 01:02:04 | GPSFan | scanline: I've been cleaning it up, but it grew a bit to 10MB. ;>( |
| 01:02:15 | Keiser_Soze | I just wanted to ask how to order a tuxscreen using paypal. |
| 01:02:59 | scanline | Keiser_Soze: There's a paypal button on tuxscreen.net |
| 01:03:04 | scanline | GPSFan: What grew to 10MB? |
| 01:04:07 | GPSFan | scanline: the jpeg, I took it into photoshop & diddled with the contrast & brightness, then did a little creative erasing, a lot more readable, but big. |
| 01:04:39 | scanline | Maybe I'll try taking gimp to it... |
| 01:04:53 | Keiser_Soze | sorry, I don't see an paypal button. |
| 01:05:31 | GPSFan | scanline: as is all you really have to do is diddle the contrast & brightness, and a lot more info just pops out. |
| 01:06:05 | scanline | Keiser_Soze: That's odd, it seems to have disappeared... Still, there are ordering instructions on the Wiki |
| 01:08:00 | boodle | BZFlag: for the hell of it (and I could care less about wince) I tried fdisk on /dev/hda and it seems to have worked... so flashable flashroms (16mb) |
| 01:09:16 | dirkx | Aye - that is nice. Yes - that makes sense. Weird - as those capabilities match more with a 1603 that with a 165x series. |
| 01:09:56 | dirkx | The cave diagram I mean. Nice. |
| 01:19:03 | Keiser_Soze | umm. I starting to feel really stupid. I didn't find any paypal ordering instructions, beyond what the amount is and the email address (Sales@TuxScreen.net). Do I just do a "send money"? |
| 01:19:13 | scanline | yes |
| 01:19:55 | Keiser_Soze | ok, great. i kept looking for an ordering form. |
| 01:21:57 | BZFlag | boodle: yep. it's a sandisk. |
| 01:22:22 | boodle | BZFlag: btw, do you know what kind of audio chip it has? |
| 01:32:01 | Russ | is here |
| 01:42:13 | Keiser_Soze | has anyone use a PCMCIA<->RS232 serial adapter? |
| 01:43:34 | dirkx | These are common in the industrial/process industry. See http://www.pretec.com/ |
| 01:47:25 | dirkx | I quite like the SL0703 series; see socket communcations or google for SL0703 two/quad port. |
| 01:49:27 | dirkx | http://www.socketcom.com -> prducts -> ruggedized -> battery friendly -> 1,2 - 16 ports pcmcia and cardbus |
| 01:49:40 | dirkx | works with freebsd/netbsd/openbsd - so linux should propably work too. |
| 01:49:56 | Keiser_Soze | cool. I found a dual for only 169 on pricewatch. |
| 01:50:11 | Keiser_Soze | is a driver of some sort needed? |
| 01:51:01 | Keiser_Soze | and does the application using it think it's using a serial port? |
| 01:51:14 | dirkx | No idea - I am no linux expert. On *BSD you map into the serial UART support - so the driver is just a ptr which is in the default install - and you just plug in the card and it is a go. |
| 01:54:14 | Russ | dirkx: what were you wanting to know? |
| 01:55:08 | dirkx | Eh- I have two puzzles; - is there any known issue with the udhcpc client - it seems to truncate packets and I am trying to find out if it is the kernel stack or the application or my network. |
| 01:55:46 | dirkx | The second issue is the speaker phone button - I am wondering how I get the phone to go off-hook and wired to the speaker phone when I just press the button iwth the hook in its cradle. Right now I just get a feeping loopback sound :-) |
| 01:56:15 | Russ | what do you mean by truncate? |
| 01:56:28 | dirkx | (I see the DHCP problem on wireless and with two types of wired cards). |
| 01:57:13 | dirkx | Not sure - tcpdump on OpenBSD report the packets as corrupted/early terminated - and in general the reply is not to the linking of ucdhpc either - it does not ack/recognize the offer. |
| 01:59:40 | EGg0 | is away: I'm reading a book. Do you remember books? |
| 02:00:42 | Russ | can you pass me a dump? |
| 02:00:57 | dirkx | Sure - tcpdump format fine ? |
| 02:01:01 | Russ | ya |
| 02:08:49 | dirkx | /dcc send russ |
| 02:11:16 | dirkx | singing of to change irc tool. |
| 02:11:49 | scanline | GPSFan: http://tuxscreen.net/download/schematics/cave_drawings.cleaned.jpg |
| 02:12:29 | Russ | nifty |
| 02:12:57 | scanline | Not perfect, but much easier to read |
| 03:02:21 | BZFlag | scanline: nice. thanx for leaving the old one. |
| 03:04:27 | scanline | BZFlag: Yeah, there might be somebody who would rather download the 2MB one ;) |
| 03:23:03 | scanline | dirkx: I put a cleaned up version of cave_drawings.jpg on tuxscreen.net |
| 03:23:11 | scanline | dirkx: http://tuxscreen.net/download/schematics/cave_drawings.cleaned.jpg |
| 03:47:40 | dirkx | After a pivot_root - I can run /bin/sh fine - but an exec on /sbin/init gives me a hang. Does that ring a bell with anyone ? |
| 03:49:00 | dirkx | Could that have something to do with /var getting list on action ? |
| 03:49:53 | dirkx | (and with /var, some /dev copy?) |
| 03:50:02 | GPSFan | scanline: WOW, that's nicer than I was able to get. |
| 03:50:18 | scanline | GPSFan: I fiddled with the contrast and cut off the extra grey areas |
| 03:52:29 | GPSFan | ;>) |
| 03:57:06 | GPSFan | scanline: I've just a few more areas in the power section the comms board to ring out and the schematics will be done, then I'll be able to put my Tux back together and blob it. |
| 03:57:52 | scanline | Woohoo :) |
| 03:58:59 | scanline | You going to build a shanip2 for your Tux? |
| 03:59:50 | GPSFan | scanline: then the fun starts, the shanip2, 5v pcmcia mod and possibly a software reset ckt. |
| 04:00:26 | GPSFan | scanline: followed immediately by a crash course in Linux. ;>) |
| 04:00:40 | scanline | hehe. No linux for you yet? |
| 04:01:43 | GPSFan | well, I've did some Linux to update Inferno, and have messed around a bit, but this is serious stuff, cross compiling, etc. |
| 04:04:20 | scanline | Well, buildroot makes things easy... when it works ;) |
| 04:04:57 | dirkx | And it *almost* works in the netbsd cross compile chain - now that is a clean environment :-) |
| 04:07:23 | GPSFan | scanline: what is a good distribution to use? I've setup Slackware before, even to the extent of compiling a kernel, and installing a Proxim wireless driver & card. |
| 04:08:10 | scanline | GPSFan: My favorite is debian. As far as installing goes, it's a little on the hard side, but if you're used to slackware you should be fine |
| 04:08:36 | GPSFan | scanline: which rev, or just the latest? |
| 04:08:53 | scanline | GPSFan: Use the unstable branch. The other stuff is too old |
| 04:09:16 | dirkx | ls |
| 04:09:49 | dirkx | Any reason why the screen should flicker *a lot* during an NFS mount ? |
| 04:16:08 | GPSFan | scanline: how much space does the whole thing need? |
| 04:16:20 | dirkx | ls |
| 04:16:46 | scanline | GPSFan: it depends on how much you install... anywhere from 20MB to several GB |
| 04:17:21 | scanline | GPSFan: I'd say a gigabyte or two should be sufficient for a good sized linux install |
| 04:17:29 | GPSFan | scanline: well to get a reasonable start so I can xcompile all the arm stuff for buildroot etc. |
| 04:17:35 | GPSFan | k |
| 04:17:36 | dirkx | GPSFan: I've got about 80Mb here for a machine which is purely set up to work with the tuxscreen. |
| 04:17:59 | dirkx | Sorry - that is 800Mb - debian is the installation. |
| 04:18:06 | dirkx | ls |
| 04:18:35 | GPSFan | thought that 80M seemed a bit light! ;>) |
| 04:19:08 | scanline | Hehe. I used to do heavy development on a machine with Slackware and a 150MB disk |
| 04:19:59 | dirkx | Yes - the slackware machine I set up to do the same with is around that size: 118731Kb |
| 04:20:23 | GPSFan | scanline: ok, in the unstable branch there is hurd-i386 and i386, what's the difference? |
| 04:25:10 | GPSFan | scanline: gotta go cu later. ;>) |
| 04:28:11 | scanline | GPSFan: Bye... hurd-i386 uses the Hurd kernel, i386 uses the Linux kernel |
| 04:47:27 | dirkx | What is the device name which needs to be created for the touchscreen ? |
| 04:48:03 | scanline | /dev/ucb1200-ts |
| 04:48:14 | scanline | c 11 0 |
| 04:48:43 | dirkx | Hm - that is what I thoughd. Somehow pgserver is not picking it up. Who generates those in /dev ? the kernel or is there a hidden mknod script ? |
| 04:49:07 | scanline | Is the device node there? |
| 04:49:22 | scanline | Also, do you have a /etc/pgserver.conf? |
| 04:49:45 | scanline | I don't know what generates them... |
| 04:49:47 | dirkx | Yes - that is auto generated ? (the devince) and I've got /etc/pgserver.conf and a compiled set of themes. Screen comes up fine with logo and bar. |
| 04:50:13 | scanline | You're using GoRK's jffs2 image? |
| 04:50:39 | dirkx | No - I am using buildroot-tux with a CVS checkout of pg-gui compiled. |
| 04:50:57 | dirkx | Or is there not yet enough functionality in buildroot-tux ? |
| 04:51:10 | scanline | Did you use the profile.tuxscreen when you compiled pgserver? |
| 04:52:55 | dirkx | yep - I used:.//configure --host=arm-uclibc --build=i686-pc-linux-gnu --with-prefix=arm-uclibc --with-profile=profiles/profile.tuxscreen --with-extra-ipath=/home/dirkx/ |
| 04:52:55 | dirkx | tuxscrteen/jpeg-6b --with-extra-lpath=/home/dirkx/tuxscrteen/jpeg-6b |
| 04:53:29 | scanline | Ok. So you should have all the right drivers compiled in... run 'pgserver -l' to be sure |
| 04:53:42 | dirkx | I get: |
| 04:53:43 | dirkx | # /usr/local/bin/pgserver -l |
| 04:53:43 | dirkx | Video drivers: fbdev |
| 04:53:43 | dirkx | Input drivers: tuxts ttykb |
| 04:53:43 | dirkx | Fonts: Utopia25[n] Lucida10[nd] Helvetica8[n] Helvetica12[nbd] Helvetica10[nd] Clean6[nf] Clean12[nfd] |
| 04:53:44 | dirkx | Optional widgets: Terminal Canvas |
| 04:53:46 | dirkx | Bitmap formats: PNM[dl] JPEG[dl] BMP[dl] |
| 04:53:54 | scanline | Hmm... that looks fine |
| 04:54:03 | scanline | have you calibrated your touchscreen? |
| 04:54:18 | dirkx | Nope. How does one do that ? |
| 04:54:25 | scanline | Ah, that's the problem :) |
| 04:54:31 | scanline | Snag fbtuxcal from CVS |
| 04:54:34 | dirkx | Well - done under inferno - but I guess that does not count. |
| 04:54:38 | scanline | right |
| 04:54:50 | dirkx | Gotcha. Ok - me comningling. |
| 04:54:59 | scanline | fbtuxcal works with any of the GUIs |
| 04:56:58 | dirkx | Ah - nice. |
| 04:57:30 | dirkx | Right - what do I do with the output ? |
| 04:59:55 | scanline | Nothing |
| 05:00:04 | scanline | it saves a file with the calibration, and loads it into the kernel |
| 05:00:11 | dirkx | Itgets saves so |
| 05:00:15 | dirkx | does it get saved to disk ? |
| 05:00:31 | scanline | /etc/tscal or something like that... |
| 05:00:42 | scanline | Ah, /etc/tsdata |
| 05:01:11 | scanline | You still need to run fbtuxcal to load the data into the kernel every time you boot though |
| 05:01:18 | scanline | it's something that should go into the init scripts |
| 05:01:30 | dirkx | Ok - so it would need write access to the disk - it is not getting that. lets make sure it can :-). |
| 05:01:42 | scanline | BTW, GoRK made jffs2 and kenrnel images for a working PicoGUI installation |
| 05:02:06 | scanline | And he was working on getting it into buildroot |
| 05:02:34 | dirkx | Right - but I need to get wireless support in there, IPsec support, and I'd like to run it diskless from a server. I.e. I pivot_root to an nfs mount |
| 05:03:11 | scanline | ok |
| 05:03:26 | scanline | picogui is network transparent :) |
| 05:04:24 | dirkx | Assuming you add bits into the kernel :-) |
| 05:04:46 | scanline | which bits? |
| 05:05:10 | dirkx | netlink support or something and the driver for the prism2 pcmcia card (for this case). |
| 05:15:51 | dirkx | Ah - here we go. pgtuxphone springs to live :-0 |
| 05:15:54 | dirkx | nice. |
| 05:16:10 | scanline | It's still pretty early in development |
| 05:16:30 | scanline | I'm probably going to rewrite it soon |
| 05:16:31 | dirkx | It works :-) which is a very good start. |
| 05:16:34 | scanline | yep |
| 05:16:57 | dirkx | How do you guys develop - on the tuxscreen - crosscompile or network mounted disk ? |
| 05:17:29 | scanline | Well, you can't compile on the tuxscreen itself unless you have a microdrive to put swap space on |
| 05:17:39 | scanline | so AFAIK everybody cross-compiles |
| 05:18:25 | scanline | At least with PicoGUI apps, if there's no arm-specific code in the app you can corss-display it over ethernet |
| 05:18:40 | dirkx | Hmm - I tried compiling over NFS mount on 100Mbit and it was bearable speedy. But crosscompiling seems easier. It is just the debugger tools - how do you folks do that - to get a quick compile-try-retry cycle ? |
| 05:18:50 | scanline | Or run it on a pgserver set up with the same video mode and themes as the tuxscreen |
| 05:18:54 | dirkx | Ah - that is true - had not thoughd about that. |
| 05:19:17 | scanline | Well, i think the best debugging tool we have so far is printf :) |
| 05:19:21 | scanline | But usually you can devel |
| 05:19:24 | scanline | ack |
| 05:19:33 | scanline | Usually you can develop apps on a desktop and just recompile for ARM |
| 05:20:05 | scanline | To work on pgtuxphone, for example, I run tuxphone and pgserver on the Tux, pgtuxphone on my desktop, and they all communicate via TCP/IP |
| 05:20:15 | dirkx | Hmm - I guess I am just spoiled by netbsd - where you always have your full ddd/gdb/gui debugger on the serial :-) |
| 05:20:27 | dirkx | And what do you make pgtuxphone then talk to - back to the tuxphone at the box ? |
| 05:20:30 | scanline | likes ddd |
| 05:20:41 | scanline | yes |
| 05:20:58 | scanline | you can use environment vars to tell it which pgserver and tuxphone daemon to connect to |
| 05:51:16 | dirkx | in /etc/exports - /tux (rw,no_root_squash) gives me a RO mount pt; just "/tux" works apart from the root uid mapping. What is happening ? |
| 06:07:22 | jacquesZzZz | anyone know the melting point of solder? |
| 06:07:57 | dirkx | Depends heavily on the alloy mox - it is typically on the roll - usually in a black triangle. Multiply by 10 for C'. |
| 06:09:50 | jacquesZzZz | this is 63/37 |
| 06:10:00 | scanline | ooh, new release of SDL! |
| 06:10:19 | jacquesZzZz | 300C seems to be prety good |
| 06:27:18 | jacquesZzZz | 350 is better |
| 07:00:38 | jacquesZzZz | 400 is even better |
| 07:02:21 | scanline | jacquesZzZz: Trying out a variable temp. iron? |
| 07:08:20 | jacquesZzZz | scanline: yep a hakko 936 ESD |
| 07:08:36 | jacquesZzZz | scanline: with a 0.2mm tip |
| 07:08:56 | jacquesZzZz | I had to raise the temp to compensate for the tip |
| 07:09:11 | jacquesZzZz | I now have one flash-unlocked TuxScreen!!! |
| 07:09:23 | scanline | nice |
| 07:11:17 | jacquesZzZz | I wonder what the proper blob image to use is? |
| 07:12:44 | | rumour has it blob is the Boot Loader OBject |
| 07:12:44 | jacquesZzZz | ibot: blob |
| 07:12:54 | jacquesZzZz | :-\ |
| 07:13:16 | jacquesZzZz | scanline: which blob are you running? |
| 07:13:42 | scanline | jacquesZzZz: 0.4, from the tuxscreen.net download directory |
| 07:13:49 | jacquesZzZz | scanline: thanks |
| 07:13:56 | scanline | I'm going to be leaving to watch the meteor shower soon... |
| 07:15:21 | jacques | hmm, I think it's cloudy here |
| 07:18:02 | scanline | is away: watching meteors |
| 07:28:49 | jacques | i have bootldr |
| 07:56:09 | jacques | damn flashing the ramdisk takes forever - is it because it has to erase the whole partition? |
| 07:58:03 | BZFlag | jacques: yes. but there may also be a timing issue with blob. |
| 07:58:06 | BZFlag | I think it |
| 07:58:25 | BZFlag | I think it is faster sometimes. can't explain why though. |
| 07:58:39 | jacques | the kernel > 800kB went really quick |
| 07:58:50 | jacques | but the ramdisk ~ 700kB took forever |
| 07:59:53 | jacques | types boot for the first time. |
| 08:00:16 | jacques | coo,, it booted :-D |
| 08:00:49 | jacques | so... what's the root password? |
| 08:02:19 | jacques | ah, there is no password |
| 08:09:37 | | jacques: what? |
| 08:09:37 | jacques | ibot: uclibc |
| 08:10:08 | jacques | I'm doing cvs buildroot and tinylogin is not compiling |
| 08:10:15 | jacques | it compiled a few days ago dammit |
| 08:10:26 | jacques | /usr/i386-linux-uclibc/usr/bin/i386-uclibc-gcc -Wall -O2 -fomit-frame-pointer -D_GNU_SOURCE -DTLG_VER='"1.00"' -DTLG_BT='"2001.11.18-08:58+0000"' -DUSE_SYSTEM_PWD_GRP -c -o tinylogin.o tinylogin.c |
| 08:10:26 | jacques | make[1]: /usr/i386-linux-uclibc/usr/bin/i386-uclibc-gcc: Command not found |
| 08:10:26 | jacques | make[1]: *** [tinylogin.o] Error 127 |
| 08:10:43 | jacques | why is is looking for /usr/i386-linux-uclibc/usr/bin/i386-uclibc-gcc ?? |
| 08:27:27 | BZFlag | you're not in buildroot-tux ? |
| 08:27:53 | jacques | BZFlag: everything built except for tinylogin |
| 08:28:18 | BZFlag | and you were in buildroot-tux and just typed make ? |
| 08:28:19 | jacques | BZFlag: tinylogin built a few days ago, so I can only assume something changed in cvs in the last couple of days |
| 08:28:33 | jacques | BZFlag: yep, just like I did it several times a few days ago |
| 08:28:56 | jacques | /usr/i386-linux-uclibc/usr/bin/i386-uclibc-gcc is obviously wrong - I have no idea where it's getting that path |
| 08:28:58 | BZFlag | hmmm.. don't know then. I'll try it later. must sleep. |
| 08:29:05 | jacques | ok |
| 08:29:12 | BZFlag | nite |
| 08:29:29 | jacques | times like this I wish there were more frequent working snapshots of cvs |
| 08:41:28 | jacques | anyone awake know why my pcmcia network card is not getting a device associated with it? |
| 10:20:21 | jacques | anyone awake? |
| 10:24:15 | threeme3 | goodmurning |
| 10:25:30 | jacques | hi |
| 10:25:48 | jacques | threeme3: do you have a tuxscreen working with pcmcia networking? |
| 10:26:57 | threeme3 | no, but i think that is not so tuxscreen specific if you use linux |
| 10:27:21 | jacques | heh |
| 10:27:43 | threeme3 | did you install pcmcia support ? |
| 10:27:44 | jacques | normal linux does not assume there is an NFS mount somewhere and refuse to boot if there isnt |
| 10:28:14 | jacques | now it's sitting there broadcasting on the network looking for its NFS mount |
| 10:28:22 | jacques | and does not seem to be going to time out |
| 10:29:41 | jacques | guess i have to reboot without network card and find the script that's trying to do that |
| 10:29:58 | threeme3 | i see, why is it trying to mount NFS ? |
| 10:30:23 | threeme3 | yep, i think too jacques |
| 10:30:42 | jacques | ok, rebooted without card in |
| 10:30:55 | jacques | i cant wait till we have a modular kernel |
| 10:31:02 | jacques | all drivers are linked into kernel right now |
| 10:31:14 | jacques | which means no dynamic pcmcia usage AFAICT |
| 10:31:23 | jacques | no hot swapping i mean |
| 10:31:25 | threeme3 | why is it ? |
| 10:31:42 | threeme3 | it is not so difficult to make a modular kernel |
| 10:32:02 | jacques | because right now there is a kernel partition and a root partition |
| 10:32:14 | jacques | and the kernel partition is 1MB (I dont think it can be any smaller) |
| 10:32:29 | jacques | so for most efficient use of space, the kernel needs to be as big as possible |
| 10:32:38 | jacques | (and still < 1MB) |
| 10:32:47 | jacques | so they built everything in that they could |
| 10:32:55 | jacques | and did not include any modules on the root image |
| 10:34:45 | threeme3 | mmm.. i see... |
| 10:35:28 | threeme3 | maybe it is possible to create a filesystem in the kernel partition |
| 10:36:18 | threeme3 | why is kernel partition 1MB ? is it because of flash size ? |
| 10:38:14 | jacques | yes, I believe so |
| 10:38:22 | jacques | I believe it cannot be any smaller |
| 10:38:46 | jacques | maybe we can do like the ipaq and mod blob to boot from jffs2 |
| 10:39:02 | jacques | then we only need bootldr and root |
| 10:47:14 | jacques | threeme3: look what I found in /linuxrc : |
| 10:47:17 | jacques | # do we have an ethernet card? |
| 10:47:18 | jacques | if ifconfig eth0 | grep -i hwaddr ; then |
| 10:47:18 | jacques | echo found eth0 |
| 10:47:18 | jacques | # change these three lines! |
| 10:47:18 | jacques | ifconfig eth0 192.168.87.6 up |
| 10:47:19 | jacques | route add default gw 192.168.87.1 |
| 10:47:21 | jacques | mount -o nolock 192.168.87.1:/home/potato-arm /mnt |
| 10:47:24 | jacques | if test -x /mnt/linuxrc ; then |
| 10:47:25 | jacques | umount /proc |
| 10:47:28 | jacques | cd /mnt/ |
| 10:47:29 | jacques | pivot_root /mnt /mnt/initrd && exec /linuxrc |
| 10:47:32 | jacques | mount /proc |
| 10:47:33 | jacques | fi |
| 10:47:35 | jacques | umount /mnt |
| 10:47:38 | jacques | fi |
| 10:53:26 | scanline | is back (gone 03:35:23) |
| 10:57:28 | jacques | hi scanline |
| 10:57:49 | scanline | Hi. I was out with two friends watching the meteor shower |
| 10:57:53 | scanline | Really cool :) |
| 10:58:09 | jacques | nice |
| 11:31:06 | threeme3 | jacques: will it boot now? |
| 11:31:29 | jacques | threeme3: I modified the linuxrc and am just about to try |
| 11:32:45 | jacques | yep it boots now |
| 11:33:49 | jacques | network works too but there's some DNS problem |
| 11:34:32 | jacques | # nslookup www.apple.com |
| 11:34:32 | jacques | Server: default |
| 11:34:32 | jacques | Address: default |
| 11:34:33 | jacques | nslookup: www.apple.com: Unknown error |
| 11:34:36 | jacques | I love unknown errors |
| 11:34:42 | jacques | so easy to debug |
| 11:40:02 | threeme3 | hihi |
| 11:58:14 | jacques | i wish dns was working |
| 12:14:13 | jacques | pardon me but buildroot is pretty stupid |
| 12:14:23 | jacques | if anything goes wrong, one is pretty screwed |
| 12:43:26 | EGg0 | is back (gone 10:43:42) |
| 12:50:31 | SteveD | discovers that the PFI pin IS required |
| 12:50:51 | SteveD | without it, the wheaties shuts itself down |
| 13:22:10 | EGg0 | jacques which Linux distro are you using? |
| 13:22:53 | jacques | EGg0: redhat 7.1 |
| 13:23:37 | jacques | biab - dinner |
| 13:23:44 | EGg0 | Hrm... I'm using Mandrake. I thought that my serial port might have been set up some funky way since it's pretty much automagically done. |
| 13:43:56 | jacques | EGg0: serial device permissions are OK? |
| 13:50:08 | EGg0 | jacques Yeah, they are. |
| 13:50:20 | EGg0 | I even tried it with root. |
| 13:59:12 | EGg0 | jacques Let me ask you a stupid question... You are connecting the null modem cable to the DB9 connector on your tuxscreen, right? |
| 14:29:38 | jacques | EGg0: yes |
| 14:30:03 | jacques | EGg0: you are sure your cable is correct? |
| 14:30:20 | jacques | EGg0: there are more than one type of null-modem adaptor |
| 14:31:12 | EGg0 | jacques Well, I had assumed it was, but I'm thinking now that it may not be. |
| 14:31:26 | EGg0 | I didn't realize there was more than one kind. |
| 14:32:25 | jacques | EGg0: yeah, they differ by how they wire the hardware flow control lines I think |
| 14:33:02 | EGg0 | Hrm... Well, what's the skinny on what I want? I'll wireup one myself. |
| 14:33:11 | EGg0 | If you know. |
| 14:35:44 | EGg0 | I wonder if I pull wires 2 and 3 off this cable and wire it straight... Maybe there's confusion with everything else that's wired up. |
| 14:36:02 | EGg0 | I was lazy and bought a cable. Guess it serves me right. |
| 14:39:35 | jacques | do you have the package for the cable? |
| 14:39:41 | EGg0 | Yeah. |
| 14:39:51 | jacques | what does it say exactly? |
| 14:40:04 | jacques | is there any indication of what type of null-modem it is? |
| 14:40:10 | jacques | or how it's wired? |
| 14:41:07 | jacques | heh, it's even possible that the cable is mis-wired |
| 14:41:29 | EGg0 | Well, all it says is that "this cable has a crossed "null-modem" wiring design. and that (of course) it was designed for use with Windows Direct Cable Connection software. No real info. I'm pulling my multimeter to check it now. Bastards :-) |
| 14:43:22 | jacques | windows... what? |
| 14:44:40 | jacques | EGg0: I have a straight-through db9-db9 cable with a null-modem adaptor |
| 14:45:01 | jacques | EGg0: I'm pretty sure my adaptor only has 2/3 crossed |
| 14:46:16 | EGg0 | THat's what I'm thinking. They've probably got all kinds of crap wired up in this for windows whatever. I don't have a null modem adapter, but I'm going to wire up a hack to see if it works. |
| 14:47:07 | jacques | EGg0: yeah, I would. or you might wanna just get an adaptor and a straight-through cable |
| 14:47:17 | jacques | depending on which is faster |
| 14:48:02 | EGg0 | Sunday in Charleston, SC, USA- Buckle of the Bible belt. I've got a better chance finding a porn store open today than an electronics anything :-) |
| 14:48:10 | EGg0 | I'm pulling my soldering gun out. |
| 14:48:29 | EGg0 | We shut down here on Sunday. |
| 14:49:23 | jacques | EGg0: sorry to hear that |
| 14:49:33 | jacques | Sunday is half the weekend |
| 14:57:29 | jacques | EGg0: HOLY CRAP |
| 14:57:55 | jacques | EGg0: look at this page: http://www.helmig.com/j_helmig/dcccable.htm |
| 14:58:23 | EGg0 | Going... |
| 14:58:25 | jacques | Serial cable: |
| 14:58:25 | jacques | Be care: DCC requires for Serial connections a "real" NULL-Modem cable with ALL wires, where you can have either a 9-pin(DB9) or 25-pin(DB25) serial connection. This information is according to the Microsoft Knowledge Base |
| 14:58:25 | jacques | article Q142324 and a TechNet article Appendix B - RAS Cabling: |
| 14:58:25 | jacques | Signal: 9-Pin 25-pin 25-pin 9-pin Signal |
| 14:58:25 | jacques | Ground 5 7 <- the ONLY not crossed wire-> 7 5 Ground |
| 14:58:26 | jacques | Transmit 3 2 <----> 3 2 Receive |
| 14:58:28 | jacques | Receive 2 3 <----> 2 3 Transmit |
| 14:58:30 | jacques | RTS 7 4 <----> 5 8 CTS |
| 14:58:32 | jacques | DSR,CD 1,6 6,8 <----> 20 4 DTR |
| 14:58:34 | jacques | CTS 8 5 <----> 4 7 RTS |
| 14:58:36 | jacques | DTR 4 20 <----> 6,8 6,1 DSR,CD |
| 14:58:41 | jacques | sorry |
| 14:58:42 | jacques | shit |
| 14:58:52 | jacques | just look at the page - it took forever to come up for me but look at the serial cable wiring diagram |
| 15:00:04 | jacques | anyway, it looks to me like your cable is not a null-modem as the world which is not microsoft knows it |
| 15:00:26 | EGg0 | That's what I was afraid of. |
| 15:00:37 | EGg0 | Dammit I hate Micro$oft. |
| 15:01:03 | EGg0 | is repackaging his gate-modem cable for return... |
| 15:01:38 | jacques | :-\ |
| 15:01:58 | jacques | I hope I'm not steering you wrong, but that's what it looks like to me |
| 15:02:21 | EGg0 | jacques You're not steering me wrong. Multimeter tests confirm it. |
| 15:02:51 | jacques | well I'm happy that it looks like your Tux is OK and it's just your cable that's bad |
| 15:04:16 | jacques | heh, I found someone else having the busybox DNS problem on the mailing list |
| 15:04:29 | jacques | the only reply I saw was "compile with uclibc" |
| 15:04:30 | EGg0 | Woohoo! Rat SHack opens today at noon (1 hr) and they have a null modem adapter in stock! |
| 15:04:43 | jacques | EGg0: excellent :-) |
| 15:05:09 | EGg0 | runs down to the garage to get his TuxScreen from under his car :-) |
| 15:06:39 | EGg0 | God, it urks me how Micro$oft calls that a "real" null-modem cable. |
| 15:19:36 | prpplague | morning all |
| 15:20:23 | EGg0 | prpplague: I think I found my problem. |
| 15:20:58 | EGg0 | I have a stupid made for Micro$oft "real" null-modem cable, with all kinds of wire crossing going on. |
| 15:22:50 | prpplague | EGg0: doh |
| 15:23:10 | prpplague | EGg0: one that has all of the control signals cross wired? |
| 15:23:45 | jacques | prpplague: that's a roger |
| 15:24:28 | EGg0 | prpplague: those bastards. :-) |
| 15:25:27 | prpplague | EGg0: computing according to M$, "if your doing something that doesn't work with our standard, you shouldn't be doing it!" |
| 15:25:43 | EGg0 | Yeah, LOL. |
| 15:27:28 | jacques | that word you keep saying, "standard"? I no think it means what you think it means. |
| 15:40:22 | EGg0 | is away: going to RatShack while my iPAQ grabs the new stack... be right back, jacq! :-) |
| 15:40:33 | jacques | heh, good luck to you |
| 15:40:45 | prpplague | EGg0: don't buy a cable , just get parts! and a soldering iron |
| 15:41:17 | EGg0 | prpplague: Have a nice Weber iron. I was being lazy. serves me right. |
| 15:41:22 | jacques | prpplague: does dns resolution work for you on the tuxscreen? |
| 15:47:00 | jacques | well, bedtime I guess |
| 15:48:51 | prpplague | jacquesZzZz: not connecting my tux to the net right, i'm doing arm dev work on it |
| 16:21:14 | prpplague | a little quite today.... |
| 16:21:53 | prpplague | what you guys don't work 7 days a week with 10-12 hours??? |
| 16:25:39 | EGg0 | is back (gone 00:45:15) |
| 17:28:15 | SteveD | SpaceCoaster: found out why the wheaties wouldn't work on my Tux... |
| 17:29:59 | SteveD | SpaceCoaster: was because I had left the PFI pin unconnected. Got my hands on a 110v convertor and things are much happier... |
| 17:30:58 | prpplague | SteveD: so you got the phone stuff working? and it was the PFI ? |
| 17:33:33 | SpaceCoaster | SteveD: cool, it must be reassuring to have them work and we discover a new piece of info on the these toys. |
| 17:36:09 | SpaceCoaster | SteveD: it makes sense, the DSP was powering itself off, the PFI might be connected to a I/O line on the DSP. |
| 17:39:28 | SteveD | SpaceCoaster: looking at the cavedrawings, there is an input to the DSP about power failure |
| 17:40:12 | SteveD | SpaceCoaster: just need to figure out what I'm supposed to put on the PFI if I'm not using the stock PSU |
| 17:40:26 | SteveD | I did update the wiki with this info |
| 17:40:36 | SpaceCoaster | SteveD: yeah, they were pretty paranoid about this thing being able to make calls when the DSP was inoperative, PFI, batteries etc. |
| 17:41:07 | SteveD | SpaceCoaster: there's some rule about that from the FCC I presume |
| 17:41:53 | SpaceCoaster | Hmm, I have mains powered phones that do nothing when power is out, they come with a note, "do not use this as your only phone". |
| 17:42:31 | SteveD | ah well - maybe its just better that way |
| 17:43:15 | SteveD | is there now a working 2.4.13 build, BTW? |
| 17:43:57 | SpaceCoaster | SteveD: not sure, I tried it last night and ran into problems, but that might just be my setup. |
| 17:45:36 | SteveD | having got things going, should I start walking forward through wheaties versions and try the newer ones? |
| 17:58:09 | SteveD | SpaceCoaster: I still have some niggles with the tuxphone daemon - DSP still seems to reboot sometimes etc - is this code issues? |
| 17:59:21 | SteveD | SpaceCoaster: like, sometimes the speakerphone button takes a few presses, dialling flips back to pulse if you wait too long, and suchlike |
| 18:02:33 | SpaceCoaster | hmm, You need the version of wheaties.c from CVS http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/tuxscreen/kernel/ |
| 18:05:12 | SteveD | currently running wheaties.c v 1.15 |
| 18:05:45 | SteveD | with some extra printk's which could perhaps be causing trouble. I'll get a clean one out the cvs |
| 18:08:14 | SteveD | SpaceCoaster: any support for the idea that the wheaties driver itself should run a thread to do the keepalives? |
| 18:09:31 | SpaceCoaster | it was considered but tuxphone needs to be running to use the DSP for anything so it just complicates the driver, you still can do anything even if it is alive without an application. |
| 18:09:46 | SpaceCoaster | s/can/can not/ |
| 18:11:23 | SteveD | SpaceCoaster: so the tuxphone daemon will be the standard way to the DSP? GUIs should talk to it rather than straight to the wheaties device? |
| 18:19:16 | Keiser_Soze | Does anyone use a PCMCIA harddrive in their Tuxscreen? |
| 18:19:57 | Keiser_Soze | There is a 170MB version here: http://www.softwareandstuff.com/h.comp.viper170.html |
| 18:20:05 | Keiser_Soze | for only 24.95 |
| 18:26:05 | SpaceCoaster | SteveD: correct |
| 18:26:50 | prpplague | Keiser_Soze: i've got a ibm microdrive, but i've not tried it on the tux yey |
| 18:27:05 | prpplague | s/yey/yet |
| 18:28:06 | Keiser_Soze | i was very surprised to see a 170MB drive for $25. hopefully it'll work okay. |
| 18:29:03 | GPSFan | Keiser_Soze: that drive is most likely a type 3, and will not fit in one slot, and when put in the other slot fills them both up. |
| 18:31:09 | Keiser_Soze | yup, it says "PCMCIA Type III ATA Card" |
| 18:31:38 | GPSFan | Keiser_Soze: also is 5V. |
| 18:32:06 | GPSFan | Keiser_Soze: Tangent has an IBM Microdrive and it works ok. |
| 18:32:06 | Keiser_Soze | GPSFAN: isn't there a hack for 5V PCMCIA? |
| 18:32:29 | GPSFan | Keiser_Soze: yes there is. |
| 18:32:30 | prpplague | Keiser_Soze: yes there is plenty of info on the wiki for that hack |
| 18:34:25 | Keiser_Soze | GPS: does the type 3 card take up both slots because of physical size, or because it uses both slots' connections? |
| 18:34:30 | prpplague | get this just had a guy tell me this on the phone "i'd love to switch to linux, but i really don't want to have to buy a new c compilier" |
| 18:36:04 | GPSFan | Keiser_Soze: a type 3 is about 2x as thick as a type 1 or 2, uses only one connector though. |
| 18:40:26 | prpplague | GPSFan: hmm, is this the same kinda stuff that tangent used - http://www.allelectronics.com/cgi-bin/category.cgi?item=TEK-5&type=store&category=445 |
| 18:41:39 | GPSFan | prpplague: yes, allelectronics is a distributor for Techniks. |
| 18:42:16 | prpplague | GPSFan: cool, need to order some |
| 18:44:08 | prpplague | GPSFan: this looks nice - http://www.allelectronics.com/cgi-bin/category.cgi?category=760&item=HPS-5&type=store |
| 18:45:42 | GPSFan | prpplague: these guys have press & peel too: http://www.elexp.com/pro_npb5.htm |
| 18:46:51 | prpplague | GPSFan: ya got them bookmarked |
| 18:47:00 | GPSFan | prpplague: that portable scope's cool, but at 5MHz bandwidth, good for audio & low speed stuff only. but $189.00 isn't bad for that. |
| 18:47:41 | prpplague | GPSFan: i was thinking about for ac line checkout, we do a lot of overseas stuff and the ac power is really dirty |
| 18:48:08 | GPSFan | prpplague: that would be a great use of that scope. |
| 18:48:22 | prpplague | GPSFan: be nice to prove to the customer that the line was the problem |
| 18:49:02 | prpplague | GPSFan: i want a do-it-yourself arm kit |
| 18:49:32 | GPSFan | prpplague: build a LART. |
| 18:51:26 | prpplague | GPSFan: ya thats what i'm thinking, but the cpu part is pretty tricky |
| 18:52:30 | GPSFan | prpplague: yep a 208 pin qfp is not real easy to put on, I've done it, and without a stereo zoom microscope and a steady hand it is next to impossible. |
| 18:53:43 | prpplague | GPSFan: i wish someone would mount the cpus on a socketed board or something where it'd be easier for prototyping and general tinkering around with |
| 19:00:33 | SteveD | SpaceCoaster: that stuff in kernel on the cvs - builds modules - is that now the "preferred" way to include these rather than patching the kernel? |
| 19:55:27 | EGg0 | is away: cooking dinner while I wait to accost Tangent and jacques with questions :-) |
| 19:58:29 | dirkx | Pgui/cross compile exports; on a 'ranlib'-ed .a file - why does arm-uclibc-gcc -o bouncyball -L../../cli_c/lib bouncyball.o -lpgui |
| 19:58:42 | dirkx | work and why does arm-uclibc-gcc -o bouncyball -L../../cli_c/lib -lpgui bouncyball.o |
| 19:58:50 | dirkx | fail - with things like |
| 19:58:55 | dirkx | bouncyball.o: In function `main': |
| 19:58:55 | dirkx | bouncyball.o(.text+0x14): undefined reference to `pgInit' |
| 19:58:56 | dirkx | bouncyball.o(.text+0x14): relocation truncated to fit: R_ARM_PC24 pgInit |
| 19:59:03 | dirkx | i.e. it is just the order which differes. |
| 20:08:56 | dirkx | ARRGAGAGG why does pgui/apps nicely adhere to Makefile.rules in most places -but not in pgboard. Just cobberd /usr/local. :-( |
| 20:10:04 | prpplague | doh |
| 20:18:47 | Russ | dirkx: my bad, some udp stacks are lazy and don't give udp checksums, I'm not checking for that |
| 20:19:23 | dirkx | Ate - let me know if you need me to test something. |
| 20:19:34 | dirkx | Whoa- what is *this* |
| 20:19:34 | dirkx | pgd = c02e4000 |
| 20:19:35 | dirkx | *pgd = c80df001, *pmd = c80df001, *pte = c03dc01f, *ppte = c03dcaae |
| 20:22:20 | Russ | verbose kernel |
| 20:23:01 | dirkx | Cheers - I'll stop worrying about it then. |
| 20:23:17 | Russ | its from when a program segfaults I think |
| 20:23:32 | dirkx | Russ: what do I type to receive that file ? (silly question - I know!) |
| 20:23:45 | Russ | /dcc get russ |
| 20:26:10 | dirkx | nasty - this DCC stuff reveals the end to end IPs. Neat. |
| 20:28:42 | dirkx | Much better: |
| 20:28:42 | dirkx | # ./udhcpc -i wlan0 |
| 20:28:43 | dirkx | Moreton Bay DHCP Client (v0.9.5) started |
| 20:28:43 | dirkx | adapter index 2 |
| 20:28:43 | dirkx | adapter hardware address 00:04:5a:cc:4e:a3 |
| 20:28:43 | dirkx | execle'ing /usr/share/udhcpc/default.script |
| 20:28:45 | dirkx | script /usr/share/udhcpc/default.script failed: Permission denied |
| 20:28:47 | dirkx | adding option 0x35 |
| 20:28:49 | dirkx | adding option 0x37 |
| 20:28:51 | dirkx | Sending discover... |
| 20:28:53 | dirkx | payload length is 252 bytes |
| 20:28:55 | dirkx | oooooh!!! got some! |
| 20:28:57 | dirkx | adding option 0x35 |
| 20:28:59 | dirkx | adding option 0x32 |
| 20:29:01 | dirkx | adding option 0x36 |
| 20:29:03 | dirkx | adding option 0x37 |
| 20:29:05 | dirkx | Sending select for 10.10.0.122... |
| 20:29:07 | dirkx | payload length is 264 bytes |
| 20:29:09 | dirkx | oooooh!!! got some! |
| 20:29:11 | dirkx | Lease of 10.10.0.122 obtained, lease time 7200 |
| 20:29:13 | dirkx | execle'ing /usr/share/udhcpc/default.script |
| 20:29:49 | dirkx | Yup - same for wired and wireless card. Trying NE2000 card... fine too. Works now with SUN's DHCP, the ICS-dhcp2 and with apple's airport dhcp. Thansk ! |
| 20:30:44 | dirkx | Russ: do you have any plans to merge this into the kernel (i.e. do what *bsd does - where you can just boot a kernel - the kernel bootp/dhcp's itself, nfs mounts accordning to given dhcp options - i.e. no disk image needed at any moment). |
| 20:35:10 | Russ | 2.2.x had that option |
| 20:35:24 | Russ | but 2.4.x moved away from that in favor of initrd |
| 20:35:57 | Russ | you might check "script /usr/share/udhcpc/default.script failed: Permission denied" |
| 20:39:56 | scanline | awakens |
| 22:29:44 | ctr | is away: brb - sorting something |
| 22:32:37 | ctr | is back (gone 00:02:52) |
| 22:35:24 | dirkx | Russ: - that script or directory is not there. That is fine - all I am doing is getting an IP - extract the mount path - and then mount root/swap from there. Then rely on the mounted /etc/fstab from there on. |
| 22:36:54 | Russ | udhcpc doesn't touch the interface |
| 22:37:07 | Russ | the scripts do everything to the interface |
| 22:37:45 | Russ | you probably want use the command line paramaters -q -n as well |
| 22:37:47 | dirkx | Russ: looking at that script - the default buildroot-tux leaves it in /root as non exec. |
| 22:41:53 | dirkx | Woa ! Tribute to sun and the java team. Their CDC and CLDC packages for J2ME cross compile cleanly for the arm. |
| 22:51:38 | dirkx | Hmm - cried victory too soon. Threading and semaphores are not quite as complete as java wants it. |
| 22:57:32 | dirkx | |
| 22:57:34 | dirkx | register |
| 23:26:45 | EGg0 | is back (gone 03:31:11) |
| 23:42:12 | prpplague | is off to nassau |
| 23:42:17 | prpplague | later guys |
| 23:42:25 | prpplague | be back in 3-5 days |
| 23:42:48 | SpaceCoaster | prpplague: have fun |
| 23:43:32 | prpplague | SpaceCoaster: i'm going to try, mostly work |
| 23:45:41 | EGg0 | is away: watching a movie with my wife. |
