| 00:43:26 | Sammy | morning all ... |
| 02:19:10 | Sammy | can anyone tell me that this happen when start kernel ? |
| 02:24:01 | Sammy | Warning: unable to open an initial console. |
| 02:24:02 | Sammy | Kernel panic: No init found. Try passing init= option to kernel. |
| 02:24:03 | Sammy | is this the problem with compilering the kernel ? |
| 03:09:03 | joev | is flashing in the ramdisk.. Hurry up! :^) |
| 03:09:31 | joev | Sweet! |
| 03:22:38 | GoRK | evening all |
| 03:24:34 | scanline | hello |
| 03:28:49 | joev | Anyone get a 802.11b card working on a TS? |
| 03:59:45 | GoRK | hmm something on the mainboard |
| 03:59:59 | GoRK | the sa1100 itself perhaps? |
| 04:02:58 | andersee | ? |
| 04:07:34 | GoRK | typing out loud |
| 04:07:39 | GoRK | thinking with my fingers |
| 04:08:05 | andersee | ahh |
| 04:08:21 | GoRK | RedBaron: the post about swapping comm boards or lcd's between working/nonworking ts units did not change the 'workingness' of the touchscreens |
| 04:08:23 | GoRK | err |
| 04:08:28 | GoRK | not redbaron |
| 04:08:32 | GoRK | that was supposed to say re: |
| 04:08:36 | GoRK | stupid autocomplete |
| 04:08:51 | Hildy | but they still all work in inferno, no? |
| 04:09:06 | GoRK | yes |
| 04:09:20 | GoRK | i need to fire up my other phone to test the touchscreen |
| 04:10:03 | GoRK | can the JTAG interface be used for any kind of debugging? Or is it just for flashing? ive got one now and im about to stick it on this phone |
| 04:10:53 | Hildy | I believe it can be used for debugging as well, but don't ask me |
| 04:11:36 | AlHaz | as i understand it, there are a bunch of possible JTAG commands, and any given device with a JTAG connector supports some subset of them |
| 04:12:00 | AlHaz | the entire idea tho is to read & write memory w/o executing code |
| 04:12:11 | andersee | I've used JTAG for kernel debugging on other systems, but it appears this one is limited to flashing stuff only |
| 04:12:38 | Hildy | but JTAG is an on-chip thing, no? |
| 04:12:49 | Hildy | as in, it's a feature of the 1100? |
| 04:20:37 | AlHaz | I think so, I mean I've seen it listed as a feature supported by processors |
| 04:24:22 | andersee | From the jflash Readme file: ./flash provides an interface to be able to program or retrieve the flash memorycontents. Simple redirect the contents of a file to ./flash and the file will be programmed into flash. Simply redirect ./flash to a file and the contents of the flash will be written to the file. Parameter of -a and -b allow the address and number of bytes to read or write to be specified. |
| 04:24:52 | andersee | So yes, we can read and write. But nobody has integreated this with gdb... |
| 04:25:40 | AlHaz | well yeah, but is that a limitation of the interface or a limitation of that code? |
| 04:25:56 | AlHaz | I heard that the interface is a little limited, but not super-limited |
| 04:27:21 | andersee | JTAG itself is fine. This code is simply limited. |
| 04:29:18 | andersee | In theory, the interested programmer could hook in gdb support and we could directly debug the kernel with gdb... That what I was able to do with JTAG when I wrote the kernel for the Ericsson BLIP. But I used a JEENI for the JTAG-gdb interface... |
| 04:30:52 | Sammy | andersee : if I want to make the command only I want to use in busybox , can I do this ? |
| 04:31:21 | andersee | huh? |
| 04:31:33 | andersee | I do not understand your question |
| 04:36:46 | Sammy | busybox can support many command , right ? but if I don't need some of that , can I delete it ? |
| 04:37:43 | AlHaz | you can recompile busbox without those commands |
| 04:37:53 | Sammy | and made less side of busybox ? |
| 04:38:25 | AlHaz | you can make busybox smaller by disabling commands you won't need before you compile it |
| 04:40:52 | Sammy | how ? should I need the source code ? |
| 04:42:44 | GoRK | takes the mainboard out of this dead f0n |
| 04:42:55 | GoRK | in one way i want the touchscreen on this one to work and in another way i don't |
| 04:44:11 | Hildy | why is it dead? |
| 04:44:13 | GoRK | hey i wonder .... would those laptop PCMCIA "docking stations" .. the ones that give you real PCI slots off of the card work in linux? |
| 04:44:31 | GoRK | Hildy: i flashed a bad inferno image on it.. it was corrupted on the tuxscreen site.. quite a while ago.. |
| 04:44:42 | GoRK | so i am finally getting around to putting some headers on it |
| 04:45:21 | Hildy | so JTAG it |
| 04:45:31 | Hildy | gork: no |
| 04:45:33 | GoRK | that is what im doing now |
| 04:45:45 | Hildy | if it gives PCI, it needs CardBus |
| 04:45:57 | GoRK | oh yeah i forgot. i knew that |
| 04:49:39 | GoRK | heh this one has some pencil doodles in it |
| 04:49:50 | Hildy | ponders buying tuxscreens |
| 04:49:54 | GoRK | i guess whoever was supposed to check and initial it couldnt get his/her pencil writing |
| 04:55:01 | andersee | did they doodle in spanish? |
| 04:56:06 | GoRK | nope just a big doodle.. squiggly circles |
| 04:56:15 | GoRK | wow my SA1100 chip has about 4 stickers on it |
| 04:56:56 | GoRK | do i have to apply power to my board to use the JTAG to read/write or not? |
| 04:57:19 | GoRK | ive never used a jtag interface before.. i have always used socketed eeproms or the like |
| 04:57:24 | GoRK | or PIC's |
| 04:57:32 | scanline | Yay for PICs :) |
| 04:58:01 | andersee | I used my JTAG dongle for my first couple of passes at hacking blob |
| 04:58:14 | andersee | You do need to have the phone plugged in... |
| 04:58:29 | andersee | Doesn't much matter what the phone is doing at the time.. |
| 05:00:02 | andersee | Sammy: so what was it you wanted to know about busybox? |
| 05:00:34 | GoRK | cool. well then i better get some scans of this thing before i put it back in the phone.. just in case it's a non working touchscreen :) |
| 05:00:40 | Sammy | I have ramdisk with jffs2 , but it use busybox . |
| 05:01:23 | andersee | Sammy: Ok. Nothing wrong with that. |
| 05:01:30 | Sammy | actually it make by erikm , so I think if I can rebuild smaller busybox |
| 05:01:51 | Sammy | include some command only I want ... |
| 05:02:09 | andersee | Sammy: Sure. Edit busybox/Config.h and comment out the stuff you don't want. Then recompile busybox. |
| 05:02:43 | andersee | Sammy: I do cover that in the busybox/README file... |
| 05:03:01 | Sammy | so I must down the source first , and then do it ? right |
| 05:03:12 | andersee | Sammy: yes |
| 05:03:45 | Sammy | so does tinylogin ? |
| 05:04:44 | andersee | Sammy: yes |
| 05:04:47 | Sammy | umm.. which compiler should I use , for x86 or for arm ? |
| 05:05:17 | andersee | depends what architecture you want to run the binary on. |
| 05:05:41 | Sammy | Ok , thanx andersee . |
| 05:05:58 | Sammy | start search the source ... |
| 05:06:02 | andersee | no problem |
| 05:59:55 | Sammy | andersee : I compiler my MTD but why isn't pass when make mkfs.jffs2 ? |
| 06:03:39 | Sammy | is that anything I miss ? |
| 06:17:41 | andersee | Are you using buildroot-tux? |
| 06:19:07 | GoRK | narf. i dont have my parallel port turned on :( |
| 06:19:16 | GoRK | 02:21:30 up 40 days, 23:02, 2 users, load average: 0.01, 0.03, 0.00 |
| 06:19:35 | andersee | bummer |
| 06:19:48 | andersee | You could build a module and insmod it |
| 06:20:09 | andersee | Or is it disabled in the bios? |
| 06:20:11 | GoRK | no it's in the bios |
| 06:20:12 | GoRK | heh |
| 06:20:15 | andersee | bummer |
| 06:20:20 | andersee | reboot city |
| 06:20:29 | GoRK | jflash says: lpt port 0x0 not available |
| 06:20:40 | GoRK | /proc/ioports lists nothing |
| 06:20:50 | GoRK | when the parport and lp modules are loaded |
| 06:21:27 | GoRK | oh i remember i turned it off to free an irq for my TV tuner card.. blah |
| 06:21:29 | andersee | Ran jflash as root? |
| 06:21:45 | GoRK | but since i can capture over the network from my tivo now, it doesnt really matter |
| 06:22:04 | GoRK | yeah it's reboot time for me |
| 06:25:53 | Sammy | how to make mtd/util ? |
| 06:27:32 | Sammy | andersee : is that step in file README ? |
| 06:28:40 | GoRK | 0778-077a : parport0 |
| 06:28:41 | GoRK | whoot |
| 06:29:01 | andersee | Sammy: do a 'cvs up && rm -rf mtd sources/mtd* && make' in your buildroot-tux directory |
| 06:29:08 | andersee | GoRK: cool |
| 06:30:23 | andersee | GoRK: btw, I found that using jflash was much slower then using the serial cable |
| 06:30:42 | andersee | GoRK: So I just flashed in blob, then used blob to flash everything else |
| 06:31:41 | andersee | GoRK: That was after I had waited about 45 min for jflash to finish flashing all 4 megs... And it was only about half done |
| 06:32:26 | GoRK | andersee: that's what i figured |
| 06:33:09 | GoRK | hey is there a way to load a splash screen w/o jtag? |
| 06:33:21 | andersee | I dunno |
| 06:33:31 | GoRK | i just saw that in there... i guess ill load one up here while im at it |
| 06:33:36 | andersee | I hacked up jflash to disable all that crap since it didn't compile |
| 06:33:49 | GoRK | hmm it compiled totally fine for me |
| 06:34:10 | andersee | Thats because I hacked up jflash... ;-) |
| 06:34:26 | GoRK | oh. heh. so 'splash' won't work on this build from cvs? |
| 06:34:32 | andersee | right |
| 06:34:44 | GoRK | how about the binaries from tuxscreen.net/downloads ? |
| 06:34:49 | GoRK | will that 'splash' work? |
| 06:34:56 | andersee | I didn't touch the docs, and I think I forgot to remove that symlink. |
| 06:35:05 | andersee | GoRK: dunno |
| 06:35:12 | GoRK | oh well ill fool with it later |
| 06:35:21 | GoRK | right now i need to get this phone back from the dead! |
| 06:36:11 | andersee | Lazurus, come forth! |
| 06:36:20 | GoRK | do i run prep? or just flash <file> ? |
| 06:37:54 | andersee | I just ran ./flash < ../tuxscreen-blob |
| 06:37:59 | andersee | or something like that |
| 06:38:50 | GoRK | yeah it's going now |
| 06:39:34 | andersee | blob is nice and small, so that should go quicklt |
| 06:43:03 | GoRK | heh. i thought it wasnt working but my serial cable had fallen out |
| 06:43:24 | andersee | heh |
| 06:43:30 | andersee | happened to me too |
| 06:43:50 | GoRK | this f0n has been LARTed! |
| 06:44:57 | andersee | cool beans! |
| 06:52:35 | andersee | GoRK: so does f0n #2's touchscreen work? |
| 06:53:21 | GoRK | dunno yet about to boot |
| 06:54:56 | GoRK | f0n heh |
| 06:55:22 | GoRK | if it does, it's becoming f0n #1 since it has the bonus of JTAG for when i totally f*** up |
| 06:55:42 | GoRK | x2vnc is about the single greatist program in the whole of the known universe |
| 06:55:58 | andersee | heh |
| 06:56:08 | andersee | That a MW app? |
| 06:56:14 | GoRK | no. it's just an x app |
| 06:56:23 | andersee | what it do? |
| 06:56:36 | GoRK | but it lets me control my linux and windows desktop like a single big multihead machine |
| 06:56:41 | andersee | I use xvncviewer |
| 06:57:01 | andersee | It it different from that? |
| 06:57:35 | GoRK | it sets up a single pixel wide window on the side of the screen and when you mouse into it it starts sending VNC mouse/kbd events to another computer.. it doesnt do any display stuff... only sends events.. |
| 06:58:14 | GoRK | so i can just mouse over to the other monitor and control some windoze apps then mouse back and poke at linux |
| 06:58:16 | andersee | Then how do you know what you are clicking on? |
| 06:58:28 | andersee | two monitors? |
| 06:58:33 | GoRK | the windows box has its own monitor .. yeah |
| 06:58:37 | andersee | ahh |
| 06:58:40 | andersee | gotcha |
| 06:59:08 | andersee | sortof like a KM switch |
| 06:59:11 | GoRK | the only thing it's missing is clipboard sync. that would be the bomb |
| 06:59:28 | GoRK | it's totally transparent though..it's really bizzare to watch |
| 07:00:16 | scanline | GoRK: Nifty. I sometimes use somthing like that between my Linux on my desktop and Mac OS X on my laptop |
| 07:00:31 | GoRK | x2x ? |
| 07:00:39 | GoRK | x2x is the same thing between 2 x sessions |
| 07:00:46 | scanline | well, it's x2vnc also |
| 07:00:51 | scanline | and the OS X vnc server |
| 07:01:20 | scanline | kinda funny having a 3-head system where one head is a laptop :) |
| 07:01:32 | GoRK | when i use my laptop here, i set it up on the left with linux desktop in the middle and use x2x on the left side between X on the laptop and the desktop and x2vnc on the right in windows |
| 07:01:51 | scanline | neat |
| 07:01:52 | GoRK | heh yeah exactly |
| 07:08:50 | scanline | is away: sleep... zzzzzzz |
| 07:09:48 | GoRK | hmm whoops i guess i didnt get power conencted very well or something |
| 07:09:54 | GoRK | coz i cant see linux on the screen at all |
| 07:13:41 | GoRK | yeah touchscreen works on it |
| 07:15:35 | andersee | bummer. |
| 07:15:46 | andersee | So both of yours work |
| 07:15:47 | GoRK | my buddy has another |
| 07:15:53 | GoRK | well i have 3 |
| 07:15:57 | andersee | oh |
| 07:16:15 | GoRK | PBX would be kind of useless w/o a few phones dont you think? :) |
| 07:16:23 | andersee | true 'nuf |
| 07:17:15 | GoRK | not to mention rather hard to test/debug |
| 07:17:55 | GoRK | i think we are going to implement a serial transport for picogui clients... so that the pbx can support phones w/o the need for them to have ethernet |
| 07:19:56 | GoRK | heh jtag is cool |
| 07:33:33 | Arc_Tangnet | wakes up |
| 07:34:52 | GoRK | hacks on x2vnc... clipboard transfer sounds yummy |
| 07:34:55 | GoRK | actually |
| 07:35:12 | GoRK | goes to bed. sleep tastes better than clipboard transfer |
| 07:35:30 | andersee | heh |
| 07:35:43 | GoRK | heh. dont you just love being able to make your computer do what you want it to?!?!? |
| 07:36:23 | Arc_Tangnet | GoRK : Don't know.. never happened |
| 07:36:30 | GoRK | hehehe |
| 07:36:35 | GoRK | later folks, i gotta sleep |
| 07:36:43 | Arc_Tangnet | Night GoRK |
| 07:42:12 | andersee | night |
| 07:47:03 | Arc_Tangnet | Morning ctr |
| 07:47:18 | ctr | mornin |
| 07:53:40 | Arc_Tangnet | Hrm.. no new jobs on jobserve today... Sucks... |
| 07:55:27 | ctr | =( |
| 08:02:20 | Arc_Tangnet | stops being an evil overlord... |
| 11:46:40 | RedBaron | morning |
| 12:38:21 | pattieja | is back (gone 15:31:20) |
| 13:04:11 | ctr | how can i confiure something to have a root prefex of /usr/local but install into /files/fakeroot/usr/local ? |
| 13:05:12 | prpplague | howdy all |
| 13:05:34 | mallum | hello |
| 13:05:43 | RedBaron | Howdy |
| 13:06:23 | prpplague | has free batteries for keyboard |
| 13:06:30 | prpplague | no free fresh |
| 13:06:56 | prpplague | argh, no coffee yet this morning |
| 13:07:21 | mallum | uh oh |
| 13:08:10 | prpplague | i could seem to get inferno remote to connect last night until i realized my keyboard batteries were as dead as disco |
| 13:11:44 | mallum | Ho CosmicPenguin |
| 13:11:49 | mallum | s/ho/hi/ |
| 13:11:52 | CosmicPenguin | mallum: hey |
| 13:12:06 | mallum | was there any progress on the ts stuff last night ? |
| 13:12:11 | ctr | CosmicPenguin: y0 |
| 13:12:18 | ctr | CosmicPenguin: looking at building ferite :) |
| 13:12:30 | CosmicPenguin | A lot of red herrings, but no real leads.... |
| 13:12:47 | ctr | ij |
| 13:12:49 | ctr | ok |
| 13:12:55 | ctr | what host do i use? |
| 13:13:16 | mallum | CosmicPenguin: ok, well thanks for the continued effort :) |
| 13:13:16 | CosmicPenguin | mallum: We know for sure that it is not the USB1200 chip, and it must either be on the processor, or something having to do with the traces from the chip to the processor |
| 13:13:39 | mallum | CosmicPenguin: ok, all that hardware stuff is beyond me .... |
| 13:13:57 | CosmicPenguin | mallum: As it emerges that more people have broken TS than working ones, this is a bigger issue all the time |
| 13:14:37 | mallum | mallum: Id say thats good as it means it'll get sorted but its not really good is it ... know what I mean ? |
| 13:16:38 | CosmicPenguin | mallum: Yep.... The really annoying thing is that this is probably a 1 line fix somewhere |
| 13:17:13 | mallum | yeah |
| 13:17:15 | RedBaron | CosmicPenguin: this USB1200 chip. Is it a USB controller? I thought that the tuxscreen did not have any usb. |
| 13:17:27 | mallum | I've gotta learn more about Linux internals |
| 13:17:47 | mallum | I have some books on the way |
| 13:18:14 | CosmicPenguin | RedBaron: Sorry, I meant to say UCB |
| 13:18:29 | CosmicPenguin | RedBaron: Its an analog to digital controller.... |
| 13:18:32 | RedBaron | CosmicPenguin: Ah. I see :) |
| 13:20:06 | mallum | CosmicPenguin: the UCB1200 is more of less the core of the tuxscreen for handling input, the screen etc ... ? |
| 13:20:15 | ctr | fudge |
| 13:20:31 | mallum | packer ? |
| 13:22:19 | CosmicPenguin | mallum: It handles the touchscreen, the audio, and (if there isn't other hardware available), the telecom. It also provides some GPIO pins |
| 13:22:29 | CosmicPenguin | which the tuxscreen uses for brightness / contrast / backlight control |
| 13:23:27 | mallum | what are GPIO pins ? |
| 13:23:29 | ctr | hrmm |
| 13:24:04 | CosmicPenguin | General Purpose IO pins... they are esentially pins that can be used for anything the developer wants to control.... |
| 13:24:16 | ctr | CosmicPenguin: i bet it's the GPIO stuff |
| 13:25:14 | CosmicPenguin | ctr: We know that the problem lies somwhere between the processor and the UCB1200 chip.... |
| 13:25:51 | CosmicPenguin | Basically, SpaceCoaster discovered that we go into a loop waiting for a bit to be set in the processsor indicating that we had a successful read from the chip... |
| 13:25:54 | CosmicPenguin | We never get it |
| 13:26:02 | mallum | CosmicPenguin: how did you learn about the hardware side of things - did you do a cs degree or did you just 'pick it up' ? |
| 13:26:03 | ctr | hrmm |
| 13:26:21 | ctr | CosmicPenguin: cant we just assume we got it right and carry one? has that been tried? |
| 13:26:49 | CosmicPenguin | mallum: I do have a CS degree, but I do embedded, so you really pck this stuff up over time... |
| 13:27:13 | CosmicPenguin | mallum: Keep in mind that i only understand hardware when it breaks the software, so you can't count on me as a general hardware guy.... :) |
| 13:27:25 | mallum | CosmicPenguin: heh :) |
| 13:27:30 | ctr | hrmm |
| 13:27:51 | ctr | i am going to have to re-write this pcre configure thingy to allow me to do cross compilation |
| 13:27:51 | CosmicPenguin | ctr: SpaceCoaster tried that, and it stopped locking up, but we couldn't read any valid data... |
| 13:27:59 | ctr | CosmicPenguin: k |
| 13:28:02 | ctr | CosmicPenguin: hm |
| 13:31:06 | ctr | CosmicPenguin: could it be the order we are doing things in? |
| 13:31:31 | CosmicPenguin | mallum: I have been fighting with UCB1200 chips for a while now... I've got another hell spawned device on my desk right now that uses the UCB1200.... :( |
| 13:31:40 | CosmicPenguin | ctr: What do you mean? |
| 13:32:45 | ctr | ok |
| 13:32:48 | ctr | now i could be wring |
| 13:32:54 | RedBaron | Does anyone know if there is anything resembling a schematic for the shannon and the communications board? I vaguely recall seeing a link on the tuxscreen site, but cant find it now. |
| 13:33:08 | ctr | but could we be initilising and accessing the hardware in an incvorrect oder? |
| 13:33:16 | ctr | ie |
| 13:42:03 | prpplague | anyone know how to get out of the "Download Software" screen? i've tried the esc+2 but that doesn't seem to be working |
| 13:43:05 | ctr | ok |
| 13:43:17 | ctr | i think iam going to go play a little HalkfLife - BlueShift |
| 13:43:44 | mallum | I got that, but havn't really played it |
| 13:43:47 | mallum | oops |
| 13:54:52 | prpplague | where's BZFlag when you need him? |
| 13:55:12 | RedBaron | prpplague: Probably off trying to have a life :) |
| 13:57:01 | prpplague | RedBaron: whats that? |
| 13:57:51 | prpplague | finaly, got out of that dam screen... |
| 14:00:21 | CosmicPenguin | prpplague: Its still pretty early in the morning here (early being a relative term of course)... |
| 14:01:24 | prpplague | CosmicPenguin: ya i know, i'm just doing a little b-n-c since i've too much other work i should be doing... |
| 14:01:38 | RedBaron | prpplague: BZFlag is probably off trying to have a life when you need him. Just Murphy rearing his ugly head. |
| 14:02:31 | prpplague | RedBaron: i meant "a life, whats that?" |
| 14:02:52 | RedBaron | prpplague: Ah. I dont know...Never had one. |
| 14:03:01 | pattieja | "this is ... a life?" |
| 14:14:22 | prpplague | thinks the name "inferno" is pretty descriptive |
| 14:14:53 | CosmicPenguin | Man, I should really send some money to these guys to try and help stop these service interruptions... |
| 14:19:21 | RedBaron | Whats up with this 'spliting' thing? I used to IRC a long long time ago..but have been on MUD's for a long time. |
| 14:20:28 | CosmicPenguin | RedBaron: From what I am told, when a machine goes down on the network, it has the potential of splitting the network in two, thus isolating folks on one side... |
| 14:21:39 | RedBaron | CosmicPenguin: Oh. the IRC server is a network of servers? |
| 14:23:28 | Tangent | RedBaron : Yep |
| 14:23:36 | Tangent | RedBaron : There's an IRC backbone too :) |
| 14:23:45 | Tangent | RedBaron : I connect to a server in london.. |
| 14:23:57 | RedBaron | Tangent: Heh. Seems like a bit over kill. but why not. :) |
| 14:24:27 | Tangent | RedBaron : Sadly not overkill anymore.. hence the splits when traffic gets busy (DOS attack etc) |
| 14:25:08 | RedBaron | Tangent: *nod* |
| 14:35:05 | RedBaron | Yegawds, windows is a pain in the rear. |
| 14:38:41 | SpaceCoaster | CosmicPenguin: hi, you gave a nice summary to mallum/ctr of yesterdays TouchScreen orgy. |
| 14:39:36 | mallum | ooh err missus |
| 14:40:39 | SpaceCoaster | mallum: did you catch the part where warlord had TWO TuxScreens completely stripped down :-) |
| 14:41:13 | mallum | nope |
| 14:41:16 | mallum | ;) |
| 14:41:51 | SpaceCoaster | it got pretty intense, we even have photos on the wiki |
| 14:42:44 | CosmicPenguin | SpaceCoaster: Yeah, unfortunately it didn't lead to much.... :( |
| 14:45:41 | mallum | SpaceCoaster: thanks for working on it :) |
| 14:51:13 | SpaceCoaster | someone else had this problem in the past, with another arm, dpathirana@sri.crossvue.com, his problem magically vanished at kernel 2.4.4 |
| 14:51:13 | CosmicPenguin | SpaceCoaster: actually, i lied. I think we pretty conclusively proved that the UCB1200 is *not* to blame |
| 14:52:46 | SpaceCoaster | CosmicPenguin: It appears to be the SIB interface that chip, so maybe something do do with the way the is being programmed up SA1100. |
| 14:53:15 | mallum | has anybody tried a 2.4.4 kernel ? |
| 14:53:21 | CosmicPenguin | SpaceCoaster: I agreee. Warlord helped us by proving that the base unit was broken |
| 14:53:32 | CosmicPenguin | And not the board with the chip on it |
| 14:53:58 | CosmicPenguin | SpaceCoaster: and we know its not ADC sync.... |
| 14:53:58 | SpaceCoaster | ahh, I didn't understand that until now. |
| 14:54:18 | SpaceCoaster | what is the ADC sync? |
| 14:56:09 | SpaceCoaster | mallum: I have a cunning plan, I am going to strip the UCB initialisation down to nothing, empty c files, and build it back up, testing as I go |
| 14:56:41 | mallum | SpaceCoaster: sounds good ... |
| 14:56:42 | CosmicPenguin | SpaceCoaster: THe ADC is the handshaking protocol between the kernel and the UCB chip... |
| 14:57:10 | SpaceCoaster | CosmicPenguin: how do we know that's good? |
| 14:57:22 | CosmicPenguin | SpaceCoaster: Apparently, there is a mode where you can sync it with the analog sampling, resulting in less noise or something, I don't understand the electronics... |
| 14:57:46 | CosmicPenguin | Anyway, if sync is enabled on the chip, you must have it enabled in the kernel, or you will get nothing... |
| 14:57:59 | CosmicPenguin | and of course, the reverse is true as well... |
| 14:58:18 | CosmicPenguin | But I had you and andersee turn on SYNC yesterday, and it made no difference.... :( |
| 14:58:49 | SpaceCoaster | We know that we have working hardware (inferno, etc), true? |
| 14:59:16 | CosmicPenguin | SpaceCoaster: true |
| 15:00:23 | prpplague | has anyone tried loading a copy of inferno from a machine that doesn't work to one that does to see if there is a difference in the inferno code? |
| 15:00:48 | prpplague | or atleast do a cksum on the two copies of inferno? |
| 15:01:29 | CosmicPenguin | prpplague: I doubt there is a difference..... |
| 15:01:56 | SpaceCoaster | CosmicPenguin: is it correct to say that on broken machines the UCB1200 is providing no functionality, nothing works? |
| 15:02:30 | CosmicPenguin | SpaceCoaster: No, i would say that the processor cannot talk to the UCB1200 - I now believe the UCB1200 is working fine |
| 15:04:35 | SpaceCoaster | CosmicPenguin: what do you think of my plan to start from a minimal driver? |
| 15:05:20 | CosmicPenguin | SpaceCoaster: Thats good for the long run (since the UCB1200 drivers aren't the greatest), but I am afraid that you wuould do a ton of work and end up at the same result.... |
| 15:05:44 | CosmicPenguin | The UCB1200 drivers do their job... I have proof sitting on my desk right now.... |
| 15:06:05 | SpaceCoaster | CosmicPenguin: It would be just to isolate the problem, then apply the solution to the current drivers. |
| 15:06:18 | CosmicPenguin | SpaceCoaster: Consider something as far fetched as this: |
| 15:06:47 | CosmicPenguin | The SIB is a bus, so technically there can be more than 1 device on the bus, right? |
| 15:06:55 | SpaceCoaster | yup |
| 15:07:09 | CosmicPenguin | So what if some of the spins had the UCB as the first item, and then the DSP as the second? |
| 15:07:11 | scanline | is back (gone 07:58:21) |
| 15:07:31 | CosmicPenguin | And then others had the DSP as the first, and the UCB as the second? What if you need to set up the DSP to talk to the UCB instead of the other way around? |
| 15:08:46 | CosmicPenguin | Thus the Inferno software would always work, because they are always setting up the DSP... |
| 15:09:09 | CosmicPenguin | Thats probably not it - but I think it shows what we're up against.... :( |
| 15:09:21 | prpplague | that was my question |
| 15:13:25 | wej | boodle? like the gin? |
| 15:13:46 | boodle | wej: heh well not exactly but yes.. there is boodles |
| 15:14:11 | wej | likes boodles. |
| 15:30:08 | warlord | Hi |
| 15:30:12 | CosmicPenguin | Doh! My switch on my deck ceased to be...\ |
| 15:30:22 | CosmicPenguin | s/deck/desk/ |
| 15:30:34 | warlord | To whom should I send bug reports about bugs in sub-packages of buildroot-tux? |
| 15:30:50 | CosmicPenguin | warlord: what have you got? |
| 15:31:08 | warlord | Two Makefiles in microwindows are missing "-lm" |
| 15:31:18 | warlord | I wanted to turn 'VERBOSE' to 'Y' in config |
| 15:31:21 | CosmicPenguin | Ok..... Thats my bad.... |
| 15:31:33 | CosmicPenguin | Do you have a patch, or shoud I just fix it? |
| 15:31:34 | warlord | And also uClibc appears to be missing values.h |
| 15:31:54 | CosmicPenguin | warlord: who uses values.h? |
| 15:32:21 | warlord | li_recognizer.c:945: values.h: No such file or directory |
| 15:32:35 | CosmicPenguin | Ah, I hate that thing...... |
| 15:32:45 | warlord | That's in demos/nxscribble |
| 15:33:08 | CosmicPenguin | warlord: New version of that will be forthcoming soon.... |
| 15:33:20 | CosmicPenguin | Not that scribble is very useful on this thing anyway (not right now) |
| 15:33:33 | warlord | True, but it should at least compile ;) |
| 15:33:39 | warlord | Anyways, the -lm bugs are in: |
| 15:33:47 | warlord | demos/nxroach |
| 15:33:57 | warlord | demos/nanox |
| 15:34:07 | warlord | Just need to add -lm to LDFLAGS |
| 15:34:15 | warlord | (otherwise they complain about missing 'sin' and 'cos' functions |
| 15:35:04 | CosmicPenguin | warlord: know what sucks? Those bugs are because of my CEO... :) Wait till I give him that bug report... |
| 15:36:24 | warlord | Because of your CEO?!? |
| 15:36:45 | warlord | Ok, this I gotta hear.. |
| 15:36:50 | AlHazWork | there's nothing like the wisdom of management |
| 15:37:42 | CosmicPenguin | warlord: My CEO (Greg Haerr) is the maintainer of Microwindows..... He's actually a very good programmer.... |
| 15:38:10 | RedBaron | AlHazWork: Sure there is...For example the squawking of a flight of geese. |
| 15:38:50 | prpplague | RedBaron: the sound of a flushing toilet? |
| 15:39:01 | AlHazWork | RedBaron: It's more like the fallout from a flight of geese |
| 15:39:03 | RedBaron | prpplague: Hehe. That too. |
| 15:40:01 | prpplague | well i think i'm about ready to unlock my flash and loaded blob |
| 15:42:09 | CosmicPenguin | warlord: Ok, fixed.... thanks for the report... :) |
| 15:46:16 | warlord | No problem.. |
| 15:49:35 | warlord | (how did you fix the values.h problem?) |
| 15:49:36 | CosmicPenguin | warlord: Well, actually, i didn't fix that one yet.... |
| 15:49:36 | warlord | Ahh. You just fixed the Makefiles and the config? |
| 15:49:36 | CosmicPenguin | warlord: ok, fixed the scribble as well |
| 15:49:58 | prpplague | anyone using tiny-x ? |
| 15:51:01 | warlord | thanks CP |
| 15:51:19 | warlord | How often are the MW tarballs released? |
| 15:52:00 | CosmicPenguin | warlord: Well, officially they are released on the "Whenever greg feels like it" schedule.... |
| 15:52:15 | warlord | LOL |
| 15:52:16 | warlord | I see :) |
| 15:52:18 | CosmicPenguin | Thats why I maintain a seperate tarball for the tuxscreen.... |
| 15:52:24 | warlord | smiles |
| 15:52:40 | warlord | So how come I've found all these bugs all within 12 hours of starting work on this project? ;) |
| 15:52:55 | warlord | Or does nobody else actually read over the build logs? ;) |
| 15:53:14 | CosmicPenguin | warlord: Well, to tell you the honest truth, i really didn't care if nsaver built or not, because I don't use it... :) |
| 15:53:32 | warlord | Hey, I dont use it either (well, as far as I know), |
| 15:53:39 | warlord | but compilation errors are evil nonetheless ;) |
| 15:53:39 | CosmicPenguin | warlord: The demos are sort of useless anyway..... |
| 15:53:46 | RedBaron | warlord: Compile warnings and errors? Pshaw. no need to fret. :) |
| 15:53:59 | CosmicPenguin | If there is something worth using, then I usually copy it into the tinytoe tree for further use... :) |
| 15:54:02 | warlord | Well, compile errors could be hiding other issues. |
| 15:54:31 | warlord | And warnings could be covering other bugs (if you actually use that code) |
| 15:54:39 | warlord | Me, I just personally like "clean compiles" |
| 15:54:58 | warlord | tinytoe? |
| 15:55:03 | CosmicPenguin | Anway, you're good to go now.... |
| 15:55:13 | warlord | Oh, right. |
| 15:55:15 | CosmicPenguin | warlord: tiny Tuxscreen Operating Environment |
| 15:55:23 | warlord | :) |
| 15:56:48 | warlord | Cool. Is there a new tarball I can grab, or will you update the buildroot-tux CVS Makefile as well? |
| 15:57:04 | warlord | is looking forward to the day of CVS commit bits :) |
| 15:57:37 | CosmicPenguin | warlord: All of the above.... it will probably few about an hour or so, becuase at some point, somebody will expect me to do some work at my real job... :) |
| 16:00:30 | warlord | Real Job!? |
| 16:00:32 | warlord | phsaw |
| 16:00:34 | warlord | ;) |
| 16:01:00 | warlord | Also, is there a 'real' bug-reporting address I should be using for things like this? |
| 16:02:09 | CosmicPenguin | warlord: tuxscreen-devel mailing list is best best |
| 16:02:17 | CosmicPenguin | s/best best/best bet/ |
| 16:09:38 | warlord | @tuxscreen.sf.net? |
| 16:10:33 | CosmicPenguin | No - tuxscreen-devel@lists.sourceforge.net See the wiki |
| 16:11:02 | warlord | I've tried. The wiki is so convoluted that it took me an hour to find the wiki page with the links to the development environments. |
| 16:13:01 | AlHazWork | try the search feature |
| 16:13:46 | warlord | That presupposes that I know the phrase I was looking for.. |
| 16:14:19 | warlord | IMHO the wiki is just not arranged well. It's TOO interconnected without a clear over-riding page hierarchy |
| 16:14:42 | prpplague | has been LARTed! |
| 16:17:19 | CosmicPenguin | warlord: Kinda like the internet itself?? :) |
| 16:19:41 | warlord | congrats, prpplague |
| 16:20:24 | warlord | CP: No. The internet is a loose set of different people, tools, and projects. |
| 16:20:40 | warlord | TuxScreen is a single project (well, perhaps it is a group of projects with a single purpose) |
| 16:20:50 | warlord | The project page should help organize the information |
| 16:20:55 | warlord | Not just be a repository for it. |
| 16:25:20 | CosmicPenguin | warlord: Ok, the CVS changes have been made, and new tarballs have been uploaded |
| 16:25:39 | CosmicPenguin | warlord: I'm not calling this a release or anything, I just snuck some bug fixes up there... |
| 16:28:32 | warlord | Of course. :) |
| 16:29:00 | warlord | What change did you make to the linux touchscreen patch? |
| 16:30:06 | CosmicPenguin | warlord: Before, the ioctl() calls scaled the values from their natural value to 0 - 255. I changed that back.... |
| 16:30:21 | CosmicPenguin | Now, brightness goes from 0x0 to 0x07 and contrast from 0x0 - 0x3F.... |
| 16:30:22 | warlord | Ahh, ok. |
| 16:32:03 | warlord | Now, if only we could figure out why the touchscreen only works on some devices... ;) |
| 16:34:27 | CosmicPenguin | Thats the 100,000 dollar question |
| 16:37:21 | warlord | Yep |
| 16:37:43 | warlord | OBTW, your Makefile patch that removed the semi-colon at the end of $(PATCH4COMMAND) causes Make to fail. |
| 16:37:54 | warlord | (at least for me) |
| 16:38:12 | prpplague | is the numberic telephone pad connected to the gpio in a matrix? |
| 16:38:56 | warlord | A better fix is to move the semi-colon outside the quotes in the PATCH4COMMAND definition above |
| 16:39:06 | warlord | Should I email a patch? |
| 16:40:01 | prpplague | i wonder if the ram stick off my intel sa-1100 multimedia board will work with tuxscreen... |
| 16:43:55 | warlord | Also, CP, you didn't change the configuration option for VERBOSE to Y :) |
| 16:44:57 | CosmicPenguin | warlord: Thats gonna have to be a personal option, I think.... |
| 16:45:16 | CosmicPenguin | warlord: do e-mail a patch.... |
| 16:45:30 | warlord | True, but there isn't a good way to change it in buildroot-tux |
| 16:45:31 | warlord | :) |
| 16:45:35 | warlord | Ok, patch sent. |
| 16:53:07 | CosmicPenguin | warlord: Its in |
| 16:53:35 | warlord | Cool. :) |
| 16:54:58 | warlord | Now to figure out why I didn't receive a copy of my patch ;) |
| 16:56:29 | warlord | (although I did receive other mail) |
| 16:59:03 | warlord | (ah, just took time -- must learn patience ) |
| 17:39:34 | warlord | hi erik |
| 17:40:05 | andersee | howdy |
| 17:40:52 | andersee | making progress? |
| 17:41:20 | CosmicPenguin | andersee: on? |
| 17:41:55 | andersee | the phone... |
| 17:41:58 | andersee | :) |
| 17:42:27 | CosmicPenguin | andersee: I finally gotr my buildroot stuff checked in, does that count? |
| 17:42:51 | andersee | thats cool |
| 17:43:08 | warlord | I've submitted a bunch of patches.. ;) |
| 17:43:08 | andersee | counts as progress in my book |
| 17:43:36 | warlord | but still nothing on why some work and some do not |
| 17:43:44 | CosmicPenguin | andersee: and I'm happy to discover that you have __UCLIBC__ defined, so I can use that to shore up any holes in my code.. :) |
| 17:43:51 | GoRK | CosmicPenguin: ahh no wonder your brightness/contrast menu always just made my screen go away w/ my kernel :) |
| 17:44:38 | CosmicPenguin | GoRK: Yeah, that made me mad too when I tried it yesterday.... I forgot that I had made that change.... :) |
| 17:44:47 | andersee | CosmicPenguin: geez, that andersee guy. he thought off everything... :) |
| 17:45:06 | CosmicPenguin | andersee: yep, he's a smart guy..... |
| 17:45:16 | warlord | who? where? |
| 17:45:29 | andersee | Someone should pay him lots of money to just sit at home and think deep thoughts. :) |
| 17:45:40 | CosmicPenguin | andersee: thats a lot better than #ifdef ARGH_STUPID_UCLIBC_WORKAROUND |
| 17:46:04 | andersee | What're you having to workaround? |
| 17:46:14 | GoRK | CosmicPenguin: i jtag'd my broken phone last night only to find that the touchscreen works. blah |
| 17:46:16 | andersee | wide char? |
| 17:46:24 | CosmicPenguin | you don't have <values.h> |
| 17:46:29 | CosmicPenguin | GoRK: Thats good! |
| 17:46:57 | GoRK | CosmicPenguin: it's bad because i dont have anything to debug |
| 17:47:08 | CosmicPenguin | GoRK: Ahh..... |
| 17:47:23 | CosmicPenguin | GoRK: I have that same problem.... but sometimes I can grab signal11's phone.... |
| 17:47:25 | warlord | oh, right, the values.h bug i reported :) |
| 17:48:01 | CosmicPenguin | warlord: Its actually the code that is stupid, but I figured if <values.h> *is* there, then use it... |
| 17:48:14 | andersee | CosmicPenguin: do now... :) |
| 17:48:19 | CosmicPenguin | so a quick #ifdef __UCLIBC__ #define MAXINT 0x7FFFFFFF #endif fixed it.... |
| 17:48:39 | andersee | CosmicPenguin: 'cd uClibc && cvs up' |
| 17:48:55 | andersee | Then you can drop the workaround.... |
| 17:50:04 | warlord | ahh, you supply values.h now? |
| 17:50:08 | andersee | I've got a couple bugs I need to fix for the government, then I'm planning on fixing uClibc sig handling and hopefully I'll have time to add large file support |
| 17:50:09 | CosmicPenguin | andersee: Why is lineo so untrusting? a blank password isn't good enough for them? |
| 17:50:18 | andersee | warlord: as of about 1 minute ago I do. |
| 17:50:30 | warlord | LOL |
| 17:50:38 | warlord | I love this real-time bug-fixing :) |
| 17:50:38 | andersee | CosmicPenguin: why is the sky green? |
| 17:50:55 | CosmicPenguin | Because you are in Orem near Utah Lake.... :) |
| 17:51:29 | warlord | FTR, A friend of mine seems to think that the S 9804 and S 9806 identifiers on the chips are probably date-stamps, and doubts that the chips themselves are the indicator of "workingness" |
| 17:51:30 | andersee | But I'm not in orem. I'm in Salt Lake (like you) |
| 17:51:42 | andersee | So the sky is only a murky grey |
| 17:51:55 | CosmicPenguin | andersee: then thats kennecott instead...... |
| 17:51:59 | warlord | Hey, I'll be in Salt Lake in December! |
| 17:52:22 | andersee | kennecottt, which I lovingly refer to as "The Blight" |
| 17:52:28 | andersee | warlord: oh, cool. |
| 17:52:28 | warlord | um |
| 17:52:38 | warlord | Yea.. IETF is there. |
| 17:52:43 | warlord | (www.ietf.org) |
| 17:53:02 | CosmicPenguin | ietf is here? lets picket! |
| 17:53:22 | warlord | what do you plan to picket about the ietf? |
| 17:53:33 | andersee | patenting of protocols? |
| 17:53:42 | andersee | I thought that was the w3c |
| 17:53:52 | warlord | that's the w3c |
| 17:54:02 | warlord | IETF specifically forbids patented technology |
| 17:54:09 | CosmicPenguin | Maybe it will be like the Simpsons - You ever see that one where people were picketing the white house with signs that said "Nothings Wrong" and "You're doing a great job"???? |
| 17:54:12 | warlord | (at least in required-to-implement protocols) |
| 17:54:30 | warlord | LOL -- nope, didn't see that |
| 17:56:29 | warlord | Anyways, at this point the buildroot seems to work without errors... But I haven't tested it, yet. |
| 17:57:22 | GoRK | i have to integrate the apps tonight then i'll have buildroot for picogui also |
| 17:57:40 | warlord | so what is picogui (as opposed to microwindows)? |
| 17:57:45 | GoRK | so if anyone wants to test that, get ready :) |
| 17:58:06 | warlord | the web site isn't very helpful at a comparrison |
| 17:58:10 | GoRK | it's another GUI. picogui has built in window manager and widgets. that's the biggest difference |
| 17:59:50 | GoRK | consequently apps are super small.. the terminal app is <6KB for instance |
| 18:01:20 | GoRK | but you lose a lot of the flexibility that you get with nano-x |
| 18:01:49 | GoRK | although if you were die-hard you could paint your own widgets onto a canvas |
| 18:02:33 | prpplague | GoRK: you done anything with tiny-x ? |
| 18:03:03 | GoRK | no |
| 18:03:12 | GoRK | i'd like to make it run just to see it |
| 18:03:31 | GoRK | but i havent had time |
| 18:04:01 | prpplague | i like using fltk, i'm just wondering how easy it will work with nono-x |
| 18:04:12 | warlord | yea, i do guess that the microwindows apps are fairly large.. |
| 18:04:34 | CosmicPenguin | warlord: Fairly large? |
| 18:04:55 | CosmicPenguin | prpplague: It works great, execpt on the Tuxscreen (damn C++ isuses) |
| 18:05:46 | andersee | CosmicPenguin: too bad Jason never offered to pay me to finish that... |
| 18:06:14 | prpplague | CosmicPenguin: cool |
| 18:06:18 | warlord | Well, my ts seemed to crash when I ran a couple copies of, say, snake |
| 18:06:34 | CosmicPenguin | andersee: You know what my recommentations were... |
| 18:06:55 | prpplague | andersee: what lacks being done? |
| 18:07:05 | CosmicPenguin | warlord: Are you sure that wasn't the window manager? Snake is only 41K |
| 18:07:49 | warlord | No, I'm not sure... But if I try to run, say, message, snake, and then the web viewer, the machine would crash |
| 18:09:09 | warlord | I mean, viewml _is_ huge.. |
| 18:09:31 | CosmicPenguin | warlord: I don't see that -- meminfo - 13944KB, 5288KB free.... |
| 18:09:48 | andersee | prpplague: well, someone needs to spend some time making global constructors and destructors work properly on ARM. Then (optionally) make libstdc++ work... |
| 18:09:53 | CosmicPenguin | warlord: But yes, viewml is _huge_ but that is because it has to be static... (right, andersee????) |
| 18:10:06 | andersee | CosmicPenguin: yup |
| 18:10:20 | warlord | why does it have to be static? |
| 18:10:29 | andersee | CosmicPenguin: becuase viewml uses c++, threads, and libstdc++ |
| 18:10:29 | CosmicPenguin | C++ does not work with glibc |
| 18:10:40 | andersee | s/glibc/uclibc/ |
| 18:10:40 | CosmicPenguin | andersee: I could kill off the threads |
| 18:10:58 | andersee | And libstdc++ to I expect |
| 18:11:08 | CosmicPenguin | andersee: probably.... |
| 18:11:12 | andersee | Probably just some debug stuff, right? |
| 18:11:33 | warlord | and including those libs would be larger than linking statically? |
| 18:11:34 | CosmicPenguin | andersee: Yeah, I would expect.... |
| 18:11:38 | CosmicPenguin | warlord: Yep |
| 18:11:42 | warlord | ahh |
| 18:11:44 | prpplague | andersee: what kinda time+$$ |
| 18:12:23 | warlord | Back to the touchscreen problem -- this is where I wish Linux had a kernel debugger :) |
| 18:12:43 | CosmicPenguin | warlord: if C++ worked with uclibc, I'll bet I could cut by 2/3 |
| 18:13:13 | GoRK | hey how do i get the viewml to pull up a webpage at all? my network is working on the phone with the microwindows image.. i type in the url and press enter and the window turns grey and the network lights flash but nothing happens |
| 18:13:13 | warlord | nods |
| 18:13:13 | andersee | CosmicPenguin: In theory, the global constructor/destructor stuff is working on ARM already (as of a week ago)... But I havn't tested the change. |
| 18:13:19 | warlord | What's keeping uclibc from working w/ c++? |
| 18:14:01 | warlord | (is there a list of bugs in these packages that I can lookup) |
| 18:14:05 | andersee | I also have a patch I need to cleanup and add which provides a uclibc g++ wrapper |
| 18:16:19 | warlord | Does viewml have SSL support? |
| 18:17:28 | CosmicPenguin | warlord: sadly it doesn |
| 18:18:28 | warlord | nods |
| 18:19:04 | warlord | Maybe I can look at that. It might be useful (instead of writing my own UI) |
| 18:19:58 | CosmicPenguin | warlord: There is actually a long list of stuff to add / modify, especially if we are going to stick t in such a small space... |
| 18:20:16 | andersee | prpplague: well, right now I've got all the work I can handle, so I'll probably need to couch someone through the dirty work of getting c++ going |
| 18:21:04 | warlord | Where is this list, CP? |
| 18:23:17 | CosmicPenguin | warlord: First step is to get rid of the dependency of threads..... |
| 18:23:43 | CosmicPenguin | next, we need to think about replacing libwww with something a little smaller... |
| 18:24:07 | warlord | Ah, you're talking specifically about the list of viewml? |
| 18:24:19 | warlord | I was hoping there was a general "TuxScreen ToDo List" |
| 18:24:34 | CosmicPenguin | SSL warlord: i've got one of those too (at least for Microwindows) |
| 18:24:41 | CosmicPenguin | s/SSL// |
| 18:25:30 | CosmicPenguin | andersee: ARGH! |
| 18:25:48 | andersee | warlord: there is a TODO list in buildroot-tux |
| 18:25:51 | CosmicPenguin | andersee: The current buildroot has the newest uClibc which has the broken signal support, huh? |
| 18:26:17 | andersee | umm. yeah, I suppose so. |
| 18:26:37 | prpplague | CosmicPenguin: whats that about broken sig support? |
| 18:27:04 | warlord | What is this about broken signals? |
| 18:27:05 | CosmicPenguin | prpplague: andersee can probably explain it better... |
| 18:27:48 | prpplague | i pester andersee enough already.... |
| 18:28:24 | warlord | Yea, me too :) Right, Erik? ;) |
| 18:28:30 | andersee | Since I reworked the header files, signals are now one-shots. |
| 18:28:30 | CosmicPenguin | prpplague: Anyway, if you install a signal handler, it only works for the first instance of that signal, and then it disappears.... |
| 18:28:41 | CosmicPenguin | Ok, so there you go..... |
| 18:28:45 | andersee | I'm hoping to get a few minutes to fix that today |
| 18:29:00 | CosmicPenguin | andersee: should I put a work around in the code, or will it be fixed soon? |
| 18:29:02 | warlord | iirc, sigaction() provides one-shot handlers |
| 18:29:45 | prpplague | don't suppose that would be related to my SIGILL problems? |
| 18:30:08 | CosmicPenguin | prpplague: Doubtful, this only pertains to signal handlers..... |
| 18:30:25 | andersee | I need to go fix some bugs in the government project I'm doing... |
| 18:30:31 | CosmicPenguin | warlord: BTW - thats probably why you thought it crashed.... |
| 18:30:34 | andersee | I'll be back later |
| 18:30:39 | CosmicPenguin | Bye! |
| 18:30:46 | warlord | CP: yea, losing signals would certainly do it. |
| 18:30:48 | warlord | See ya, Erik |
| 18:30:48 | andersee | Hopefully with time to fix sighandling... |
| 18:31:02 | CosmicPenguin | andersee: Ok, I will put a work around in for now.... |
| 18:31:03 | prpplague | later |
| 18:58:31 | prpplague | under arm linux when the arm board is put to sleep and the power button is pressed hows that handled? |
| 19:04:41 | warlord | power button? |
| 19:08:01 | prpplague | warlord: when you use the linux apm to put the system to sleep how is the wake procedure handled? is the an arm apm doc? |
| 19:09:05 | warlord | I've never actually tried putting it to sleep that way, so i dont know. |
| 19:09:32 | CosmicPenguin | prpplague: Did you ask in the #handhelds or #ipaq room? They are the folks who have done most of the work in that area |
| 19:10:17 | prpplague | CosmicPenguin: no, i typed before i thought, i'll ask there... |
| 19:16:02 | warlord | Hmm... I wonder why /proc/devices/ucb1200/register locks up on machine but works on the other.. |
| 19:17:35 | CosmicPenguin | warlord: same problem as all the other issues |
| 19:19:02 | warlord | So it's a related problem, then? |
| 19:22:40 | CosmicPenguin | Its *THE* problem |
| 19:22:40 | warlord | Mmm. |
| 19:22:40 | warlord | I see. |
| 19:22:40 | warlord | Well, let me see if my own-built kernel works.. Then I'll see if I can look into it more. |
| 19:25:40 | warlord | First let's see if it works on my working ts. Then I'll know that I've successfully built the kernel. (The blob cmp's the same, so I have no doubts about that :) |
| 19:27:07 | warlord | Ok, kernel works. Now to move to non-working ts :) |
| 19:27:25 | prpplague | CosmicPenguin: i really dislike the handhelds.org channel |
| 19:27:40 | CosmicPenguin | If I take the Win32 API out of microwindows, I can save soo much room... do you guys suppose anyone will want to use it? |
| 19:27:42 | CosmicPenguin | prpplague: why? |
| 19:27:57 | AlHazWork | prpplague: is it one of those "if you have to ask you're not cool enough to know" channels? |
| 19:28:11 | prpplague | AlHazWork: seems that way to me |
| 19:28:21 | AlHazWork | prpplague: ever try #unix on efnet? |
| 19:28:54 | CosmicPenguin | prpplague: Some of those guys hang out in here too... I wonder why they would be so evil.... |
| 19:29:35 | prpplague | CosmicPenguin: i can't explain the attitude in here is much different |
| 19:30:15 | CosmicPenguin | prpplague: Maybe its because we are a little group of people struggling together with a limited product |
| 19:30:31 | CosmicPenguin | while every yokel with a credit card has an Ipaq.... |
| 19:30:57 | warlord | CosmicPenguin: make it a compile/configure option? |
| 19:31:11 | warlord | Some people might want it. |
| 19:31:25 | prpplague | CosmicPenguin: maybe so |
| 19:31:46 | CosmicPenguin | warlord: its definately there, I just want to pull it out of the default build.... |
| 19:31:49 | warlord | Hey, *I* don't have an IPaq. ;) |
| 19:31:58 | CosmicPenguin | warlord: are you a yokel? |
| 19:32:08 | warlord | I have no problems with losing Win32 :) |
| 19:32:16 | warlord | CP: Not last I checked ;) |
| 19:33:06 | CosmicPenguin | warlord: Yeah, I'm talking about those guys who want to "try that LInyucks thingy" |
| 19:34:53 | prpplague | lol |
| 19:35:18 | CosmicPenguin | warlord: saved 65K! |
| 19:36:07 | warlord | Heh |
| 19:36:16 | warlord | WOW! 65K! *snickers* |
| 19:36:38 | CosmicPenguin | warlord: you making fun? |
| 19:36:40 | warlord | hey, any idea why I'm getting this error during bootup: |
| 19:37:03 | warlord | fatfs: bogus logical sector size 34276 |
| 19:37:21 | warlord | CP: Yes. I just don't see 65K as a huge savings. |
| 19:37:41 | CosmicPenguin | warlord: fatfs is build into the kernel, so it trys to see if the root fs is FAT |
| 19:38:08 | CosmicPenguin | warlord: 65K is another GUI application or two.... |
| 19:38:12 | prpplague | warlord: i got that when my initrd image was damaged and the kernel tried to see it as a fat fs(on other arm systems) |
| 19:38:38 | prpplague | warlord: did you store the initrd image on a fatfs? |
| 19:39:43 | prpplague | CosmicPenguin: when you put the tuxscreen to sleep, is the kernel still running in the background or does the system just record the current status? |
| 19:39:47 | AlHazWork | I get the fatfs: warning every time i boot without exceptin |
| 19:39:49 | AlHazWork | exception |
| 19:40:15 | prpplague | AlHazWork: thats odd |
| 19:40:36 | CosmicPenguin | prpplague: Hmm.... I don't know if you can put the tuxscreen to sleep or not.... will it ever wake up? |
| 19:40:59 | CosmicPenguin | prpplague: its just a warning.... if you want to get rid of it, compile out fatfs support....... |
| 19:41:11 | CosmicPenguin | Of course, that would limit your ability to read CFs |
| 19:41:29 | prpplague | BZFlag: know much about arm-linux apm? |
| 19:41:52 | prpplague | CosmicPenguin: i don't get a warning |
| 19:41:59 | CosmicPenguin | prpplague: but when you put the Ipaq to sleep, the system saves the state and turns everything off.... |
| 19:42:04 | CosmicPenguin | prpplague: really/ |
| 19:42:11 | CosmicPenguin | s/\//?/ |
| 19:43:44 | prpplague | CosmicPenguin: hmm, so when the ipaq starts back up does bootldr run again? |
| 19:43:52 | CosmicPenguin | prpplague: no |
| 19:44:09 | warlord | nope -- using jffs2 |
| 19:44:40 | AlHazWork | I thought I'd heard that the boot loader on arm systems had to be semi-involved in waking up |
| 19:45:06 | prpplague | AlHazWork: thats what i was getting at, i didn't see any wake procedures in blob |
| 19:45:38 | CosmicPenguin | AlHazWork: I think that was the old style... hold on the guy in the cube next to me has been working on this... |
| 19:46:18 | AlHazWork | I know a guy who's been working on arm power management a lot, but he's not in today, and I've stuck his neck out a few too many times on the subject |
| 19:46:20 | CosmicPenguin | Yeah, it seems to all be contained within the kernel now.... he even has alarms that wake up the system. He hits the button and an minute later it wakes back up |
| 19:46:43 | CosmicPenguin | with a very annoying alarm wav, I might add.... |
| 19:47:04 | prpplague | so the kernel intercepts the wakeup interupt? |
| 19:47:49 | CosmicPenguin | prpplague: yes |
| 19:48:05 | prpplague | CosmicPenguin: ok thanks, tell your buddy thanks too |
| 19:49:00 | prpplague | i want everyone here to know, if you guys are ever in east texas, look me up, i'll spring for a steak-n-beer |
| 19:50:20 | GoRK | prpplague: where are you in tx? im in amarillo |
| 19:51:20 | prpplague | tyler |
| 19:52:00 | GoRK | cool. my packets go through there all the time :) |
| 19:54:13 | RedBaron | BZFlag: I got that power supply today. When I get home, I'll try it out and let you know if the tuxscreen works. |
| 19:56:51 | prpplague | GoRK: nice amarillo |
| 19:59:30 | CosmicPenguin | AlHaz: Did the pcnet change to pcmcia config ever make it into buildroot? |
| 19:59:32 | warlord | GRF.. "make kernel" does nothing if linux/arch/arm/boot/zImage exists |
| 20:00:22 | warlord | That was annoying.. |
| 20:00:55 | CosmicPenguin | Ooops..... |
| 20:00:58 | CosmicPenguin | It didn't.... |
| 20:02:05 | CosmicPenguin | Wait a second - bind tulip???? |
| 20:02:51 | AlHazWork | tulip is a pci chip . . . |
| 20:03:15 | CosmicPenguin | AlHaz: Yeah, different card name... |
| 20:03:24 | warlord | So why is orrinco built into the kernel? |
| 20:03:35 | AlHazWork | warlord: Good question, especially as it doesn't work |
| 20:04:22 | CosmicPenguin | warlord: the kernel is a little like RedHat's -- everything is built in. Thats good for beginners, but more advanced folks would probably end up rebuilding anyway |
| 20:05:30 | warlord | Um, buildroot certainly does not build everything by default. |
| 20:05:40 | warlord | Besides, one would think you would want it as a module, not compiled into the kernel. |
| 20:06:00 | warlord | and for the record, the ucb1200 lock/unlock does not cause the lockup |
| 20:06:08 | AlHazWork | there are a bunch of modules built into the kernel because we've got almost a meg of flash to store the kernel |
| 20:06:32 | CosmicPenguin | warlord: What do you mean? |
| 20:06:40 | AlHazWork | so stuff that might be useful was added until it was filled |
| 20:06:55 | AlHazWork | if we update to a blob that can read jffs2, we'll stop doing that entirely |
| 20:07:18 | warlord | I'm playing with the kernel trying to track down what's failing with the ucb1200.. |
| 20:07:25 | warlord | reading from io and registers fails. |
| 20:07:29 | CosmicPenguin | warlord: right... |
| 20:07:43 | warlord | Well, the register read does a lock/read/unlock process. |
| 20:07:54 | warlord | I verified that the lock/unlock isn't the culprit. |
| 20:08:42 | CosmicPenguin | Right.... the culprit is line 78 in ucb1200_sa100.c |
| 20:09:15 | warlord | You mean line 77 (the while loop)? |
| 20:09:30 | CosmicPenguin | And the CWC bit on line 86 is broken too we think, but it is broken to our favor, so it works.... |
| 20:09:40 | CosmicPenguin | warlord: yep... |
| 20:11:41 | warlord | Hmm. Interesting.. |
| 20:12:09 | warlord | Where is the definition of Ser4MCDR2? |
| 20:12:34 | CosmicPenguin | thats the MCP data register 2... its in the SA1100 documentation.... |
| 20:13:37 | CosmicPenguin | http://tuxscreen.net/download/sa1100/278088.pdf.gz |
| 20:14:38 | warlord | hm. |
| 20:17:18 | warlord | Do we have any docs for the Digital StrongARM? |
| 20:17:54 | prpplague | CosmicPenguin: will you ask your cube-mate if he/she can recomend some reading material on the arm apm? |
| 20:19:59 | CosmicPenguin | prpplague: he said "yeah, the kernel source" |
| 20:20:42 | CosmicPenguin | or look for pm_helper on handhelds archive |
| 20:24:39 | BZFlag | heh, yes apm wake up starts at 0x0 just like reset. there is a register set to indicate a resume. |
| 20:24:53 | BZFlag | the bootloader needs to see this and then "do the right thing" |
| 20:25:18 | BZFlag | bootldr (crl) has support, afaik blob does not yet. |
| 20:25:32 | BZFlag | though we can take it from bootldr ;-) |
| 20:27:13 | BZFlag | under bootldr, the register holds a resume vector. I don't recall the exact symantics, it might be a code vector or it might be a data area with further saved state info as well as a resume jump point. |
| 20:27:47 | BZFlag | the bootloader should do as little as possible resume code, and jump to the OS entry point for the rest. |
| 20:28:06 | GoRK | BZFlag: i was fooling around with jtag and andersee said that all the other programs (ie other than "flash") do not compile cleanly.. are there precompiled binaries that do? |
| 20:28:28 | GoRK | BZF; the makefile still creates all the links and things |
| 20:28:38 | BZFlag | there have been changes in who does what under linux-arm. ie: I think the bootloader used to setup the serial port, but it did not have knowledge as to what state it should be in, the kernel did |
| 20:28:46 | BZFlag | so the kernel does that piece now iirc. |
| 20:29:34 | BZFlag | GoRK: jflash is cvs wants to see tucscreen cvs blob to compile against, and perhaps some now outdated kernel source too. |
| 20:29:48 | BZFlag | I should clean that up at some point. |
| 20:30:05 | prpplague | BZFlag: thanks for the info |
| 20:30:20 | BZFlag | You might try Russ' mtd over jtag driver. it's cleaner, though I have not rebuilt that in a while either. |
| 20:30:27 | prpplague | CosmicPenguin: thanks as well |
| 20:30:47 | BZFlag | prpplague: sure. ;-) |
| 20:31:22 | GoRK | BZFlag: actually all i wanted to do is play with 'splash' and the gpio watcher.. is the jflash in the .tgz on downloads/ compiled with support for all that stuff? |
| 20:31:24 | CosmicPenguin | prpplague: not that I was much help, but np... :) |
| 20:31:47 | prpplague | you know one of things i really dislike about the m$ culture is the way it creates a knowledge cast system, the linux community is totaly different |
| 20:33:59 | AlHazWork | prpplague: except for the im-smarter-than-you-nyah-nyah bunch |
| 20:35:05 | AlHazWork | also known as the you-dont-code-so-you-dont-matter bunch |
| 20:35:52 | scanline | There's also the i-done-code-but-I'm-smarter-than-you-'cuz people that are often found on slashdot |
| 20:36:05 | scanline | s/done/dont/ |
| 20:37:28 | CosmicPenguin | scanline: like the 'cuz-you-don't-use-<insert distribution here>' people? |
| 20:38:04 | scanline | yep |
| 20:40:59 | BZFlag | GoRK: I think so, yes. |
| 20:41:33 | BZFlag | GoRK: Russ has blob splash patches now too. That's probably the better route to take. |
| 20:41:57 | GoRK | BZFlag: sweet. Yes that is nicer :) |
| 20:42:53 | GoRK | a little of that, a little LPP and we have a cheeful boot |
| 20:45:27 | warlord | I wonder why I'm getting an rtc 1Hz interrupt on my non-working but not on my working ts? |
| 20:46:10 | warlord | I also appear to not be receiving GPIO interrupts. |
| 20:46:50 | CosmicPenguin | warlord: hmm... that is interesting indeeed..... |
| 20:47:46 | warlord | smiles |
| 20:48:04 | warlord | It was occurring to me that if we're missing an interrupt, it could cause this sort of lockup. |
| 20:48:24 | CosmicPenguin | no, I doubt thats the thing, since the UCB communications happen over the SIB |
| 20:49:40 | CosmicPenguin | How are you finding this out? Do you have a scope hooked up, or is this just in code? |
| 20:50:00 | warlord | True, but if the kernel never gets interrupted that the data has changed. For example, I know we don't look at IRQs 18 and 19 (the Serial4 IRQs) |
| 20:50:23 | CosmicPenguin | warlord: I would expect thats because we poll them.... |
| 20:50:51 | CosmicPenguin | or rather, poll the Serial4 ports... |
| 20:51:53 | warlord | hm |
| 20:53:26 | CosmicPenguin | Note this: The SIBDOUT pin of the UCB1200 is tri-stated for the bit 64 and higher in the SIB frame to prevent bus conflicts with other slave devices. However when SIB_ZERO is set the SIBDOUT pin is forced to zero from bit 64 onwards to prevent the SIBDOUT line from floating. THis feature is needed when the UCB1200 is the only slave device connected to the bus |
| 20:56:29 | BZFlag | GoRK: I think the blob splash code should get added to blob cvs, but not enabled unless you do a --enable-splash on the configure line. |
| 20:57:45 | warlord | But swapping the UCB1200s did nothing last night.. It was the mainboard, not the i/o board, that affected workingness.. |
| 20:58:02 | CosmicPenguin | warlord: but the i./o board is connected by the SIB... |
| 20:58:03 | warlord | Unless there is a bad resistor on some boards |
| 20:58:06 | CosmicPenguin | and the SIBOUT pin is on the processor |
| 20:58:18 | CosmicPenguin | s/pin/bit/ |
| 20:59:37 | warlord | So what are you saying, then.. I don't get where you're going (I knew my EE degree would be useless ;) |
| 21:00:44 | CosmicPenguin | warlord: So what if the sibdout pin *is* floating? |
| 21:01:05 | warlord | Why would it be floating on some main boards and not on others? |
| 21:01:27 | warlord | wishes he had a logic analyzer to test the pin :) |
| 21:01:31 | CosmicPenguin | warlord: I guess it would depend on *if* there are other slaves on the bus, and *where* they are.... |
| 21:02:04 | warlord | Do we have TuxScreen schematics somewhere? |
| 21:03:28 | warlord | Or someone with a logic analyzer and a non-working TS? |
| 21:03:43 | CosmicPenguin | warlord: My whole point is this: There is only one place where the processor can talk to the UCB1200 chip, thats the MCP port - So we have to assume that our problems lie there.... |
| 21:03:44 | warlord | Also, if it were bad hardware, wouldn't Inferno not work? |
| 21:04:14 | CosmicPenguin | So we only have a limited number of registers that we can tweak, and it just happens that the SIBOUT bit is one of them.... |
| 21:05:38 | warlord | Do we set SIB_ZERO? |
| 21:06:35 | CosmicPenguin | Not unless the UCB1200_ADC_SYNC flag is set.... |
| 21:06:51 | CosmicPenguin | ucb1200_sa1100.c:203 |
| 21:08:27 | warlord | WHich I dont think it is in our configuration... |
| 21:08:39 | CosmicPenguin | nope |
| 21:08:58 | warlord | So what happens if we set that flag? |
| 21:09:52 | CosmicPenguin | ? |
| 21:11:05 | warlord | Well, what if we set UCB1200_ADC_SYNC to force SIB_ZERO to be set? The only other ramification I can see is that start_adc() will set 'reg' differently in ucb1200_generic |
| 21:12:42 | warlord | I can try to add arch->flags |= UCB1200_ADC_SYNC in the SHANNON config and see what happens... |
| 21:13:13 | CosmicPenguin | warlord: did you get the SA1100 spec? |
| 21:14:46 | warlord | Yes, but it's absolutely huge :) |
| 21:15:58 | SpaceCoaster | warlord: SpaceCoaster fades in...chapter11...fades out |
| 21:17:07 | CosmicPenguin | SpaceCoaster: you are the reigning master of this after yesterday, you should be involved with this discussion |
| 21:17:13 | warlord | Funny that -- I went straight to chapter 11... But I feel like hacking, not reading ;) |
| 21:17:44 | CosmicPenguin | "You must first fill your head with knowlege before you can break blocks of ice with it" |
| 21:18:40 | warlord | Yea, yea.. patience.. How long's THAT gonna take? |
| 21:20:17 | warlord | I feel like I'm learning Tai Kwon Leap |
| 21:20:21 | warlord | :) |
| 21:20:40 | warlord | (and FTR, setting ADC_SYNC didn't solve the problem ;) |
| 21:20:49 | CosmicPenguin | damn |
| 21:20:59 | CosmicPenguin | howsabout just setting the SIBOUT without the SYNC |
| 21:21:40 | CosmicPenguin | also see function sa1100_mcp_enable? |
| 21:21:51 | GoRK | did anyone think just to press really *hard* on the touchscreen? maybe that's all you really need to do!!!!!!!!!!!!!!!! |
| 21:22:15 | CosmicPenguin | See where we turn on bit MCCR0_MCE , we do our thing, and then turn it off again/ |
| 21:22:31 | CosmicPenguin | What if we were to leave it on all the time? |
| 21:23:03 | CosmicPenguin | Or, we need to at least make sure that the registers are set up correctly..... |
| 21:23:54 | CosmicPenguin | and I quote: Clearing MCE resets the MCP's FIFOs. However, MCP data register 3, the control and the status registers are not reset. The user must ensure these registers are properly reconfigured before reenabling the MCP |
| 21:25:33 | warlord | Well, right now I'm testing the setting of the sib_zero without all the adc_sync stuff (basically, I just added "|| 1" on line 203 |
| 21:27:19 | warlord | But that didn't help. |
| 21:28:00 | warlord | I'm not sure how we could leave the MCCR0_MCE bit on all the time.. |
| 21:28:27 | CosmicPenguin | WHy not? |
| 21:28:27 | warlord | I mean, you want to leave sib_enabled all the time? |
| 21:28:37 | CosmicPenguin | sure |
| 21:28:49 | warlord | Um, ok.. I can disable sib_disable :) |
| 21:28:56 | CosmicPenguin | we don't watch the interrupts anyway |
| 21:29:12 | CosmicPenguin | Just make sure that bit stays on |
| 21:29:48 | warlord | Um, sure. |
| 21:32:55 | warlord | Updating kernels via blob is just so cool... So much easier/better than inferno |
| 21:34:04 | warlord | Nope. Setting SIB_ZERO and also disabling sib_disable didn't do anything :( |
| 21:34:28 | CosmicPenguin | how are you testing that? |
| 21:34:40 | warlord | cat /prot/drivers/ucb1200/io |
| 21:34:51 | warlord | er s/prot/proc/ |
| 21:34:51 | CosmicPenguin | ok |
| 21:34:58 | prpplague | warlord: i know you guys must be frustrated, i am just watching... |
| 21:35:05 | warlord | thats ok, prpplague |
| 21:35:18 | warlord | i'm actually not frustrated.. Just.. confused. |
| 21:35:43 | warlord | Let me try disabling sib_disable without actually setting sib_zero |
| 21:35:49 | CosmicPenguin | man, a logic analyzer would be worth alot right now |
| 21:35:59 | CosmicPenguin | I know of half a dozen pins I would like to know the state of |
| 21:36:42 | warlord | *chuckles* |
| 21:36:56 | warlord | your kingdom for a logic analyzer? |
| 21:37:35 | prpplague | where's a vulcan when you need one? |
| 21:37:58 | warlord | slaps prpplague around a bit with a large trout |
| 21:38:54 | prpplague | thanks i feel better now |
| 21:40:03 | warlord | anytime |
| 21:40:11 | warlord | any other suggestions, CP? |
| 21:40:30 | CosmicPenguin | not right now.... |
| 21:40:32 | AlHazWork | we have a few logic analyzers at lineo . . . . |
| 21:40:38 | AlHazWork | big agilent logic analyzers |
| 21:41:14 | CosmicPenguin | AlHaz: We had some of those babies at Cabletron.... |
| 21:41:15 | warlord | Are you sure that that code is actually setting SIB_ZERO properly? |
| 21:41:34 | CosmicPenguin | warlord: *shrugs* |
| 21:41:49 | CosmicPenguin | Can you rent 'em? |
| 21:42:56 | CosmicPenguin | http://www.harlanlabs.com/datacom.html |
| 21:43:41 | CosmicPenguin | http://www.atecorp.com/flyer2.htm |
| 21:44:12 | AlHazWork | yeah, you can |
| 21:45:42 | warlord | What is bit 0x200? |
| 21:49:08 | CosmicPenguin | Doesn't way |
| 21:49:12 | CosmicPenguin | s/way/say/ |
| 21:49:27 | warlord | Hrm. |
| 21:49:47 | CosmicPenguin | bit 15 is the SIB_ZERO |
| 21:50:37 | CosmicPenguin | Wait, I'm an idiot.... |
| 21:50:53 | CosmicPenguin | That sets the direction of the 9th GPIO pin to as output.... |
| 21:51:01 | CosmicPenguin | And then it sends data to that pin..... |
| 21:54:01 | warlord | I'm mostly trying to figure out what the 0 and 1 are in the setting of sib_zero |
| 21:54:08 | warlord | (and the 0x200) |
| 21:54:11 | CosmicPenguin | line? |
| 21:54:54 | warlord | ucb1200_sa1100.c:207 |
| 21:55:19 | CosmicPenguin | the 0 and the 1 are the UCB1200 registers |
| 21:55:48 | CosmicPenguin | so it sends a 0x200 to register 1, which sets the direction of the I/O to output |
| 21:56:15 | warlord | Actually it sends SIB_ZERO | 0x200 to register 1 |
| 21:56:18 | CosmicPenguin | and then it sends a 0x200 to register 0 which sets the I/O port high.... |
| 21:56:27 | warlord | The 0x200 to register 0 |
| 21:56:50 | CosmicPenguin | Ok... |
| 21:57:06 | warlord | Which looks wrong... |
| 21:57:14 | warlord | It's setting SIB_ZERo on the direction register, |
| 21:57:17 | warlord | but not the data register? |
| 21:57:37 | CosmicPenguin | The SIB_ZERO is just a control bit of sorts.... |
| 21:57:55 | CosmicPenguin | You have the UCB1200 sheet? |
| 21:58:00 | warlord | Nope |
| 21:58:06 | warlord | Just the sa1100 |
| 22:20:06 | BZFlag | http://TuxSreen.net/wiki/view//TouchScreen <- ucb1200 data sheet link. |
| 22:59:12 | CosmicPenguin | GoRK: are you there? |
| 23:10:16 | RedBaron | Sweet! My tuxscreen finaly works. :) The power supply that came with it was bogus. never worked. the the replacement that BZFlag sent works. |
| 23:13:24 | GoRK | CosmicPenguin: yeah |
| 23:17:48 | BZFlag | RedBaron: glad to hear it. I've got a few extra PS here, so so whatever you like with the non-functional one. If you're not going to do anything with it and want to slow boat it to me, I might hack it sometime. |
| 23:22:01 | RedBaron | BZFlag: Okay. Great. I'll probably play around with it some. but if I end up not doing anything, I'll send it back. |
| 23:24:17 | CosmicPenguin | If a company says that their chips are not designed for use in life support systems, are they essentially admiting that their chips will crash at some point? :-) |
| 23:25:39 | GoRK | CosmicPenguin: i'm here was there something you were needing |
| 23:26:32 | CosmicPenguin | GoRK: Sorry no...... I want thinking that you were the one who worked on GnoPhone, but it was kram instead.... |
| 23:26:54 | RedBaron | CosmicPenguin: I used to work in the defense industry. Lots of software products had disclaimers saying that their stuff was not for use in the defense industry. I suppose when peoples lives were dependent the the software not crashing... |
| 23:28:10 | CosmicPenguin | RedBaron: Yeah, its just legal stuff... I was trying to be a little humourous - especially since that disclamier came off of the UCB1200 spec.... |
| 23:29:13 | RedBaron | CosmicPenguin: *nodder* Well...I'm glad they put that disclaimer on their...because they were right. Heh. I worked with various CORBA ORB products...and they worked fine for casual use. But nothing serious like we needed. |
| 23:29:20 | GoRK | CosmicPenguin: no but i'm planning to build up an IAX phone for picogui |
| 23:29:34 | CosmicPenguin | GoRK: of course.... :) |
| 23:30:36 | GoRK | and i hack on asterisk a bit |
| 23:38:36 | GoRK | laer oof= |
| 23:38:42 | GoRK | later |
| 23:56:00 | RedBaron | Hmm...Does anyone elses tuxscreen display waver a lot with the inferno software? |