| 00:02:34 | CosmicPenguin | Because they are hella complicated.... |
| 00:03:44 | SpaceCoaster | they just wanna sell a book for $100 bucks, saw a really neat book, "Understanding Telephone Electronics", looked fun. |
| 00:05:19 | CosmicPenguin | Did you get it? |
| 00:06:41 | CosmicPenguin | This caller ID stuff is pretty cool...... Every time you get a call, the DSP sends you the updated time from the caller id.... |
| 00:06:55 | CosmicPenguin | That way, as long as the phone keeps ringing, you will always stay somewhat on time.... |
| 00:07:33 | SpaceCoaster | nah, I have to budget at the moment, lots of toys to buy, have to think about which ones. I might buy it later if I still want it, only $21 at amazon. If it looks useful for the TuxScreen the maybe. |
| 00:08:11 | SpaceCoaster | cool, I run rdate at the moment, does it have a battery backed up clock? |
| 00:08:48 | Lenolium | http://cosmic.censoft.com/dspspec.html <-- That appears to be the spec |
| 00:09:40 | CosmicPenguin | SpaceCoaster: I think so... you need to send it the current time in seconds..... so I assume it keeps up to date.... |
| 00:10:09 | CosmicPenguin | But it is maintained by the kernel.... its not CMOS style or anything |
| 00:13:00 | CosmicPenguin | well, folks..... its time for me to mosey on home and get ready for vacation..... |
| 00:13:12 | CosmicPenguin | Have fun with the spec, I hope to see cool things when I log on again..... :) |
| 00:13:28 | SpaceCoaster | when will that be? |
| 00:13:41 | CosmicPenguin | Next monday |
| 00:13:57 | SpaceCoaster | k, have fun, win a million! |
| 00:14:07 | CosmicPenguin | thanks |
| 00:14:08 | CosmicPenguin | bye! |
| 00:16:36 | SpaceCoaster | scanline: are you in? |
| 00:16:47 | scanline | yep |
| 00:17:10 | SpaceCoaster | Did you get the spec? |
| 00:17:19 | scanline | yes |
| 00:17:23 | scanline | super nifty |
| 00:17:47 | scanline | wonders where CosmicPenguin got the information from... |
| 00:18:32 | SpaceCoaster | there's some info in http://tuxscreen.net/download/??? |
| 00:19:11 | scanline | I see |
| 00:19:37 | SpaceCoaster | ssshhh |
| 00:19:52 | scanline | So it was "reverse engineering" ;) |
| 00:20:20 | SpaceCoaster | definitely |
| 00:21:10 | SpaceCoaster | scanline: glad to see that GoRK (and yourself) have progressed the standalone calibration prog, looks nice, sorry that my math seems to be a bit upside down! |
| 00:21:14 | ctr | night all |
| 00:21:19 | SpaceCoaster | bye boris |
| 00:21:26 | ctr | am going to hit the sack |
| 00:21:32 | sammy | morning all |
| 00:21:47 | ctr | spce: you know this ucb1200 hacking you have been doing - how didyou learn tthat stuff? |
| 00:22:09 | scanline | SpaceCoaster: Yeah, I'm glad to have a way to calibrate without Microwindows |
| 00:22:23 | scanline | did GoRK fix the upside-down math? |
| 00:22:43 | SpaceCoaster | ctr: trial and error, and error, and error... |
| 00:23:33 | SpaceCoaster | scanline: he was going to do that tonight, he was working with my math from a week ago so I might already have fixed it here but I am happy to pass the torch on to him. |
| 00:23:50 | scanline | yep |
| 00:24:05 | SpaceCoaster | scanline: it was making my head hurt and it shouldn't be that difficult. |
| 00:24:15 | scanline | I was hoping that throwing together some framebuffer graphics code would jumpstart work on a standalone calibrator |
| 00:24:53 | SpaceCoaster | Yeah sometimes it just needs a spark to get it moving. |
| 00:25:39 | SpaceCoaster | I have dreams of running Java/Swing on this box without X, i.e. direct frambuffer graphics, Savaje do it. |
| 00:25:53 | scanline | can you do that with Kaffe? |
| 00:26:58 | SpaceCoaster | no kaffe doesn't support swing yet and it seems to be pretty behind the times. I want to replace the graphics layer in Sun's JVM but I don't yet have a clue |
| 00:28:00 | SpaceCoaster | ctr: I have a basic driver, that I stripped down for the ucb1200 stuff which I will share with you, I can talk you through building a kernel etc if you want. |
| 00:28:05 | Lenolium | scanline: How big is the picogui binary, and a library for applications to link against? |
| 00:28:31 | ctr | SpaceCoaster: that would rock :) |
| 00:28:37 | scanline | checks... |
| 00:28:37 | ctr | SpaceCoaster: but i have to hit the sack first |
| 00:29:02 | SpaceCoaster | ok, when you have time, where are you. which timezone? |
| 00:29:25 | Lenolium | scanline: On ARM.. (of course) |
| 00:29:36 | SpaceCoaster | ctr: btinternet, hmm, blighty |
| 00:29:54 | scanline | Lenolium: In GoRK's jffs2 image, pgserver is 343584 bytes and libpgui.so is 162621 bytes |
| 00:29:59 | scanline | I don't know why libpgui.so is so big... |
| 00:30:01 | SpaceCoaster | ctr: ? |
| 00:30:08 | ctr | sorry? |
| 00:30:13 | ctr | it's 2:30am |
| 00:30:17 | ctr | and i am feeling it |
| 00:30:23 | SpaceCoaster | ctr: in the uk? |
| 00:30:29 | ctr | yup |
| 00:30:35 | scanline | Lenolium: In GoRK's image i think all the optional parts of pgserver are included, so it can definitely be made smaller if necessary |
| 00:30:50 | ctr | ok |
| 00:30:52 | ctr | night |
| 00:30:58 | SpaceCoaster | ctr: bye |
| 00:31:03 | ctr | SpaceCoaster: will hook up with you tomorrow or sometime soon |
| 00:31:08 | SpaceCoaster | k |
| 00:31:09 | ctr | SpaceCoaster: am very interested =) |
| 00:31:16 | Lenolium | scanline: Ahh... so that leaves plenty of space for a kernel, and an application or two. Now... how small can we make the MySQL library... |
| 00:31:24 | SpaceCoaster | good you can help with the DSP! |
| 00:31:55 | scanline | Lenolium: I'm not sure why libpgui.so is so big... not much code there. Is there something about the ARM architecture that makes shared libraries big? |
| 00:32:03 | GoRK | hello |
| 00:32:20 | scanline | helo |
| 00:32:31 | GoRK | i didnt do any compiling tricks in order to make the lib smaller. there are a couple i think that can be done |
| 00:33:01 | Lenolium | GoRK: Is that library statically linked against ulibc? or does the ld on ARM support multiple dynamic library linkages? |
| 00:33:28 | GoRK | Lenolium: actually, i'm not sure. Let me look |
| 00:33:38 | scanline | libpgui is 4472 lines of C, including comments and blank lines. Shouldn't be that big of a binary... |
| 00:34:05 | GoRK | hmm no it says it is dynamic ... at least that's what ldd thinks |
| 00:34:40 | GoRK | it's about the same size on x86 |
| 00:34:49 | scanline | hmms... |
| 00:35:00 | Lenolium | does gcc support -Os on ARM? (These are all beginner questions, I know, but hey, I just got my tuxscreen today) |
| 00:35:26 | retsaoCecapS | hi Gork |
| 00:35:43 | GoRK | hi. are you trying to tell me that the screen is all flipped? |
| 00:35:44 | scanline | Hmm. on x86, libpgui shrinks to less than half it's previous size when stripped |
| 00:36:05 | GoRK | :) |
| 00:36:48 | retsaoCecapS | oh how the mighty have fallen, yesterday glory, today I don't know which way is up:-( |
| 00:37:24 | GoRK | spacecoaster: i take it that's a yes? |
| 00:37:37 | Lenolium | scanline: BTW, can PicoGUI do AA fonts? |
| 00:37:51 | scanline | Lenolium: not yet |
| 00:39:28 | GoRK | scanline: yeah i just got libpgui on arm down to 48788 bytes |
| 00:39:30 | scanline | Lenolium: PicoGUI reads .bdf bitmapped fonts |
| 00:39:35 | retsaoCecapS | GoRK: I never give a straight answer, apparently :-) |
| 00:39:35 | scanline | GoRK: nice |
| 00:40:39 | Lenolium | scanline: Ahh. So, AA is out of the question? Or can .bdf support transparancy? |
| 00:41:36 | GoRK | .bdf's are bitmapped fonts only |
| 00:41:45 | scanline | Lenolium: You could do AA, it would just require modifications to the font rendering code. PicoGUI doesn't read BDF directly, there's some scripts that turn BDF into a C structure at compile-time |
| 00:42:26 | GoRK | if you were to make a bitmap font that's triple-wide you could easily code subpixel AA font blitting routines |
| 00:42:30 | Lenolium | scanline: Ahh.. cool, so you could theoretically do PNG's and turn them into C structures. |
| 00:42:41 | GoRK | wouldn't even take png's |
| 00:42:58 | scanline | Lenolium: but it would still need modifications to the font code. Right now it expects 1bpp data |
| 00:43:18 | Lenolium | scanline: Yah, that I understand. |
| 00:43:48 | scanline | GoRK: ooh. subpixel AA in PicoGUI would be cool |
| 00:44:08 | Lenolium | GoRK: Yah, that actually wouldn't be too hard. |
| 00:44:16 | GoRK | bitmapped fonts are always going to be 1bpp. but if you had three horizontal pixels for each single lcd pixel, then all you have to do is process 3 bits per pixel and bit shift the R, G, and B values a certain amount based on the 3 bits |
| 00:44:45 | GoRK | it would be very fast actually and coding for subpixel AA would actually be easier than trying to support "regular" AA for crt's (which isnt really nessessary for picogui imo) |
| 00:45:24 | Lenolium | The reason I want AA, is because of the extremely small resolutions, it would make small fonts much easier to read. |
| 00:45:55 | scanline | I think the TuxScreen is the first embedded system PicoGUI runs on that has a color LCD |
| 00:46:16 | GoRK | scanline: well it will run on the ipaq now too :) |
| 00:46:22 | scanline | :) |
| 00:46:29 | GoRK | well.. barring a pointing device driver |
| 00:46:42 | GoRK | and yopy |
| 00:46:59 | scanline | Well, if you include the iPaq you might as well include my laptop. iPaq is on the high-end of embedded systems :) |
| 00:47:37 | Lenolium | Hrm... perhaps we should make it flexable, have three different engines, with #DEFINE from here or there. One do the subpixel rendering, one using greyscale AA (with 4bpp bitmap fonts or something of the sort), and one using the current straight bitmap rendering. |
| 00:47:53 | Lenolium | Ugh... if we went with greyscale AA, we would have to do alpha-blending. |
| 00:48:21 | scanline | Wouldn't be hard to add alpha blending to PicoGUI |
| 00:48:23 | GoRK | to do any kind of aa we have to alpha blend |
| 00:48:34 | scanline | It'd just be another LGOP code |
| 00:49:51 | Lenolium | GoRK: The only problem with subpixel rendering is this: We have a limited palette, so adding all those colors for subpixel rendering would start to eat away the 256-colors we have. |
| 00:50:24 | scanline | Lenolium: Right now PicoGUI doesn't have good palette support. It uses a fixed 2-3-3 RGB palette |
| 00:50:49 | GoRK | Lenolium: not necessarily. picogui currently scales a full color palette to 2-3-3 |
| 00:51:03 | retsaoCecapS | Lenolium: Is there any info on how the palette is structured, I tried FBIOGCMAP and didn't get anywhere? |
| 00:51:39 | scanline | SpaceCoaster: In PicoGUI, or tuxscreen in general? |
| 00:51:45 | Lenolium | Oh, well then, that would make it a bit easier. |
| 00:51:54 | GoRK | spacecoaster: what is the purpose of swapping minx and maxx ? |
| 00:52:02 | retsaoCecapS | scanline: tuxscreen only |
| 00:52:15 | GoRK | in your routines |
| 00:52:50 | retsaoCecapS | err, trying to get them to make sense, i.e. max > min |
| 00:53:30 | GoRK | oh ok i get it now.. nm i am dumb :) |
| 00:53:37 | GoRK | no wonder things are still upside down |
| 00:53:57 | retsaoCecapS | did i do it twice, for luck? |
| 00:54:00 | scanline | SpaceCoaster: FBIOPUTCMAP takes a struct fb_cmap. I think FBIPGCMAP would too |
| 00:54:27 | GoRK | spacecoaster: naw i changed the test i think |
| 00:55:28 | SpaceCoaster | back to normal, if Gork got confused, I don't feel so bad. |
| 00:56:15 | SpaceCoaster | scanline: didn't want to drag you away, do you have a pointer to some code? |
| 00:56:22 | Lenolium | needs to flash Unlock his tuxscreen. Looking for a 12V power supply. |
| 00:57:14 | Lenolium | So, anyone in the sugarhouse area with a good 12V power supply to unlock my flash? |
| 01:01:18 | scanline | SpaceCoaster: oops, didn't notice your question.... there's some in framebuffer_palette() in fbtuxcal's fbvideo.c |
| 01:01:46 | scanline | the ioctl() expects colors to be in the range of 0 to 0xFFFF |
| 01:02:45 | SpaceCoaster | scanline: thanks, i did this sort of stuff a while ago, bit rusty now |
| 01:03:10 | scanline | The framebuffer device is fun :) |
| 01:04:15 | SpaceCoaster | but 8 bits only and fixed size is a hackers dream! |
| 01:04:33 | scanline | yep! Like Mode 0x13 in DOS |
| 01:06:04 | SpaceCoaster | hmm, DOS, never used it, first OS Unix, last OS Unix, but I catch your drift, thanks again, see ya |
| 01:06:19 | scanline | bye |
| 01:18:20 | GoRK | spacecoaster: can you try fbtuxcal now? |
| 01:18:39 | scanline | spacecoaster is gone |
| 01:18:54 | GoRK | whopops |
| 01:19:01 | GoRK | drat |
| 01:19:06 | GoRK | well i think it is fixed |
| 01:19:14 | GoRK | i need someone with a formerly not working screen to test it |
| 01:19:44 | GoRK | it prints a lot of stuff also |
| 01:23:05 | Aelhaz | not working in what respect? |
| 01:23:07 | Aelhaz | flipped? |
| 01:23:59 | GoRK | yeah it was flipped on spacecoasters but worked on mine |
| 01:24:07 | GoRK | i was interested in what was up |
| 01:24:24 | GoRK | did it work for you? |
| 01:24:46 | GoRK | at home |
| 01:24:56 | RedBaron | GoRK: Your talking about the touchscreen being flipped right? |
| 01:25:26 | GoRK | no im wondering if fbtuxcal works on all touchscreens really |
| 01:26:34 | RedBaron | GoRK: okay. |
| 01:27:24 | GoRK | my screen has always worked but it requires x and y flip |
| 01:27:55 | GoRK | i think perhaps the ones that work as of spacecoaster's last kernel patch maybe don't require the flip, but i dont have one to test for sure, so... |
| 01:28:50 | Aelhaz | GoRK: haven't tried it yet . . . . soon . . . |
| 01:29:49 | Aelhaz | the screen at work needs to be flipped, the one at home doesn't, but both have always worked |
| 01:30:26 | Aelhaz | hold on, I'll boot up the box that's connected to the serial console on it |
| 01:30:36 | GoRK | Aelhaz: oh ok i didnt know some required flipping that already worked... hrm ok |
| 01:30:46 | RedBaron | GoRK: I believe you can flip the connector wire for the touchscreen. I pulled my tux apart over the weekend. and I htink if you flip the connector, the x and y lines flip |
| 01:31:57 | GoRK | RedBaron: hey that would definately be a good way to test it :) |
| 01:32:24 | GoRK | takes his stupid phone apart again |
| 01:32:43 | GoRK | once just to peek, once to do jtag, once because i forgot a screw, now this :) |
| 01:33:05 | RedBaron | GoRK: The problem with that wire is that once you pull it off, you can loose its orientation. so I wouldnt be surprised at all if some were just put together wrong |
| 01:33:17 | GoRK | or i could just wait for people to tell me it's broken :) |
| 01:33:39 | RedBaron | GoRK: *Chuckle* |
| 01:34:19 | RedBaron | I made sure to plug mine back in the way the wire wanted to lay out. And it worked correctly. So hey. What ever that means. :) |
| 01:59:59 | Aelhaz | OK, something weird is going on here |
| 02:00:22 | Aelhaz | with 0.3 the touchscreen on this tux is right side up |
| 02:00:37 | Aelhaz | but with buildroot-tux from this afternoon it's flipped |
| 02:01:44 | Aelhaz | At least in Familiar 0.4 it's flipped, maybe microwindows addresses it differently? |
| 02:04:28 | GoRK | hrm |
| 02:04:37 | GoRK | does fbtuxcal work? |
| 02:05:17 | Aelhaz | haven't tried yet |
| 02:05:35 | Aelhaz | have you updated fbtuxcal since this afternoon or can i keep using the binary I've got? |
| 02:05:44 | GoRK | yes i updated it a small bit |
| 02:06:00 | GoRK | and it also spits a bunch of debug info |
| 02:06:05 | Aelhaz | hmm, that's a problem then, I don't have a current buildroot here |
| 02:06:17 | GoRK | i can send you a binary |
| 02:06:23 | Aelhaz | ok |
| 02:06:28 | Aelhaz | email it to alhaz@xmission.com? |
| 02:07:19 | GoRK | http://www.blurbco.com/~gork/tuxscreen/fbtuxcal |
| 02:09:46 | Aelhaz | that works too |
| 02:14:43 | GoRK | did it work? |
| 02:15:05 | Aelhaz | dunno yet, hold on |
| 02:26:23 | Aelhaz | No. |
| 02:26:29 | GoRK | it's flipped |
| 02:26:29 | GoRK | ? |
| 02:26:41 | Aelhaz | yeah, and doesn't appear to be accurately calibrated |
| 02:26:48 | Aelhaz | maybe familiar is doing something odd |
| 02:26:51 | GoRK | can you paste the output? |
| 02:27:09 | Aelhaz | no, it's connected to another box |
| 02:27:17 | Aelhaz | I can probably get it for you somehow tho, hold on |
| 02:34:38 | Aelhaz | # ./fbtuxcal |
| 02:34:38 | Aelhaz | TL(60, 60, 860 818) |
| 02:34:38 | Aelhaz | TR(580, 60, 114 835) |
| 02:34:38 | Aelhaz | BL(60, 420, 866 144) |
| 02:34:38 | Aelhaz | BR(580, 420, 119 162) |
| 02:34:39 | Aelhaz | maxx:863 minx:116 |
| 02:34:41 | Aelhaz | maxx:949 minx:30 |
| 02:34:43 | Aelhaz | maxy:826 miny:153 |
| 02:34:45 | Aelhaz | maxy:938 miny:41 |
| 02:35:05 | Aelhaz | funny. the serial ports on my desktop box have trouble doing 115k with the tux, but my notebook has no trouble at all |
| 02:35:28 | GoRK | hrm and it doesnt work..... that is screwed up.. can you cat /etc/tsdata on the phone and tell me what those 6 numbers are |
| 02:35:51 | Aelhaz | 30 41 949 938 1 1 |
| 02:36:05 | GoRK | ok well that should be exactly right |
| 02:36:17 | GoRK | what GUI are you running? |
| 02:37:20 | Aelhaz | Familiar, the ipaq distribution, on a CF card |
| 02:37:38 | GoRK | ok well something is pretty wrong then |
| 02:37:47 | Aelhaz | maybe what's wrong is familiar |
| 02:38:24 | GoRK | maybe. hrm i think i might know one other thing that's wrong |
| 02:38:30 | GoRK | well |
| 02:38:31 | GoRK | no |
| 02:38:44 | GoRK | maybe |
| 02:38:50 | Aelhaz | with 0.3 familiar had the touchscreen right-way 'round |
| 02:38:58 | Aelhaz | what are you testing yours on? |
| 02:39:04 | GoRK | the GUI should also be responsible for telling the ts device what the screen resolution is |
| 02:39:09 | GoRK | both picogui and nano-x |
| 02:41:19 | Aelhaz | hmm |
| 02:41:26 | GoRK | what that info tells me is that the bottom right hand pixel of your screen is physical screen coordinates 30,41 and the top left pixel is 949,938... since the screen addresses 0,0 as the top left, that means that you need x and y flip enabled... and /etc/tsdata show that they are (the last 2 numbers are 1) |
| 02:42:13 | Aelhaz | hmm |
| 02:42:55 | GoRK | so when you move the stylus around in the bottom right corner, the cursor moves around in the top left? |
| 02:43:03 | Aelhaz | Right |
| 02:43:07 | Aelhaz | but Not Terribly Accurately |
| 02:43:43 | Aelhaz | I've tried to mess around with xcalibrate in familiar, and it doesn't work fer shit on the tuxscreen, maybe it's saved some calibration stuff somewhere that it'sloading . . |
| 02:45:01 | GoRK | hrm... maybe... if the GUI messes with the calibration ioctl's then it's liable to hose fbtuxcal's data |
| 02:45:27 | Aelhaz | hmm, something's not right here |
| 02:45:33 | GoRK | that data you posted shows that fbtuxcal should be working |
| 02:49:59 | scanline | Yay. Except for handling HTML tags, this parser works :) |
| 02:50:57 | GoRK | so what does it do exactly, remove HTML tags? |
| 02:51:35 | scanline | It passes HTML tags to a function (that doesn't do anything yet), handles whitespace, stuff like , and converts the text into a divnode tree for word wrapping |
| 02:54:05 | Aelhaz | well, familiar is not loading calibration stuff |
| 02:54:14 | Aelhaz | maybe it's addressing the touchscreen ass backwards . . . |
| 02:55:12 | andersee | Anybody had a chnce to read over the "Shannon DSP specification"? http://cosmic.censoft.com/dspspec.html |
| 02:55:20 | scanline | yes |
| 02:55:51 | GoRK | oh yeah that's what i was going to look at |
| 02:56:24 | andersee | I think that little baby is going to get things rolling.... |
| 02:56:48 | andersee | Well, I'm off to buy some Mtn Dew for another long night... |
| 02:56:54 | GoRK | heh. im already fiddling with libiax on the phone |
| 02:57:04 | GoRK | are you going to hack on the dsp or something else? |
| 02:57:11 | andersee | is away: I'm busy |
| 02:59:02 | GoRK | wonders if anyone has an arm-gdbserver/and x86 gdb host for it .... :) |
| 02:59:09 | GoRK | maybe i should ask in handhelds.org channel |
| 02:59:14 | GoRK | if they dont just call me stupid |
| 02:59:46 | scanline | I can just use ethereal to trace the network communication between client and server |
| 02:59:55 | scanline | likes looking at hex dumps of picogui packets... |
| 03:00:51 | Aelhaz | Hmm |
| 03:01:50 | GoRK | Aelhaz: did you ever try running my picogui image on your phone at home? |
| 03:02:00 | Aelhaz | GoRK: No, but, hold on while I experiment |
| 03:02:04 | Aelhaz | I got it to work |
| 03:02:08 | Aelhaz | but I'm not sure how |
| 03:02:20 | Aelhaz | <-- test guy, won't report dick until it's reproducable |
| 03:02:23 | GoRK | but you didnt have to modify fbtuxcal? |
| 03:02:32 | Aelhaz | right |
| 03:03:19 | GoRK | well that is partially good news for me.. If i can get some people with one or two zeros in those last slots i'll be fairly confident that it works |
| 03:03:26 | Aelhaz | OK |
| 03:03:56 | Aelhaz | Well, here, my plan was that I'd calibrate the screen and then add fbtuxcal to linuxrc |
| 03:04:03 | GoRK | does the calibration appear accurate? |
| 03:04:31 | Aelhaz | I hadn't been specifying a path so it wasn't running, and then familiar was coming up flipped |
| 03:04:38 | Aelhaz | now it works |
| 03:04:41 | Aelhaz | loose nut behind the keyboard |
| 03:05:10 | GoRK | ah! the ID-10-T error. Problem exists between chair and keyboard, etc. |
| 03:05:10 | Aelhaz | GoRK: Sorta |
| 03:05:31 | Aelhaz | it's probably technically accurate |
| 03:05:47 | Aelhaz | but looks odd because the tip of the stylus extends past where the pointer is on the screen |
| 03:05:53 | GoRK | Aelhaz: if you tap too lightly it gets a bad reading. you can run fbtuxcal -f to overwrite it |
| 03:06:14 | Aelhaz | looks like it's off by a few pixels |
| 03:06:18 | GoRK | Aelhaz: yeah it does taht a little bit. it's not going to be pixel accurate without tweaking the kernel driver more to take into account more than simple linear distortion |
| 03:06:34 | Aelhaz | GoRK: it's way too easy to tap one spot four times and get totally bogus calibration |
| 03:07:05 | GoRK | which i am going to play with at some point in time... I believe that I can get a pixel-accurate calibrator that will never leave you with a bad calibraiton |
| 03:07:16 | scanline | Might be a good idea to reject a tap if it's too close to the previous one, to prevent double-tapping |
| 03:07:51 | GoRK | scanline: or just sleep for 1/2 second after you read the successful tap would do it also |
| 03:08:00 | scanline | yep |
| 03:08:07 | scanline | well.. |
| 03:08:13 | scanline | wouldn't the data still be in a buffer then? |
| 03:08:32 | Aelhaz | yeah, ignore everything for a half a second |
| 03:08:34 | Aelhaz | that should do it |
| 03:08:35 | GoRK | not if you flush it after that 1/2 second |
| 03:08:48 | scanline | yeah |
| 03:35:29 | Tangent | goes to bed.. night all |
| 03:38:45 | Aelhaz | guh. |
| 03:39:00 | Aelhaz | familiar is giving me so much shit . . . |
| 03:40:50 | andersee | Aelhaz: And all this time I thought that was a result of eating... |
| 03:40:54 | andersee | ;-) |
| 03:41:00 | GoRK | lol |
| 03:41:22 | Aelhaz | ah, a joke. Funny. |
| 03:41:47 | Lenolium | Yay! I found a 12V power supply, thank you fan header! |
| 03:42:08 | andersee | I'm back fromthe store, cany and Mtn Dew in hand... |
| 03:42:11 | andersee | err. |
| 03:42:12 | andersee | candy |
| 03:42:41 | scanline | please email me some mountain dew :) |
| 03:42:50 | Lenolium | andersee: You are in Orem, right? |
| 03:43:02 | andersee | Lenolium: Salt Lake actually |
| 03:43:23 | Lenolium | andersee: Ahh, cool, as am I. (Sugarhouse, to be exact) |
| 03:43:23 | Proteus | Heh. Lots of people from Yewtar hacking away with these things? |
| 03:44:20 | andersee | Proteus: Helps to be close to the source of all goodness (in this case, the source of the phones) |
| 03:44:40 | andersee | Proteus: Also helps that Tim want evangelizing to all the local Linux users groups... |
| 03:45:09 | andersee | s/want/went/ |
| 03:45:46 | Proteus | andersee: Gotcha. |
| 03:45:48 | Proteus | *lol* |
| 03:45:51 | Lenolium | Not to mention the fact that it's easier to buy one when you can just drive for a bit and pick one up. |
| 03:45:58 | Proteus | I think I could turn this box into a 32 channel synth. :-) |
| 03:46:11 | Proteus | Yeah - but I've got a POB and PayPal. |
| 03:46:16 | Proteus | I'd order 10 if I had the cash right now. |
| 03:49:30 | andersee | Proteus: I got two -- that was the limit of my play money. |
| 03:51:32 | Lenolium | I got two today, hopefully more soon, My girlfriend seems pretty impressed with them, so I don't think she'll complain too much when more start showing up. |
| 03:54:28 | Proteus | Anyone get schematics of this thing yet? |
| 04:00:32 | Aelhaz | Well, rxvt being munged up in familiar is NOT a font availability issue |
| 04:07:13 | rw-rw-rw- | hello Erik, I am on this server net too, I am usually in #linuxnewbies |
| 04:09:49 | GoRK | hahaha |
| 04:09:50 | rw-rw-rw- | anyway, I am just cooking it all up under the new uClinc |
| 04:09:54 | rw-rw-rw- | uClibc |
| 04:10:13 | | privet, andersee |
| 04:10:13 | andersee | rw-rw-rw-: howdy |
| 04:10:40 | rw-rw-rw- | previt? russian? |
| 04:10:46 | andersee | no idea? |
| 04:10:54 | | Sorry, andersee, I'll keep my mouth shut. |
| 04:10:54 | andersee | ibot shut up |
| 04:11:07 | andersee | silly robot |
| 04:11:22 | rw-rw-rw- | I was the person just mailing the list on the shell exit on child signal |
| 04:11:30 | Aelhaz | ok, like many things in the Familiar distribution, the hacks that make rxvt-aa do it's magic don't work if you don't have a tiny screen . . . |
| 04:11:36 | Aelhaz | the regular rxvt works |
| 04:11:37 | rw-rw-rw- | thought I might just pop in here |
| 04:11:48 | andersee | k |
| 04:11:54 | andersee | any progress |
| 04:12:11 | rw-rw-rw- | just about to build a rootfs.gz and boot it up |
| 04:14:33 | Lenolium | I guess you can't build a null modem connection when one cable only has pins 1,3,5,8 connected (it is a serial mouse I was about to sacrifice) |
| 05:05:43 | andersee | Howdy all |
| 05:06:00 | andersee | I just fixed the uClibc-on-ARM signal handling bug... |
| 05:06:29 | andersee | So updating you libc and recompiling stuff is probably in order |
| 05:07:38 | Lenolium | andersee: You don't happen to have a spare 9-pin null modem cable you would be willing to give up, do you? :) |
| 05:07:50 | andersee | I've got one... |
| 05:08:15 | andersee | I've also got a spare null modem-izing adaptor for a regular 9 pin serial cable |
| 05:08:28 | Lenolium | ... that you would be willing to trade for a late night greasy burrito? |
| 05:08:40 | andersee | Hmm.... |
| 05:08:44 | andersee | Hmmm..... |
| 05:08:46 | Lenolium | (the cable, not the adapter. I don't have any 9-pin stuff lying around) |
| 05:08:50 | andersee | Nah |
| 05:09:12 | andersee | I like my cable -- it a nice belkin one |
| 05:09:26 | andersee | Cost me like $10 bucks |
| 05:09:41 | Lenolium | Ahh.. I was hoping you had extras... I guess I should see if I can steal one from work. |
| 05:12:20 | GoRK | i use an ethernet cable as my null modem |
| 05:12:30 | GoRK | w/ two of those little db9<->rj45 adaptors |
| 05:12:42 | GoRK | i use those for everything |
| 05:12:54 | Lenolium | GoRK: The problem is, I need the ends. |
| 05:13:09 | Lenolium | GoRK: So unless you are in Salt Town also, I can't even steal those from you. |
| 05:13:46 | GoRK | nope. |
| 05:13:53 | GoRK | not there |
| 05:15:44 | Lenolium | Hrm.. I wonder how many pins are wired in this funny device. |
| 05:15:57 | Lenolium | And, I don't need any handshaking on this serial cable to the TuxScreen, right? |
| 05:17:16 | Russ | right |
| 05:17:21 | Russ | no flow control |
| 05:18:36 | andersee | Russ: howdy |
| 05:18:53 | andersee | Russ: tests done? |
| 05:19:08 | Russ | midterms done |
| 05:19:13 | Russ | merging a kernel now |
| 05:19:28 | andersee | Russ: BTW, I just fixed an ugly arm specific sighandling but in uClibc... |
| 05:19:30 | andersee | cool |
| 05:19:51 | andersee | rmk's latest is looking pretty goo |
| 05:19:56 | andersee | err. good |
| 05:19:56 | GoRK | haha HAPPY TONE and SAD TONE |
| 05:20:13 | GoRK | wonders what HAPPY TONE sounds like..... |
| 05:20:20 | Lenolium | GoRK: So do I... |
| 05:20:22 | scanline | wahoo. I have <p> and <br> tags working |
| 05:20:30 | scanline | hehe. happy tone :) |
| 05:20:57 | Lenolium | scanline: Making a web browser? |
| 05:21:11 | scanline | Lenolium: Yep. HTML rendering support for PicoGUI |
| 05:22:08 | GoRK | scanline: do you have support in there for <blink> ... make sure you dont forget <blink> |
| 05:22:18 | scanline | Noooo! Not <blink>! |
| 05:22:43 | GoRK | scanline: initially it was meant as a joke, but you know what... it actually might be a very useful tag on a handheld or something |
| 05:22:45 | scanline | Heh. I'm not even sure how I'd implement blink... Prolly have to have a linked list of nodes that are blinking, and update them on a timer. Not worth it... |
| 05:22:52 | rw-rw-rw- | andersee: I take back that comment about my filesystem not pivoting with msh, it's fine, I overwrote my linuxrc script with the busybox installed one because I still had it defined in my Config.h |
| 05:22:57 | scanline | what would it be useful for? |
| 05:23:28 | andersee | rw-rw-rw-: ahh. oops. |
| 05:23:54 | GoRK | BLINKING! |
| 05:24:12 | scanline | haha |
| 05:24:20 | GoRK | you could have it blink anything, images, whole divnodes!!!! |
| 05:24:31 | scanline | Ooh... We could make blinking buttons! |
| 05:24:41 | scanline | Can't click it while it's invisible |
| 05:24:45 | GoRK | they appear. they dissaper! where did it go!?! |
| 05:25:07 | scanline | even better- blinking text fields |
| 05:25:40 | scanline | it should be illegal to use <blink> and <form> in the same file :) |
| 06:00:32 | GoRK | hey CVS didnt catch my change to this one file.. is there any way i can force it to be more picky? |
| 06:01:58 | Lenolium | GoRK: Yes, let me see if I remember. |
| 06:02:11 | Lenolium | GoRK: Easy way: touch file_i_changed |
| 06:02:22 | GoRK | actually it picks up that it was changed but it shows -0+0 lines |
| 06:02:36 | scanline | was the change only whitespace? |
| 06:02:48 | GoRK | nope |
| 06:02:52 | GoRK | it was a patch to a diff |
| 06:03:24 | Lenolium | GoRK: cvs commit -Af file_you_changed ( I believe) |
| 06:03:27 | GoRK | and there was an Id: tag at the top of one of the C files that was breaking the patch applying properly and i changed the line to reflect what it should have been |
| 06:03:52 | Lenolium | ooooh. |
| 06:04:07 | Lenolium | That might be the problem. Aren't those Id: tags dynamically generated? |
| 06:04:15 | GoRK | e.g. the diff itself is the file in cvs |
| 06:04:49 | GoRK | -A is not recognized |
| 06:05:40 | GoRK | we'll see if this works |
| 06:06:15 | BZFlag | patches need to be checked in with -ko or -kb |
| 06:06:32 | BZFlag | cvs diff does not look in $foo entries otherwise. |
| 06:06:40 | GoRK | ahh the lovely nuances of CVS :) |
| 06:07:06 | BZFlag | and Id etc get's auto changed on checkin if neither flag is set. |
| 06:07:24 | GoRK | which is probably why it broke in the first place |
| 06:07:26 | GoRK | heh ok |
| 06:07:29 | BZFlag | GoRK: that's true of any version control software I've ever used. ;-) |
| 06:07:42 | BZFlag | GoRK: cvs admin -ko file.name |
| 06:08:58 | BZFlag | all non-text files in cvs should be -ko as well. ie: tar files. |
| 06:09:17 | BZFlag | well, those should be -kb, but that has the same effect. |
| 06:15:08 | GoRK | BZFlag: thanks for the info. I did not know that was the reason it was breaking :) |
| 06:18:38 | Lenolium | GoRK: Yup, it'll break some binary files pretty quickly that way too. :) |
| 06:20:33 | GoRK | hmm so it looks like the clock is in the DSP.. which explains why linux loses the time every time it boots |
| 06:23:40 | Lenolium | GoRK: Well, the DSP has a function to get the time, so it shouldn't be all that hard to grab. And when you do, can you play the HAPPY_SOUND? :) |
| 06:24:27 | GoRK | it's all about the HAPPY SOUND |
| 06:26:06 | Lenolium | Hrm... I don't see anything in here that would even allow us to get close to doing VoIP though. |
| 06:27:37 | GoRK | DSP_LOAD_MSG_DATA, DSP_USRP_RDY_4_NEXT_BLK, DSP_START_REC, DSP_STOP_REC, DSP_PLAY_MSG DSP_STOP_MSG |
| 06:28:14 | Lenolium | Alright... I am wrong. |
| 06:28:19 | Lenolium | Not the first time that has happened... :) |
| 06:28:29 | GoRK | well they are under unused/unknown |
| 06:28:36 | GoRK | it's possible they dont work |
| 06:28:58 | GoRK | i am still completely scared of the lucent document that says they had to modify the ROM for VoIP things |
| 06:29:59 | Lenolium | Where is that lucent document? |
| 06:30:28 | GoRK | http://www.lucent.com/minds/techjournal/jul-sep2000/pdf/paper14.pdf |
| 06:35:00 | Lenolium | Oohhh.. looks interesting... |
| 06:40:08 | Lenolium | prepares to unlock his flash |
| 06:49:50 | Lenolium | Gotta reboot, and re-enable my serial port. |
| 06:49:53 | Lenolium | *plong* |
| 06:50:26 | Proteus | Yum. |
| 07:00:21 | Lenolium | *pling* |
| 07:24:51 | scanline | is away: sleep |
| 08:29:05 | Lenolium | WooHoo! |
| 08:29:09 | Lenolium | FlashUnlock COMPLETE |
| 08:53:13 | Lenolium | *plong* |
| 12:27:52 | RedBaron | Morning all |
| 12:28:07 | Sammy | everin |
| 12:28:09 | mallum | hi |
| 12:29:47 | Tangent | Hi |
| 12:31:30 | RedBaron | Wheee...X just crashed on my desktop. |
| 12:42:37 | prpplague | morning all |
| 12:46:49 | RedBaron | Howdy prpplague |
| 12:51:17 | ctr | is away: shower |
| 12:59:33 | prpplague | well bellsouth now has 8 out of 9 frame-relay circuits running, maybe we can get this finished this morning so i can play with my tuxscreen |
| 13:16:51 | prpplague | mallum: you alive this morning? |
| 13:18:04 | mallum | prpplague: yes I am this morning |
| 13:18:49 | prpplague | mallum: btw, i like your web site, saw your site about a year ago, final put 2+2 together |
| 13:20:52 | mallum | prpplague: thanks ... which one ? |
| 13:49:18 | prpplague | mallum.com |
| 13:49:42 | mallum | prpplague: ahh, havn't updated that in a while |
| 13:50:55 | prpplague | mallum: oh well, the site really left a mental image |
| 13:50:55 | mallum | prpplague: check http://10.am too |
| 13:51:09 | mallum | prpplague: though its currently a bit rusty |
| 13:53:27 | prpplague | looks nice, i'll bookmark it for future ref |
| 15:01:57 | AlHazWork | So, pardon my ignorance, should it be sufficent to "cvs up" "make clean" "make" in order to build new buildroot? |
| 15:14:52 | AlHazWork | rc.pcmcia still missing PCIC=i82365 |
| 15:15:25 | AlHazWork | and still missing te100pc16r.conf |
| 15:15:26 | AlHazWork | minor buglets, tho |
| 15:55:58 | Lenolium | *pling* |
| 15:58:08 | RedBaron | Give me one ping, Vincine. One ping. |
| 16:00:40 | SpaceCoaster | RedBaron: pong |
| 16:00:58 | RedBaron | SpaceCoaster: Hey...I asked for one ping... |
| 16:01:17 | SpaceCoaster | RedBaron: ooops, ping |
| 16:01:22 | Lenolium | is reading the stuff in http://tuxscreen.net/download/howdidthisgethere/shandsp.c , very, very interesting. |
| 16:01:23 | RedBaron | SpaceCoaster: hehe |
| 16:09:32 | AlHazWork | you realize that if linux were windows, our browsers would download, compile, and try to execute that code . . . |
| 16:10:07 | RedBaron | AlHazWork: and it would all run in VB. including the kernel. |
| 16:10:15 | RedBaron | AlHazWork: well...*try* to run |
| 16:10:52 | AlHazWork | walk, at best . . |
| 16:12:31 | RedBaron | printf("\n======================== P2OS Variables ==========================\n"); |
| 16:12:32 | RedBaron | printf("KEYWORD CURRENT VALUE\n"); |
| 16:12:32 | RedBaron | printTable(TABLE_VARIABLE); |
| 16:12:37 | RedBaron | gah |
| 16:12:39 | RedBaron | sorry about that. |
| 16:12:51 | AlHazWork | I don't care at all about the modem, but it sure would be neat if the rest of the telephony functions worked |
| 16:15:48 | AlHazWork | best case scenario it would be a lousy modem |
| 16:16:42 | RedBaron | AlHazWork: I dont see any reason to use the tux as anything other than a small display on a local ether network (wired or wireless). |
| 16:17:13 | AlHazWork | RedBaron: if you really wanted to you could use it as a stand-alone device |
| 16:17:32 | AlHazWork | I've browsed the web in inferno in it. it was painful. |
| 16:17:50 | RedBaron | AlHazWork: Eh. I bet it was. |
| 16:18:15 | RedBaron | AlHazWork: probably would have been better with a 32M SODIMM in it...but still |
| 16:18:28 | AlHazWork | it wasn't a memory issue |
| 16:18:35 | RedBaron | okay |
| 16:18:41 | AlHazWork | it was an "Oh man it's rendering it AGAIN" issue |
| 16:18:43 | RedBaron | yah...keep forgetting there is no swap. |
| 16:21:35 | RedBaron | show_constants(); |
| 16:21:49 | RedBaron | AlHazWork: I guess it was at 133Mhz and not 206mhz? |
| 16:22:13 | scanline | is back (gone 08:57:22) |
| 16:23:28 | AlHazWork | RedBaron: Yeah, but it's like a 28k modem anyway |
| 16:34:35 | Lenolium | Alright, I'm preparing to load linux on my TuxScreen... any words of advice/warning? or do the instructions work, and the latest versions run? |
| 16:35:49 | AlHazWork | what version blob / linux are you putting on? |
| 16:38:08 | Lenolium | Hrm.. I guess I should build linux first, eh? |
| 16:38:15 | AlHazWork | yeah |
| 16:38:32 | AlHazWork | if you've got disk space & cycles to spare, buildroot-tux is a good way to go about it |
| 16:38:49 | AlHazWork | newer versions of blob run at 115k and don't require you to switch speeds, one thing to look out for |
| 16:40:17 | AlHazWork | my duron system at home has trouble with 115k, dunno why, doesn't make sense |
| 16:40:22 | Lenolium | I see the blob homepage is up to 2.0.4 |
| 16:40:42 | AlHazWork | don't use blob from the blob homepage |
| 16:40:55 | AlHazWork | use blob from the tuxscreen ftp or cvs |
| 16:41:31 | Lenolium | Ahh.. I was thinking of using the V0.4 release. |
| 16:42:15 | AlHazWork | best not to fool around too much with the boot loader unless you have jtag |
| 16:42:36 | Lenolium | AlHazWork: Wise words... wise words. |
| 16:44:20 | Lenolium | I hate inferno though, XFS's inode numbers are too big for it to handle, and it pukes. |
| 16:46:36 | GoRK | whoops i forgot to include a file |
| 16:46:52 | scanline | is away: class |
| 16:54:42 | RedBaron | AlHazWork: Does the kernel not support the 32M SODIMM? The page on the wiki seems to indicate that it would just work. |
| 17:12:04 | GoRK | RedBaron: the linux kernel is not the issue. it's blob. Blob does not support anything larger than 8mb sodimm in addition to the 8mb onboard. theoretically there's nothing keeping us from supporting up to the 64mb sodimm w/ blob either |
| 17:14:51 | RedBaron | GoRK: Hmmm. So blob cant just run ignoring everything beyond 8M? I thought that once the kernel was started, it ignored all boot loaders and most of bios and figured things out itself. |
| 17:16:26 | GoRK | RedBaron: I dont know the internals of it, but I dont think it's quite that simple... |
| 17:16:47 | RedBaron | GoRK: I dont doubt that. :) |
| 17:17:10 | GoRK | RedBaron: on ARM especially, the bootloader has to do a bunch of stuff that it doesnt ordinarily have to do on other architectures |
| 17:17:53 | RedBaron | GoRK: *nod* I just recall that on x86 the linux kernel ignores a lot of what BIOS tells it, since most of the BIOSes are horribly broken. |
| 17:19:39 | GoRK | RedBaron: for instance when a machine comes back from a power management sleep or suspend, it starts at the bootloader. The bootloader has to 1) recognize that it's coming from suspend and not a clean powerup 2) do normal system register setup and 3) jump to the resume vector that was in one of the processors registers |
| 17:19:51 | GoRK | RedBaron: on ARM, not x86 |
| 17:20:03 | RedBaron | GoRK: *nod* |
| 17:22:10 | RedBaron | GoRK: Unless there are some things that are dependent on memory size, I dont see (in general) a reason why blob cant just ignore the fact that there is more than 8M ram. Like things hanging off of addresses beyond physical memory and being mapped directly to 8M+1byte or something like that. Granted, I know nothing of blob and the arm architecture. :) |
| 17:25:19 | GoRK | RedBaron: i dont either, but i do know that something's amiss otherwise it'd probably already work :) |
| 17:26:10 | AlHazWork | Someone mentioned something about the SA1100 datasheet saying that if there are multiple memory banks they all have to be the same size |
| 17:26:26 | AlHazWork | currently the blob sees the on-board memory as 2 banks of 4 and the sodimm as another 2 banks of 4 |
| 17:27:18 | RedBaron | AlHazWork: Hrm...So why does the wiki say that a 32M SODIMM works? |
| 17:27:30 | AlHazWork | RedBaron: it works, as 8 megs, and in inferno it works as 32 megs |
| 17:27:54 | RedBaron | GoRK: I guess so. I just hadnt heard about it not working. |
| 17:28:00 | RedBaron | AlHazWork: Ah. |
| 17:28:33 | GPSFan | AlHazWork: That was me, read section 10.1 of thw SA1100 data sheet. I'm also almost done with the schematic of the DRAM & SODIMM part of the mainboard. |
| 17:29:11 | AlHazWork | gps:*nod* |
| 17:29:20 | AlHazWork | 16 megs of ram is a bit tight for some stuff |
| 17:30:04 | GPSFan | I think, that if the refresh page size is the same, that the SA1100 can be fooled into having non-identical sized bandk. |
| 17:30:05 | RedBaron | AlHazWork: Yah..once you hit 32M, your basicly good to go. 40M is just icing. :) |
| 17:30:27 | AlHazWork | RedBaron: Yup |
| 17:33:55 | AlHazWork | makes distclean, cvs up's, and attempts to buildroot w/ lpp and microwindows |
| 17:36:01 | Lenolium | Gaar! I hate inferno. |
| 17:36:31 | GoRK | Lenolium: it's the hip thing to do. |
| 17:36:56 | signal11 | are the patches at http://russ.dhs.org/files/tuxpatches/ the latest? |
| 17:36:59 | RedBaron | GoRK: flashing blob/linux is the hip thing to do |
| 17:37:03 | Lenolium | GoRK: It won't find my blob. |
| 17:37:11 | signal11 | they're for 2.4.7, dated in august |
| 17:38:13 | GoRK | signal11: if you're building from scratch, use buildroot-tux from the tuxscreen sourceforge project |
| 17:38:30 | GoRK | signal11: russ has some work done against 2.4.9 but last i heard was having many stability problems, etc |
| 17:38:43 | signal11 | i just want whatever is stable + touch screen fix |
| 17:38:50 | signal11 | i'll look at builtroot-tux. thanks |
| 17:39:49 | Lenolium | Where do I need to put the blob in order for Inferno/Win32 to see it? |
| 17:39:50 | GoRK | buildroot has really a suite of patches ... 2.4.7 + rmk's arm patches + russ's shannon patches + touchscreen fix + touchscreen brightness/contrast/backlight controls + optional Linux Progress Patch |
| 17:40:10 | RedBaron | Lenolium: In the directory structure that the inferno installer makes |
| 17:40:20 | GoRK | Lenolium: should just be somewhere under your inferno root directory. i used <inferno-root>/tmp |
| 17:40:34 | RedBaron | Lenolium: In linux I did /usr/local/inferno/<file> |
| 17:41:51 | Lenolium | I run: c D!/tmp/blob F!all , the tuxscreen goes into "reading" with a progress bar at the bottom, Inferno goes back to the ; prompt |
| 17:43:12 | RedBaron | Lenolium: Try just c D!tmp/blod F!all |
| 17:43:18 | RedBaron | blob |
| 17:43:37 | RedBaron | I dont think inferno interprets leading /'s correctly |
| 17:44:36 | Lenolium | RedBaron: Same exact thing. |
| 17:45:21 | RedBaron | Lenolium: In windows...where is the blob file located? |
| 17:46:08 | Lenolium | C:\users\inferno\, c:\users\inferno\tmp, c:\users\inferno\Nt\386\, c:\inferno\ |
| 17:47:28 | RedBaron | Lenolium: Hmm...just noticed a new comment on the infernoremote page saying that in win98, someone couldnt send files over 4K in size. You running NT? |
| 17:47:31 | Lenolium | My flash is totally unlockd and everything. |
| 17:47:43 | Lenolium | Nope, I'm running 98 |
| 17:48:14 | RedBaron | Okay. maybe its the same problem you having. The blob file is 14K. |
| 17:48:29 | RedBaron | takes off for a meetin...whee |
| 18:06:38 | Lenolium | Curse you inferno! |
| 18:08:19 | Lenolium | emuinit: unable to load /dis/install/install.dis: inode number too large |
| 18:08:37 | GoRK | Lenolium: make a loopback ext2 filesystem |
| 18:08:45 | GoRK | Lenolium: you are on XFS right? |
| 18:09:00 | Lenolium | GoRK: Yup. |
| 18:09:06 | signal11 | entry-armv.S: Assembler messages: |
| 18:09:08 | signal11 | entry-armv.S:781: Error: Internal_relocation (type 187) not fixed up (OFFSET_IMM) |
| 18:09:19 | signal11 | during kernel compile |
| 18:09:33 | signal11 | are my tools out of date or something? |
| 18:09:35 | GoRK | signal11: argh i saw that before... what was it |
| 18:09:46 | GoRK | signal11: what toolchain are you using? |
| 18:09:48 | signal11 | sept 8. |
| 18:09:56 | signal11 | is the date on my /skiff directory |
| 18:10:17 | signal11 | sept 8 2000 |
| 18:10:26 | GoRK | are you compiling this natively on a skiff? |
| 18:10:48 | GoRK | or x-compiling? |
| 18:10:58 | signal11 | x-compiling |
| 18:11:15 | signal11 | i have used this toolchain successfully before when i first got the tux screen |
| 18:11:25 | signal11 | but i was manually applying patches |
| 18:11:40 | GoRK | hrm... i have seen that message before but i forgot what i did. argh |
| 18:12:45 | GoRK | i am using the toolchain from emdebian. Others seem to have determined that it works the best (especially if you are using debian) but they provide rpm's too |
| 18:13:24 | signal11 | i found a web page that says to remove constants.h and rebuild |
| 18:14:01 | signal11 | seems to be working so far |
| 18:14:02 | GoRK | BZFlag do you remember the entry-armv.S: Assembler messages: / entry-armv.S:781: Error: Internal_relocation (type 187) not fixed up (OFFSET_IMM) thing? |
| 18:14:08 | signal11 | yep |
| 18:14:24 | signal11 | gork: http://lists.arm.linux.org.uk/pipermail/linux-arm-kernel/2001-July/003952.html |
| 18:14:35 | GoRK | hrm.. i don't remember doing that at all |
| 18:14:44 | GoRK | signal11: are you using buildroot? |
| 18:15:12 | signal11 | yes |
| 18:15:20 | signal11 | i did "make world" |
| 18:16:52 | Lenolium | *plong* |
| 18:17:14 | Lenolium | *pling* |
| 18:17:53 | Lenolium | Uhm... what's the proper command syntax for mounting a loopback device.. I managed a hard lockup |
| 18:18:15 | GoRK | -o loop |
| 18:18:34 | Lenolium | mount -o loop file |
| 18:18:39 | GoRK | e.g. mount -o loop /path/to/file /mnt/point |
| 18:18:44 | Lenolium | Ahh... -k- |
| 18:19:06 | GoRK | right.. make sure your file is sized in blocks of 1024 |
| 18:19:14 | GoRK | if you're using ext2 on it |
| 18:19:54 | GoRK | or multiples of 512 bytes works too i guess |
| 18:20:23 | GoRK | anyway i dont seem to remember that it's absolutely necessary but it can't hurt |
| 18:21:42 | Lenolium | Looks like I'm gonna have to resize it. The installer is 90M |
| 18:22:02 | Lenolium | Oh, wait, that's with my 64M loopback partition-file |
| 18:25:41 | GoRK | 83854 /root/tuxscreen/inferno |
| 18:25:54 | GoRK | err |
| 18:25:56 | GoRK | 82M /root/tuxscreen/inferno |
| 18:26:33 | Lenolium | Yeah, 64M is not enough for the installer and inferno. |
| 18:28:24 | Lenolium | Alright, going for a 256M ext2 file. |
| 18:29:49 | Lenolium | Alright, installed, it uses 50M, without the installer files hanging around. |
| 18:30:51 | scanline | is back (gone 01:43:59) |
| 18:34:57 | Lenolium | Now it complains of this: panic: loading "/dis/emuinit.dis": file does not exist |
| 18:35:02 | Lenolium | What manual didn't I read? |
| 18:43:19 | Lenolium | Alright, symlinked /usr/inferno to where I actually put inferno (/inferno) and now it loads. |
| 18:46:21 | Lenolium | SWEET! |
| 18:46:24 | Lenolium | BLOB IS LOADED |
| 18:48:18 | BZFlag | Lenolium: congrats! |
| 18:48:47 | BZFlag | please add any comments that might be helpful to others on the InfernoRemote page. |
| 18:49:25 | Lenolium | Inferno hates RobAagaard would probably be appropriate. |
| 18:49:27 | BZFlag | to work... |
| 18:58:21 | Lenolium | Hrm... having problems saving the ramdisk to flash. |
| 18:58:56 | Lenolium | Urg... I don't want to reboot now... |
| 19:00:28 | Lenolium | Alright, the kernel and blob are fine, we'll keep those, and try the ramdisk again. |
| 19:03:50 | Lenolium | Wahoo! It booted! |
| 19:04:11 | Lenolium | I can't login, but it booted. |
| 19:06:57 | AlHazWork | can't log in? |
| 19:07:28 | Lenolium | I think I should probably stick my SODIMM back in there first. :) |
| 19:08:17 | AlHazWork | yeah, linux may have issues without it |
| 19:10:47 | Lenolium | Ahhh.... much better |
| 19:11:20 | Lenolium | Now all I need is a 3.3V pcmcia ethernet card. |
| 19:29:07 | signal11 | w00 h00, a working touch screen. |
| 19:29:10 | Lenolium | *pling* |
| 19:29:11 | signal11 | you guys rock |
| 19:29:26 | signal11 | i just need to flip X and Y |
| 19:29:32 | Lenolium | Well, I guess I can do PPP for now. |
| 19:30:23 | Lenolium | is back (gone 00:01:13) |
| 19:31:37 | AlHazWork | signal11: fbtuxcal seems to handle that at this point |
| 19:31:46 | AlHazWork | but so far i've only tested it with familiar |
| 19:32:24 | signal11 | cool. i'm so excited. after a month of no touch screen, i finally can make it work |
| 19:34:10 | AlHazWork | I'd like to be able to use the tux in my livingroom to manage phone numbers |
| 19:34:15 | AlHazWork | since i can only remember about 5 at a time |
| 19:34:55 | AlHazWork | be nice to just prod at the phone a bit and have it dial for me |
| 19:35:40 | signal11 | alhazwork yeah, i plan to put it in the living room so i don't have to walk _all the way_ into the next room to browse |
| 19:35:46 | Lenolium | AlHazWork: Have you tried the phone in Inferno mode? it does exactly that. :) |
| 19:36:09 | signal11 | yeah, that inferno app was nice |
| 19:36:31 | signal11 | i can't believe it won't work as an answering machine though; that seems like the most important feature |
| 19:36:37 | AlHazWork | lino: Yeah, but the inferno gui sux |
| 19:36:58 | signal11 | with the dsp driver and a pcmcia hard drive, would it be possible to build an answering machine out of it? |
| 19:37:29 | AlHazWork | signal11: Sure |
| 19:37:33 | AlHazWork | prolly |
| 19:37:49 | AlHazWork | there's 512k of serial data flash we're not doing anything with, as far as we know, too |
| 19:39:27 | signal11 | yeah, if it does dtmf detection there should be a way to get at the raw audio information. |
| 19:40:22 | signal11 | but maybe not; the inferno program couldn't do it |
| 19:40:33 | signal11 | but that might be to the limited amount of flash |
| 19:40:56 | Lenolium | DSP_START_REC, and DSP_STOP_REC Should be pretty handy for us. |
| 19:41:14 | Lenolium | signal11: http://cosmic.censoft.com/dspspec.html |
| 19:43:06 | Lenolium | Since I'm a embedded sytems newbie, how exactly do you talk to the DSP, do you have to do it through a kernel driver? Or can I just access the GPIO pins from userspace? |
| 19:43:52 | signal11 | you could write a driver to let you access the pins |
| 19:43:59 | signal11 | you need a kernel driver |
| 19:44:26 | Lenolium | Ahh.. -k- |
| 19:50:56 | AlHazWork | LPP only greybars half way before MicroWin fires up |
| 19:51:35 | Lenolium | GoRK: So, when am I going to get serial console support in linux with the LPP? |
| 19:53:17 | signal11 | why doesn't serial console support work? |
| 19:53:30 | signal11 | does the default make world include lpp? |
| 19:54:12 | Lenolium | signal11: Don't know, supposidly LPP breaks it. (according to the wiki) |
| 19:57:48 | AlHazWork | my serial console works here |
| 19:58:21 | signal11 | still 900 phones left? |
| 19:58:27 | signal11 | or is the topic out of date |
| 19:58:50 | AlHazWork | i think there's still over 900 |
| 19:58:53 | AlHazWork | *shrug* |
| 19:59:32 | signal11 | i want to get another one just in case something happens to this one. but i want to delay as long as possible, of course |
| 20:00:03 | RedBaron | Yah. I've been thinking the same thing. |
| 20:00:22 | Lenolium | AlHazWork: Does it show bootup messages? |
| 20:00:37 | Lenolium | now just needs to get a 3.3V PCMCIA ethernet card. |
| 20:01:21 | signal11 | the kingston is treating me right |
| 20:04:45 | AlHazWork | lino: they all write over eachother on the bottom line of the screen |
| 20:11:07 | DrXenon | do out-of-the-box tuxscreens work? that is, are they fully-functioning shannons, or are they brain-damaged until you transplant a linux brain into them? |
| 20:11:51 | RedBaron | DrXenon: They have Inferno loaded on. With some custom software to do phone type stuff like store phone numbers and what not. |
| 20:12:27 | RedBaron | DrXenon: So yah...they work. :) but it all depends on what you wanna do with it. |
| 20:12:45 | Lenolium | DrXenon: They are actually very useful out of the box. But, hey, if we can add new features on top of what it can already do, I'm all for it. |
| 20:13:58 | RedBaron | Lenolium: I never used inferno much with my phone. I got the point where it seemed to want me to subscribe and said screw it and installed linux. Did it actualy want you to subscribe and give away personal info? |
| 20:14:32 | DrXenon | thanks for info. was asking merely for information. i ordered mine about 2 weeks ago, i guess, but parcels are taking a long time to get through customs. am suffering from a lack of something to hack on! |
| 20:14:45 | Lenolium | RedBaron: Mine dialed up, but I don't think anything was listening, so I hit cancel, and it went away. |
| 20:15:07 | RedBaron | Lenolium: Ok. |
| 20:20:11 | BZFlag | power is out here. |
| 20:21:22 | pattieja | UPS? |
| 20:21:35 | DrXenon | i think he means DOWNS |
| 20:22:48 | BZFlag | heh, /me is on a laptop with hubs etc on UPS |
| 20:24:46 | Lenolium | I guess 'RequestConfiguration: Bad Vcc' means that the card is 5V? |
| 20:25:15 | BZFlag | or at least asked for 5v first, yeah. |
| 20:25:51 | BZFlag | I can read my mail at home just fine. nothing here in the office though. ;-) |
| 20:26:54 | BZFlag | welcome casey! |
| 20:27:17 | ylekyot | Thx tim. |
| 20:32:03 | as_ | Evnin all. |
| 20:33:54 | as_ | Hiya BZ ;] |
| 20:33:56 | BZFlag | power still down, hubs still up. ;-) |
| 20:34:06 | ProtWork | Heh. :-) |
| 20:35:13 | Lenolium | *pling* |
| 20:35:29 | BZFlag | dns, mail down, web server, fw and ipmasq box still up. ;-) |
| 20:36:14 | signal11 | excellent |
| 20:36:22 | ProtWork | tf is this about? |
| 20:36:37 | signal11 | how long will your ups last? |
| 20:37:01 | BZFlag | don't know. never tested it. when http://www.lineo.com/ goes away, then you will have your answer. ;-) |
| 20:37:19 | BZFlag | Aelhaz is also in my office without power. |
| 20:37:43 | BZFlag | heh. not as much luck though. ;-) |
| 20:38:11 | BZFlag | the net is speedy with everybody else off it. ;-) |
| 20:39:33 | Aelhaz | Woo. |
| 20:47:58 | signal11 | well that answers the question eh |
| 20:48:52 | Lenolium | *plong* |
| 20:49:56 | Lenolium | *pling* |
| 20:50:21 | Lenolium | Well, there goes that idea, I don't think my Dell laptop will do 3.3V cards. |
| 20:50:54 | Lenolium | I was attempting to force my spare ethernet card driver down to 3.3v, to see if it would work. |
| 20:52:21 | signal11 | get a cf card; they are supposed to work at either voltage |
| 20:52:36 | signal11 | this kingston card does; and i've used it in a dell laptop |
| 20:53:30 | Lenolium | Either that, or the PCMCIA stuff can detect which voltage the card really needs. |
| 20:54:23 | Lenolium | Because setting Vcc to 33 returned with: smc91c92_cs: RequestConfiguration: Bad Vcc |
| 21:06:24 | SpaceCoaster | this DSP is the meanest chip ever, to read a byte you have to write a byte first, nothing for nothin |
| 21:07:35 | pattieja | is away: bible study + dinner (mmmm :) |
| 21:08:52 | SpaceCoaster | download kernel... |
| 21:10:03 | SpaceCoaster | flash kernel...boot kernel... if I've got i wrong my DSP flash is toast |
| 21:11:09 | SpaceCoaster | hmmm...nothing much happened, it hangs :-(, time for printk's |
| 21:14:26 | Aelhaz | why would your dsp flash be toast? |
| 21:15:40 | SpaceCoaster | well, when writing to the DSP you can either write the registers or the flash, both lines are asserted and you drop the line on the one you want to write to |
| 21:15:54 | Aelhaz | hmm |
| 21:16:06 | Lenolium | SpaceCoaster: So you're working on getting the DSP working? Great! |
| 21:16:21 | Aelhaz | SpaceCoaster: maybe you can physically protect the flash |
| 21:16:30 | SpaceCoaster | I guess writing to flash is more tricky so I probably ok, just chatting |
| 21:16:33 | Aelhaz | often you can power or clip a pin and make flash read-only |
| 21:17:05 | SpaceCoaster | I feel brave today :-) (or maybe foolish). |
| 21:17:57 | Aelhaz | they tell me the power came back on at work shortly after i left |
| 21:17:58 | Aelhaz | figures |
| 21:18:14 | SpaceCoaster | Lenolium: only trying to talk to it and flash the lights, i think it will be a while before any progress is made |
| 21:18:31 | Lenolium | SpaceCoaster: Yeah, I know, All I want is the HAPPY_SOUND |
| 21:19:02 | Lenolium | HAPPY_TONE, sorry. |
| 21:19:26 | Aelhaz | yes, HAPPY_TONE is critical |
| 21:19:44 | SpaceCoaster | who is Tone and why is he so happy? |
| 21:19:54 | Lenolium | echo 1 > /dev/tones/happy |
| 21:20:40 | SpaceCoaster | Good interface |
| 21:23:51 | SpaceCoaster | Aelhaz: I was going for flashing lights but tones are much more DSPish, it's a design change! |
| 21:27:14 | Aelhaz | hehe |
| 21:27:47 | ProtWork | Heh. Any raw specs on Wheaties? |
| 21:28:15 | SpaceCoaster | wot are the odds of installing a new, very experimental, kernel and my ethernet card stopping working at the same time, 100% :-( |
| 21:31:37 | SpaceCoaster | I think the DSP part is http://tuxscreen.net/download/dsp1650.pdf |
| 21:32:06 | Tangent | ProtWork : I just couldn't resist... http://www.wheatiesenergycrunch.com/wec/intro/energycombo/5.jpg |
| 21:33:33 | ProtWork | *lol* |
| 21:33:35 | SpaceCoaster | tee hee |
| 21:33:38 | ProtWork | Tangent: cool. :-) |
| 21:34:01 | ProtWork | Also, do we have a complete set of schematics for Shannon besides the scans on tuxscreen.net? |
| 21:34:08 | ProtWork | If not, I could re-capture them. |
| 21:37:44 | SpaceCoaster | ProtWork: I think someone (GPSFan or RedBaron maybe) is doing this, they would probably appreciate some help/verification. I have seen PCMCIA PDFs that were produced this way. |
| 21:39:15 | SpaceCoaster | ProtWork: yeah it is GPSFan, good stuff |
| 21:39:26 | RedBaron | SpaceCoaster: Whatcha talking about? |
| 21:39:39 | RedBaron | SpaceCoaster: The 5v pcmcia stuff? Thats all gpsfan. Not me |
| 21:40:23 | SpaceCoaster | RedBaron: yeah i said maybe, vague memories, I checked my logs and corrected myself |
| 21:40:44 | RedBaron | SpaceCoaster: *nod* No problem. I just got back and wasnt sure what you were talking about. :) |
| 21:40:59 | SpaceCoaster | RedBaron: me neither :-) |
| 21:41:30 | RedBaron | SpaceCoaster: *grin* |
| 21:41:45 | Aelhaz | BZFlag: It figures that as soon as i'd given up and gone home the power would come back on |
| 21:41:57 | SpaceCoaster | RedBaron: I was built without error correction, mostly random data... |
| 21:42:26 | SpaceCoaster | RedBaron: how goes the mobile division? |
| 21:43:04 | RedBaron | SpaceCoaster: Whelp...I've been off being interfered by RL for the last couple of days. But I just got all the parts to do the power supply and will hack it up after dinner |
| 21:43:20 | RedBaron | SpaceCoaster: And hopefully not fry my tux. :) |
| 21:43:58 | SpaceCoaster | RedBaron: "interfered by RL" oooh err missus |
| 21:44:16 | BZFlag | Aelhaz: heh. well eforge took forever to come back... |
| 21:45:14 | Aelhaz | BZFlag: I've got to wait for stuart to give me documentation before i can go much further with what I'm testing, anyway |
| 21:45:34 | Aelhaz | that guy just *loves to include stuff that doesn't show up in the gui that's critical to actually getting a board running . . . |
| 21:45:48 | Aelhaz | "oh, you just go to commandline and type 'foo'" |
| 21:46:18 | RedBaron | Aelhaz: GUI? whats this that you speak of? *grin* |
| 21:47:25 | RedBaron | SpaceCoaster: Yah...real life...gets in the way of the fun stuff. |
| 21:48:14 | RedBaron | SpaceCoaster: I got the voltage regulators that I need, the supporting capictors for them..and I borrowed a battery from work for the power source. And we'll see what happens. |
| 21:48:38 | Aelhaz | RedBaron: My job is to test the Embedix SDK, which is like buildroot but, bigger, and gooier . . . |
| 21:49:06 | SpaceCoaster | RedBaron: why are you building a powersupply? |
| 21:49:56 | RedBaron | Aelhaz: Ah. I have only looked at buildroot with microwin so far. Whats embedix use for gui stuff? |
| 21:50:24 | Aelhaz | RedBaron: it's a QT based gui with a big tree-view of packages, options, files, etc |
| 21:50:45 | Aelhaz | point/click compile time configuration for an entire embedded linux setup |
| 21:50:59 | Aelhaz | w/ automated builder |
| 21:51:52 | RedBaron | SpaceCoaster: I gotta have some electronics to regulate the voltage from the batteries. The Tux needs +5V and +3.3V. Its kind of hard to find batteries that give that voltage, and the voltage from batteries float around a lot. So, thats what I mean by power supply |
| 21:52:14 | Aelhaz | I'm working on a board support kit written by a developer in the UK who likes to leave things implied rather than explicitly stated . . . |
| 21:52:16 | RedBaron | Aelhaz: I see. The QT/E? |
| 21:52:26 | Aelhaz | RedBaron: oh, no, this is a full blown X11 app |
| 21:53:00 | Aelhaz | requires gobs of disk space and ram to run |
| 21:55:38 | RedBaron | Aelhaz: Okay. |
| 21:57:19 | Aelhaz | builds small linux installations on demand |
| 22:01:43 | RedBaron | of to eat dinner |
| 22:01:55 | RedBaron | hmmm...'off' as well. |
| 22:42:56 | Lenolium | SpaceCoaster: Uhm, do we have happy tone yet? :) |
| 22:45:22 | SpaceCoaster | I do but my tuxscreen doesn't |
| 23:00:03 | SpaceCoaster | Lenolium: wanna talk DSP |
| 23:00:11 | SpaceCoaster | Lenolium: ? |
| 23:00:12 | GoRK | dsp! |
| 23:00:21 | GoRK | i want to talk about the dsp |
| 23:00:38 | GoRK | i have happy tones ringing in my brain |
| 23:00:53 | GoRK | there are sad tones too |
| 23:01:15 | GoRK | i am scared that those functions listed under unknown/unused will not even work and we'll not be able to do audio through the dsp |
| 23:01:25 | SpaceCoaster | GoRK: great! Sorry but I'm a flashy LED sort of guy. |
| 23:01:59 | SpaceCoaster | do you know if the real time clock chip hangs off the DSP? |
| 23:02:30 | Aelhaz | think happy tones . . |
| 23:02:48 | GoRK | it seems the DSP can control the clock |
| 23:03:10 | GoRK | maybe there are two clocks. one that the dsp uses and one that the system uses |
| 23:03:23 | SpaceCoaster | do you see this on a schematic or something? |
| 23:03:43 | GoRK | no but i see that the DSP has commands to set the time |
| 23:03:57 | GoRK | and the time is also settable from the processor |
| 23:04:14 | SpaceCoaster | the DSP seems to restore the time from a clock, maybe it's a real clock chip with a battery and not just a timer interrupt thingy |
| 23:05:23 | SpaceCoaster | anyone know hardware and which chips are clock chips, i think they are pretty well known/standard? |
| 23:06:44 | SpaceCoaster | GoRK: the interface to the DSP is really hokey, 1 bit, software handshake, delays arrrgggh |
| 23:07:09 | GoRK | it's only on 1 gpio pin? |
| 23:08:33 | SpaceCoaster | well several INPUT/OUTPUT/RTS/CTS/CLOCK, but they must be twiddled by software a bit at a time. |
| 23:09:15 | GoRK | heh like talking to a PS/2 mouse |
| 23:11:08 | SpaceCoaster | I see why they had to modify the hardware to transfer audio from the DSP, audio aint coming across the SPI (the name of this serial link). |
| 23:12:31 | GoRK | supposedly they only modified the dsp firmware |
| 23:13:11 | Aelhaz | hmm, but the sa1100 and the ucb1200 together should be capable of audio, right? |
| 23:13:21 | SpaceCoaster | GoRK: didn't scotty talk to a mouse in Star Trek 22 the movie, like that eh? |
| 23:14:12 | SpaceCoaster | KRoG: fun eh? |
| 23:14:18 | KRoG | yeah |
| 23:14:36 | KRoG | KRoG is cooler than retsaoCecapS |
| 23:16:02 | SpaceCoaster | true, retsaoCecapS sounds like a medicine for the elderly! |
| 23:16:43 | SpaceCoaster | Aelhaz: Does the speaker/mike/telephone line hang off the DSP? |
| 23:17:54 | GoRK | ooh i know that one. Yes! |
| 23:18:43 | SpaceCoaster | k but the UCB1200 seems to talk to the modem, and not to the DSP. |
| 23:20:03 | GoRK | maybe that is the modification they were talking about... perhaps lucent reworked the dsp to wire it up to the ucb1200 |
| 23:20:08 | SpaceCoaster | the DSP doesn't seem to be connected to the rest of the hardware except via the SPI. |
| 23:21:16 | SpaceCoaster | probably, don't throw away your answering machine just yet |
| 23:22:02 | SpaceCoaster | Aelhaz: what would the DSP be doing for us? |
| 23:22:49 | SpaceCoaster | GoRK: do you know? |
| 23:24:00 | GoRK | well the dsp does stuff like generate DTMF and route signals in between speakerphone, etc. etc... it makes the phone part work.. supposedly the DSP also *is* the modem.. kind of like a software modem done in firmware instead |
| 23:24:14 | Aelhaz | hmm |
| 23:24:43 | Aelhaz | GoRK: well yeah, USRobotics modems used to all be DSP based |
| 23:25:17 | SpaceCoaster | hmm, indeed if it's the modem then it probably can talk to the UCB1200 directly but I bet it don't send audio. |
| 23:26:43 | GoRK | that wouldnt matter if you put a CODEC in the firmware. then it can talk as a modem only the data is compressed audio |
| 23:27:21 | SpaceCoaster | well the UCB1200 is a codec, would that help? |
| 23:27:37 | GoRK | i dunno |
| 23:28:16 | GoRK | takes apart a phone |
| 23:28:23 | SpaceCoaster | me neither, I was hoping for a hardware guru to provide a free tutorial :-( |
| 23:28:28 | SpaceCoaster | hehe |
| 23:29:41 | GoRK | i am not that man |
| 23:30:39 | ProtWork | Hardware guru for what? |
| 23:30:39 | SpaceCoaster | but thanks in any case, wanted a break from bit fiddling. |
| 23:30:47 | SpaceCoaster | ahaha |
| 23:31:16 | SpaceCoaster | can you scroll back and read the previous stuff? |
| 23:31:54 | ProtWork | A bit - interested in the SPI communications for the DSP? |
| 23:32:36 | SpaceCoaster | yeah, I think I have that understood, it's gross but workable. |
| 23:33:21 | SpaceCoaster | The software loads the time via the DSP and it appears to be a real time clock chip, do you know if we have one? |
| 23:34:39 | ProtWork | Checking. |
| 23:38:56 | ProtWork | I think the DSP pulls the time from the phone line. |
| 23:39:18 | SpaceCoaster | cool |
| 23:39:23 | ProtWork | Via CID. |
| 23:39:32 | ProtWork | That's the best I can get you right now. |
| 23:40:35 | SpaceCoaster | thanks, can I ask another question/ |
| 23:40:37 | SpaceCoaster | ? |
| 23:40:42 | ProtWork | Of course. |
| 23:41:22 | SpaceCoaster | we think the DSP does the modem stuff and connects to the UCB1200, the codec, what does a codec do? |
| 23:43:07 | ProtWork | A Codec is a fancy name for an integrated A/D and D/A converter - converts analogue signals into digital and vice-versa. |
| 23:44:56 | ProtWork | I need to get the spec for the UCB1200, and I can give you more info on it. |
| 23:46:29 | SpaceCoaster | k, it makes sense |
| 23:48:06 | SpaceCoaster | so the DSP takes the phone line modem whistles, processes it digitally, converts it to analogue again and sends it to the codec which converts it to digital again! |
| 23:48:43 | ProtWork | *nod* |
| 23:49:36 | SpaceCoaster | hmm, so unless the DSP can munge voice type audio and send that to the codec we can't get audio into RAM for VOIP, answering machines etc. |
| 23:53:28 | SpaceCoaster | ProtWork: ? |
| 23:53:52 | ProtWork | SpaceCoaster: Hangon. Checking. There may be a passthru function. This is why that Lucent paper was talking about munging up DSP code for VoIP. |
| 23:56:19 | GoRK | there is a JTAG port on the DSP daughtercard |
| 23:57:04 | SpaceCoaster | JTAG is my friend, it rescues broken TuxScreens! |
| 23:57:42 | GoRK | heh it may also let us reprogram the DSP |
| 23:57:55 | GoRK | or get the code off of it |
