| 00:05:35 | Tangent | gets climate control working in his office.... sweet :D |
| 00:07:24 | GPSFan | Tangent: Hi, I found a ps > pdf converter, and was able to get the schematics looking better in both formats, there is now a wiki page http://tuxscreen.net/download/schematics/ |
| 00:07:37 | Tangent | GPSFan : Excellent :) |
| 00:08:18 | Tangent | GPSFan : Looks like you've done the whole Tux now :) |
| 00:08:41 | GPSFan | Oops the wiki page is: http://tuxscreen.net/wiki/view//ScheMatics and the description is at http://tuxscreen.net/wiki/view//KenMcguire |
| 00:09:28 | GPSFan | Not quite yet, but the CPU and associated logic (pcmcia,dram, flash) is at about 85% |
| 00:10:54 | GPSFan | Tangent: next project is a bit of cleanup, and work on the reset circuit. Then on to 5V PCMCIA. |
| 00:11:13 | GPSFan | Tangent: BTW I got my PCMCIA 2.1 spec. |
| 00:11:13 | Tangent | GPSFan : Excellent work :) |
| 00:11:20 | Tangent | GPSFan : Oh well done |
| 00:11:51 | GPSFan | Tangent: if you need any excerpts, I can qoute it for you. |
| 00:12:24 | Tangent | GPSFan : When I get around to it, I may well require bits of the spec |
| 00:13:46 | Tangent | Heh.. I just went to check on the temperature in my machine room... It's so so cold now :) |
| 00:18:23 | Tangent | That means that soon, one of my Tux's can control it all |
| 00:18:40 | Tangent | In fact, if minicom works, then it can already control it all :) |
| 00:40:56 | Tangent | watches the clock tick... |
| 00:46:21 | Sammy | morning all... |
| 00:46:37 | scanline | good morning |
| 00:46:50 | Tangent | Sammy : morning |
| 01:14:12 | GPSFan | SpaceCoaster: you still awake?? |
| 01:14:22 | SpaceCoaster | GPSFan: indeed |
| 01:15:01 | GPSFan | SpaceCoaster: check out http://tuxscreen.net/download/Mongoose.jpg |
| 01:15:27 | GPSFan | SpaceCoaster: I just pulled the dsp board off and it was a bit of a supprise. |
| 01:16:11 | SpaceCoaster | what are all those empty pads? |
| 01:16:26 | GPSFan | Hmmm... indeed! |
| 01:17:06 | SpaceCoaster | did you look at ic902/3 on the cave_drawings? |
| 01:17:39 | GPSFan | I think they were the flash or the DRAM the drawing was too fuzzy to read the PN of the chip. |
| 01:18:12 | Tangent | Adding some extra flash would be nice |
| 01:19:37 | GPSFan | Section 2.2 of the dsp doc's says that the 1633 is a masked ROM device. |
| 01:20:03 | SpaceCoaster | well that puts paid to reprogramming it! |
| 01:20:38 | GPSFan | Guess so, I guess Lucent was real confident of their DSP code. |
| 01:23:15 | SpaceCoaster | I think it can be made to boot from other memory so it might not be the end of it. I suspect thought that we can work with what we have. |
| 01:25:05 | SpaceCoaster | GPSFan: are you ever going to put your tuxscreen back together? |
| 01:26:53 | SpaceCoaster | GPSFan: I guess that they have set the secure bit, after all they didn't have a need to read it. |
| 01:27:19 | SpaceCoaster | GPSFan: pity I was looking forward to speaking DSP :-( |
| 01:28:59 | GPSFan | Well, there are still the built in functions, one could also populate the chips and replace the 1633 with a 1650 ;>) |
| 01:29:37 | SpaceCoaster | you could! |
| 01:29:53 | GPSFan | I guess that would be their development system, then you need compilers and an education is DSP programming. |
| 01:31:09 | SpaceCoaster | I saw a good interesting book "practical DSP programming" difficult to justify for an unprogrammable DSP :-( |
| 01:34:52 | GPSFan | SpaceCoaster: I'll get it back together when I'm done with the schematics. The CPU board might make a nice embedded platform, All the usefull I/O comes outr J600, 601. |
| 01:35:33 | SpaceCoaster | GPSFan: the DSP manual mentions that it uses Motorola SPI for it's serial interface, this is also one of the modes supported by the SSP on the SA-1100, I wonder why they didn't use that, instead of bit twiddling. |
| 01:36:26 | SpaceCoaster | GPSFan: I want to do the same as RedBaron and assemble a mobile unit. |
| 01:37:49 | Alehaz | GPSFan: what about that unpopulated header? what's it hooked up to? |
| 01:38:02 | GPSFan | SpaceCoaster what about a display? |
| 01:38:49 | GPSFan | Alehaz: still on the list to ring out, I have several pins located but I won't get around to it till Mon, or Tues. |
| 01:39:39 | SpaceCoaster | GPSFan: I was hoping to take the boards out add batteries and glue it onto the back of the display that it has already. |
| 01:40:12 | Alehaz | GPSFan: ok, thanks |
| 01:40:38 | GPSFan | SpaceCoaster: you might have a problem seeing it in anything brighter than normal room light. |
| 01:41:10 | GPSFan | Alehaz: check the wiki, I post a message when it's up. |
| 01:42:26 | Alehaz | ok |
| 01:42:42 | SpaceCoaster | It's not as bad as it looked with inferno, the brightness/contrast tweaks a lot and colors help enormously. |
| 01:43:17 | SpaceCoaster | It was also mainly for in-house use, wireless surfing in bed! |
| 01:43:58 | GPSFan | That's good, but sunlight readable displays are somewhat of an art, but for bedroom surfing it was made to order! |
| 01:44:35 | SpaceCoaster | how much current do you guess it would use? |
| 01:44:54 | GPSFan | A sunlight readable display or a bedroom surfer appliance? |
| 01:45:14 | SpaceCoaster | bedroom surfer |
| 01:45:35 | Alehaz | I've got some sunlight readable displays . . . they're both greyscale tho |
| 01:46:54 | GPSFan | I believe that RedBaron took some pwr measurements without the display and was supprised at how little it drew. I was refering to full color displays, 30W+ for backlight, although ther are some reflective displays comming out of the lab soon. |
| 01:47:49 | GPSFan | Look ath the color iPaq it is reflective with a front light for nighttime operation. |
| 01:48:16 | SpaceCoaster | ahh that is what the frontlight stuff is in the driver directory |
| 01:49:22 | SpaceCoaster | as an aside the Motorola SPI documented in the SA-1100 spec, page 321 is exactly what I programmed in the wheaties driver |
| 01:49:54 | GPSFan | yep, the iPaq display is purely reflective, so there is nowhere to shine a backlight through, but they do have a front light. |
| 01:50:40 | GPSFan | Did you bit twiddle in wheaties.c? |
| 01:50:48 | SpaceCoaster | GPSFan: clearer all the time, thanks, I guess reflective is even better when powered by sunshine. |
| 01:51:13 | SpaceCoaster | bit twiddling like crazy |
| 01:52:08 | GPSFan | I've seen some Sharp, Kyocera, and Hosiden reflective & transflective displays, the active kind are great in the sun, the more sun the better. However they suffer from the user shadow effect. |
| 01:53:07 | SpaceCoaster | hehe |
| 01:53:18 | GPSFan | be back in a bit... |
| 02:00:31 | GPSFan | back, If you have the power, 30 watts (and the right AR coatings) will make a SVGA laptop type display visible in full sun. |
| 02:01:27 | Alehaz | oh, I also have an EL display that's not doing anything. |
| 02:01:43 | Alehaz | it's real pretty if you don't mind nothing but one brightness of amber |
| 02:03:24 | GPSFan | the first laptop had an EL display, from Sharp. |
| 02:04:00 | Alehaz | this EL display is a sharp, fwiw |
| 02:04:08 | Alehaz | EGA rez |
| 02:05:03 | Alehaz | I've been considering hooking it up to the ucsimm, for, something |
| 02:05:29 | Alehaz | there's an infrared touchscreen bolted to it, too |
| 02:06:03 | GPSFan | I drove an EL EGA display right off a n EGA controller, only had to adjust the timing a bit to make it work. |
| 02:06:20 | Alehaz | neat |
| 02:06:56 | Alehaz | i think this one's designed to be driven by a monochrome lcd controller, but I haven't read the specs very deeply |
| 02:07:25 | Alehaz | it's an old industrial touchscreen computer. aluminum box with a single board 386sx in it. |
| 02:07:48 | Alehaz | allegedly waterproof, tho it's almost definately not waterproof now |
| 02:07:55 | Alehaz | well, no, it's definately not |
| 02:08:16 | GPSFan | the SA1100 could drive it, the SA1100 is quite flexible in what sort of displays it can drive. |
| 02:09:00 | Alehaz | oh yeah, I'm sure |
| 02:09:24 | Alehaz | I have a couple ucsimms that work gave me, too, tho, which have no display at all |
| 02:10:01 | Alehaz | but should be able to drive it |
| 02:10:26 | Alehaz | but it's way down at the bottom of the list of things to fiddle with |
| 02:11:23 | GPSFan | uCsimms are cool, though a bit overpriced. |
| 02:12:11 | Alehaz | yeah |
| 02:12:38 | GPSFan | roughly the same price as a cerfcube, for a LOT less capability. |
| 02:12:45 | Alehaz | especially considering it doesn't do all that much |
| 02:12:51 | Alehaz | how much is the cerfcube now? |
| 02:14:00 | GPSFan | I'd like to use one for a remote telemetry system, I have a wireless connection to the Internet, and our remote sites need monitoring, they have ethernet connectivity to the main sight via wireless. |
| 02:15:08 | GPSFan | http://www.intrinsyc.com/Merchant2/merchant.mv?Screen=SFNT&Store_Code=01 $379US |
| 02:18:49 | Alehaz | they never seem to get cheaper than about that much tho |
| 02:19:37 | Alehaz | I've heard people say that the rule of thumb is that you sell hardware for 10x what it costs to make . . . |
| 02:20:20 | GPSFan | I wish I could buy some stuff for only 10x parts costs. |
| 02:20:37 | Alehaz | heh |
| 02:20:56 | GPSFan | well, I'm off to ZZzzz land night all. |
| 02:21:09 | Alehaz | a lot of the problem has to be the low volume |
| 02:21:29 | SpaceCoaster | see ya GPSFan |
| 02:21:45 | GPSFan | in many cases low volume is the only problem, bye.. |
| 02:22:27 | Alehaz | cya |
| 02:35:45 | GoRK | SpaceCoaster: does wheaties.c work with devfs? |
| 02:36:31 | SpaceCoaster | no but it could, I use a major number (60) reserved for experimental drivers |
| 02:36:57 | GoRK | just wondering... i dont intend to use devfs, but still curious |
| 02:38:12 | SpaceCoaster | yeah, I think it basically involves not setting the major number, i.e. setting it to 0. |
| 02:38:54 | SpaceCoaster | I have reduced the interbit delay from 5 to 1us and it still works! |
| 02:39:18 | GoRK | is it reliable? |
| 02:40:19 | SpaceCoaster | doesn't break so far, the interface is self clocked and should run at upto 2mhz so 0.5mhz should be ok (that is 2x1us). |
| 02:41:26 | SpaceCoaster | Bill Danielson might be able to scope it and tell us how close we are to breaking stuff. |
| 02:45:21 | SpaceCoaster | the transfer to/from the 512kB flash seems to use the SSP mode directly, the flash memory supports motorola SPI as does the SA-1100 and the DSP. |
| 02:45:53 | GoRK | cool so it may be easier to access it in software than we thought? |
| 02:46:27 | GoRK | another idea i had was that if the dsp is able to play audio data from the flash, it may not be possible to lay a filesystem on top of it and preserve that capability |
| 02:47:41 | SpaceCoaster | true, but a filesystem view would still be neat, maybe construct the filesystem in ram and have the data in flash! |
| 02:47:57 | GoRK | either way |
| 02:48:31 | SpaceCoaster | I just like the idea of the filesystem for getting at storage. |
| 02:49:41 | GoRK | is there a filesystem that is more suitable for a serial flash device? e.g. fixed size file index at the beginning of the flash, all files must be accessible linearly (e.g. no fragmentation) ??? |
| 02:52:23 | SpaceCoaster | hmmm |
| 02:53:43 | SpaceCoaster | do you know if the DSP can read from this flash directly? |
| 02:53:58 | Tangent | Something like tar, or ar, tape backup formats... |
| 02:54:15 | GoRK | SpaceCoaster: i do not :) |
| 02:54:34 | GoRK | SpaceCoaster: hunch |
| 02:55:41 | GoRK | im flashing a kernel with the ioctl. i am ready to go |
| 02:55:54 | SpaceCoaster | cool |
| 03:03:38 | SpaceCoaster | GoRK: I don't see anything for connecting the DSP to the flash. The both talk to the SA-1100 on the same pins with two GPIOs acting as enables. |
| 03:09:36 | Lenolium | *pling* |
| 03:10:34 | SpaceCoaster | GoRK: I'm going to bed in a moment, just a closing thought, the throughput should be fine going from flash-->sa1100-->dsp even perhaps coming out of the kernel into userspace so a jffs filesystem might work. |
| 03:10:59 | SpaceCoaster | mail me your thoughts |
| 03:12:04 | Lenolium | GoRK: So, has wheatiesd.c been abused yet? |
| 03:20:03 | GoRK | nope not yet |
| 03:20:07 | GoRK | i was about to do that |
| 03:21:44 | GoRK | or start one |
| 03:41:11 | Lenolium | notices how his comments got a few comments out of you guys. "poor man's reset" |
| 03:41:54 | GoRK | you shoulda done the software reset... it's just as good as the hardware one |
| 03:41:59 | GoRK | almost |
| 03:42:35 | Lenolium | GoRK: Well, the good news is that SpaceCoaster decided to add an ioctl() to hardware reset it. |
| 03:43:04 | GoRK | yeah i suggested that to him and he put it in there |
| 04:16:10 | StarKruzr | Hey all |
| 04:16:16 | GoRK | hi |
| 04:16:43 | StarKruzr | Anyone have any information as to how many units Riker has left? |
| 04:16:55 | StarKruzr | I know he got /.ed so I'm afraid he may suddenly be out of stock |
| 04:17:07 | GoRK | like almost 900 if not more |
| 04:17:13 | scanline | Last time BZflag set the topic it was 'more than 800 left' |
| 04:17:41 | StarKruzr | okay |
| 04:17:41 | StarKruzr | cool |
| 04:17:47 | StarKruzr | Do you guys all have one of these? |
| 04:17:53 | GoRK | i have a couple, yes |
| 04:18:04 | StarKruzr | :) Are they as cool as they sound? |
| 04:18:08 | scanline | yes |
| 04:19:15 | scanline | They run Linux nicely, the hardware is very hackable |
| 04:19:21 | StarKruzr | Are there more apps available than what are listed there on the page? |
| 04:19:27 | StarKruzr | I imagine |
| 04:19:31 | StarKruzr | no disk access time |
| 04:19:37 | StarKruzr | or on the order of ns anyway |
| 04:19:38 | scanline | But there aren't any useful apps yet |
| 04:19:40 | GoRK | StarKruzr: there are pretty much no apps at all yet |
| 04:19:54 | StarKruzr | Hm |
| 04:20:07 | scanline | No web browser yet, but that should come soon |
| 04:20:07 | StarKruzr | Except for the stuff listed there |
| 04:20:11 | StarKruzr | the voicemail and so on |
| 04:20:16 | GoRK | StarKruzr: and actually flash ram is hundreds of times slower than a hdd |
| 04:20:20 | scanline | Microwindows' browser isn't working, and PicoGUI's browser is still under development |
| 04:20:25 | StarKruzr | GoRK: You're kidding! |
| 04:20:29 | GoRK | StarKruzr: if you read that it says those applications are possible |
| 04:20:31 | StarKruzr | had no idea |
| 04:20:43 | GoRK | StarKruzr: none of the exist yet |
| 04:21:13 | Lenolium | StarKruzr: Currently, under linux, you can get it to beep at you once every second and a half. :) Oh, and it'll run a shell. |
| 04:21:23 | StarKruzr | Lenolium: WHEE |
| 04:21:29 | Lenolium | (It will actually run X, if you can get yourself a NFS mounted root) |
| 04:21:46 | Lenolium | StarKruzr: But, fortunately, it does come with inferno and some pretty useful apps pre-installed. |
| 04:21:51 | GoRK | it wil l run debian with nfsroot :) |
| 04:21:54 | StarKruzr | Lenolium: Yeah |
| 04:21:59 | StarKruzr | Wish I could afford two. |
| 04:22:03 | Lenolium | They are just of beta quality at best. |
| 04:22:10 | StarKruzr | I'd buy one to use Inferno with, and one to dev on. |
| 04:22:32 | StarKruzr | Are there emus available I can use to dev with? |
| 04:22:51 | scanline | It's pretty much like any other Linux system except for the DSP |
| 04:22:55 | Lenolium | StarKruzr: That's what I have done, I'm probably gonna try to grab four more, to make sure I have enough for everyplace I want one in my house, so after I hack them to bits, I'll be happy. |
| 04:23:11 | scanline | And the DSP development is still in it's early stages |
| 04:23:29 | Lenolium | StarKruzr: Seems like most of the app development is going to be running under PicoGUI, which will run natively under linux |
| 04:23:48 | StarKruzr | Lenolium: That sounds like a great plan, but as of right now I'm trying to find a decent job till I get commissioned in the Air Force |
| 04:23:59 | StarKruzr | <-- has around $200 to his name ;) |
| 04:24:28 | StarKruzr | PicoGUI is one of those things that writes straight to /dev/framebuffer IIRC |
| 04:24:30 | scanline | StarKruzr: You'll want to save about $50 of that to get a good PCMCIA network card |
| 04:24:45 | StarKruzr | scanline: Thankfully, I have one |
| 04:25:04 | scanline | StarKruzr: it might not work on the TuxScreen. currently it only supports 3.3v cards |
| 04:25:11 | StarKruzr | ohhh |
| 04:25:14 | StarKruzr | shood |
| 04:25:15 | StarKruzr | er |
| 04:25:16 | StarKruzr | shoot |
| 04:25:19 | StarKruzr | need CardBus then |
| 04:25:33 | StarKruzr | IIRC CardBus is 3.3 and regular PCMCIA is 5.0 or something |
| 04:25:36 | GoRK | no. cardbus is 32 bit. those dont work either |
| 04:25:43 | StarKruzr | ah. |
| 04:25:43 | GoRK | you need 16 bit 3.3v |
| 04:25:54 | GoRK | iirc most cardbus cards are 5v anyway |
| 04:25:56 | StarKruzr | I didn't know those existed. |
| 04:26:05 | StarKruzr | huh |
| 04:26:06 | scanline | StarKruzr: it's hard to tell what cards will work without testing them |
| 04:26:09 | StarKruzr | Recommend a good one? |
| 04:26:20 | scanline | I'm very happy with the netgear FA411 |
| 04:26:31 | scanline | It's about $45 |
| 04:26:37 | collord | StarKruzr: some PCMCIA cards say they need cardbus, just because they know they can get 3.3v out of them... |
| 04:26:54 | StarKruzr | collord: Gotcha |
| 04:27:04 | StarKruzr | collord: But they're actually 16-bit |
| 04:28:15 | scanline | StarKruzr: PicoGUI usually uses /dev/fb0 on embedded systems, but it has many options for video output including a windowed mode in X11 |
| 04:28:51 | StarKruzr | okay |
| 04:28:56 | StarKruzr | ew, X |
| 04:29:18 | StarKruzr | wishes X would quietly go away and be replaced by something faster |
| 04:29:50 | scanline | Well, actually it doesn't use X directly, it uses SDL |
| 04:29:58 | scanline | So anything that SDL supports will work |
| 04:30:15 | StarKruzr | okay |
| 04:30:24 | collord | like hascii |
| 04:30:27 | collord | ;) |
| 04:30:59 | StarKruzr | this thing is going to be a lot of fun to hack once I actually get one |
| 04:32:06 | scanline | StarKruzr: Actually, PicoGUI has an ncurses driver that can produce decent-looking text GUIs :) |
| 04:34:17 | StarKruzr | Hm |
| 04:34:24 | StarKruzr | So we should be able to do that now, no? |
| 04:34:45 | StarKruzr | I seem to recall someone saying something about graphical interfaces choking but text being OK |
| 04:34:55 | StarKruzr | (so far) |
| 04:34:57 | scanline | on the tux? |
| 04:35:03 | StarKruzr | Yeah |
| 04:35:12 | scanline | So far the tux will run microwindows, picogui, and X nicely |
| 04:36:32 | StarKruzr | Oh! |
| 04:36:38 | StarKruzr | w00ts |
| 04:37:10 | scanline | There's a page on the Wiki with info on the GUIs supported so faar |
| 04:41:16 | Lenolium | StarKruzr: Basically, with a NFS-root, it makes a very small linux machine with a touchscreen. And it has a pretty non-functional set of phone equipment. As soon as we get around to really hacking on the DSP, then it'll be fully functional. |
| 04:46:04 | StarKruzr | cool |
| 04:46:10 | StarKruzr | How much DSP work has been done so far? |
| 04:48:28 | scanline | We can send raw commands to the DSP and make it beep, flash the LEDs, and dial the phone |
| 04:48:45 | scanline | I think that there's work in progress on a daemon for DSP interaction |
| 04:54:04 | Lenolium | scanline: Yup. It's a work in progress, unfortunately, I have other work I'm getting paid to hack on tonight... :( |
| 04:54:24 | scanline | hehe |
| 04:57:01 | StarKruzr | :) |
| 04:57:09 | GoRK | heh this is going nicely over here |
| 04:57:19 | StarKruzr | I wonder how much a production unit would cost if it actually really took off |
| 04:57:25 | scanline | If I weren't doing math homework now, I'd be debugging PicoGUI |
| 04:57:31 | Lenolium | StarKruzr: Around $350 |
| 04:57:33 | GoRK | the original target price was like 650 |
| 04:57:37 | StarKruzr | GoRK: Yeah |
| 04:57:51 | GoRK | err they wanted them for 350 but by the time they made these the were looking at 650 |
| 04:57:55 | GoRK | which would not have gone over |
| 04:57:57 | StarKruzr | I wish I could find it in me to just randomly code like that. I envy your talent, scanline. |
| 04:58:38 | scanline | programming is fun :) |
| 04:58:49 | StarKruzr | I used to find it so |
| 04:59:03 | StarKruzr | I sort of still do, but I'm a horrible debugger and an even worse algorithm designer |
| 04:59:18 | StarKruzr | I guess I'm just out of practice |
| 05:01:26 | StarKruzr | I think I'm destined for management. ::barf:: |
| 05:02:29 | scanline | just try not to be a PHB ;) |
| 05:05:13 | scanline | (actually, file on #picogui called me an 'annoying boss'...) |
| 05:24:06 | StarKruzr | hm |
| 05:24:11 | StarKruzr | Is there a PicoGUI website? |
| 05:24:33 | scanline | http://picogui.org |
| 05:24:37 | StarKruzr | cool |
| 05:26:36 | StarKruzr | wow |
| 05:26:40 | StarKruzr | this is really nice-looking |
| 05:26:46 | scanline | I need to post some more recent screenshots |
| 05:26:55 | scanline | yeah, the theme system is kinda nifty |
| 05:27:20 | scanline | curses at differential equations... |
| 05:28:13 | Lenolium | Ugh.. |
| 05:28:19 | Lenolium | I like integrals way better. |
| 05:28:30 | scanline | yeah |
| 05:29:18 | StarKruzr | this reminds me of QNX, somewhat |
| 05:29:56 | scanline | deja vu |
| 05:30:58 | scanline | a comparison between QNX and PicoGUI came up in this channel a few days ago also |
| 05:31:01 | StarKruzr | You've JUST started college? Where? |
| 05:31:14 | StarKruzr | oh, colorado |
| 05:31:44 | StarKruzr | You might know my friend Jon Eggers? |
| 05:31:47 | scanline | yep. university of colorado at boulder |
| 05:31:52 | scanline | no |
| 05:31:55 | StarKruzr | ah well |
| 05:31:56 | scanline | it's a big school |
| 05:31:58 | StarKruzr | wild shot |
| 05:32:10 | scanline | it's always worth a try though |
| 05:32:15 | StarKruzr | true |
| 05:32:25 | StarKruzr | another friend of mine, Deidre Brucker, is a TA there |
| 05:32:44 | scanline | what class? |
| 05:32:53 | StarKruzr | oh hell if I know :-D |
| 05:32:58 | scanline | hehe |
| 05:32:59 | StarKruzr | she's an OpenBSD hacker IIRC |
| 05:34:12 | StarKruzr | dude |
| 05:34:19 | StarKruzr | did you write this whole program yourself? |
| 05:35:32 | scanline | StarKruzr: most of it |
| 05:36:18 | scanline | I started it in March 2000, and started getting some help on it in October |
| 05:37:11 | StarKruzr | You, my friend, are going to be a very rich man |
| 05:39:08 | scanline | I'd prefer not to be... |
| 05:39:28 | StarKruzr | o_O |
| 05:39:31 | StarKruzr | And why is that? |
| 05:40:45 | scanline | I'd rather work on open source software |
| 05:41:46 | StarKruzr | You CAN do both. Heck, work on a startup that does OSS consulting or something. |
| 05:42:14 | StarKruzr | You KNOW there's a market for something like PicoGUI |
| 05:42:46 | scanline | Yeah, I've already had a summer job working on picogui |
| 05:42:52 | StarKruzr | Nice :) |
| 05:43:07 | StarKruzr | I'm probably not going to be very rich doing what I want to do either, actually |
| 05:43:44 | StarKruzr | Spacecraft avionics design, but I'll be in the military, which brings down your paycheck a few 10s of Ks ;) |
| 05:44:33 | scanline | sounds like a fun job |
| 05:44:54 | StarKruzr | space has always been my first love, with computers trailing closely behind |
| 05:45:15 | StarKruzr | but you have to do what you love or you'll be miserable |
| 05:45:21 | scanline | I have some friends like that |
| 05:45:28 | scanline | definitely |
| 05:46:25 | StarKruzr | I figure I'll get the AF to pay for my EE degree |
| 05:46:42 | StarKruzr | they have a free grad school program, if you can get into it |
| 05:47:05 | scanline | neat |
| 05:47:55 | scanline | Seems like there are a lot of people interested in computers and the AF |
| 05:48:07 | StarKruzr | You know others, too? |
| 05:48:10 | scanline | I know at least two |
| 05:48:13 | StarKruzr | wow |
| 05:48:25 | StarKruzr | Good to know the AF is a geek haven |
| 05:50:15 | StarKruzr | well |
| 05:50:28 | StarKruzr | here I was gonna go to sleep and the stupid /. mailer comes into my inbox |
| 05:50:40 | StarKruzr | :p |
| 05:50:54 | scanline | hehe |
| 05:50:58 | cory | heh |
| 05:51:03 | cory | all new aol 7.0! |
| 05:51:16 | collord | i'm glad you like it! |
| 05:51:20 | cory | hehe |
| 05:51:35 | cory | I'm trying to load cnn.com but all I get is ads so far :( |
| 05:52:00 | scanline | use lynx, links, or w3m and I guarantee you no image ads :) |
| 05:52:24 | cory | hehe |
| 05:52:35 | cory | I like w3m more than either lynx or links |
| 05:53:00 | cory | but I like having other images |
| 05:53:00 | cory | hehe |
| 05:53:13 | StarKruzr | uses lynx for text-only |
| 05:53:25 | scanline | my favorite is w3m |
| 05:53:31 | StarKruzr | or whatever's on the CS dept server. Actually it might be w3m, I forget |
| 05:53:35 | StarKruzr | Haven't been in there in a while |
| 05:53:42 | StarKruzr | (forgive the CTCP, all, just curious) |
| 05:53:54 | cory | I mean, just because lynx came from KU doesn't mean it's cool ;) |
| 05:53:58 | cory | goes to KU |
| 05:54:07 | scanline | Once PicoGUI has a web browser, I'm gonna run it with the picogui ncurses driver :) |
| 05:54:22 | StarKruzr | heheh |
| 05:54:23 | StarKruzr | nice |
| 05:54:37 | StarKruzr | I really need to stop being lazy and put my stupid Linux box together |
| 05:54:39 | collord | cory: going over the features of AOL 7.0, i'd like to know what "Built-in Broadband Content" means... |
| 05:55:08 | StarKruzr | collord: probably means they put stuff on your HD that would normally be reserved for sending via broadband |
| 05:55:14 | cory | collord: heheh |
| 05:55:35 | collord | daily newsfeeds for the next decade, already on AOL 7.0 CD! |
| 05:55:40 | StarKruzr | read: "we bloat and make few apologies" |
| 05:55:41 | StarKruzr | collord: LOL |
| 05:56:21 | collord | (or at least the next fifteen issues of Time, and trailers for the next four WB films) |
| 05:56:37 | StarKruzr | :D |
| 05:58:30 | StarKruzr | tries to decide on a distro for his new machine |
| 05:58:40 | StarKruzr | or rather, rebuilt machine |
| 05:58:42 | scanline | yellow dog linux :) |
| 05:58:49 | StarKruzr | heh |
| 05:58:55 | StarKruzr | if it was a Mac, I'd be all over that |
| 05:59:01 | cory | hmmm |
| 05:59:05 | cory | I have an ibook.. |
| 05:59:14 | StarKruzr | k |
| 05:59:16 | scanline | the ibook runs linux nicely- i have one |
| 05:59:17 | cory | is yellow dog ok? it looks nice but I'm not usually a redhat/rpm kind of guy |
| 05:59:34 | StarKruzr | will NOT put RedHate on his box |
| 05:59:40 | scanline | I'm a debian guy, but YDL has good Mac hardware support. I've been meaning to try debian for PPC |
| 05:59:51 | cory | it's not "RedHat" per se, but it does use rpm |
| 05:59:55 | StarKruzr | was a Debian guy too |
| 06:00:00 | scanline | So far I've tried YDL and SuSE, and I like YDL better |
| 06:00:16 | StarKruzr | I've been thinking about Slack |
| 06:00:26 | StarKruzr | I dunno though |
| 06:00:39 | StarKruzr | I've gotten really accustomed to apt :) |
| 06:00:46 | cory | debian rules |
| 06:00:50 | scanline | Yeah, I feel lost without apt-get |
| 06:01:22 | StarKruzr | I just need to learn how to code and I'll actually be able to say I'm a Linux user |
| 06:01:22 | cory | somebody should write a song about apt-get |
| 06:01:32 | scanline | I'd listen to it :) |
| 06:01:36 | cory | you can be a Luser without coding starkruzr ;) |
| 06:01:43 | collord | apt-get install apt-praises |
| 06:01:50 | StarKruzr | LOL |
| 06:02:12 | collord | 493 new packages installed, 3012 upgraded |
| 06:02:52 | scanline | Need to download 2.4GB, continue? |
| 06:03:30 | collord | postinst script failed with error code -99999: you're screwed :) |
| 06:04:16 | StarKruzr | postinst script failed with error code 666: now executing command: rm -rf / |
| 06:05:23 | scanline | postinst script failed with error code -666: now executing command: echo ATDT > /dev/hda |
| 06:05:37 | StarKruzr | LOL HAHAHAHA |
| 06:06:01 | cory | wrinkles his nose |
| 06:06:12 | collord | apt-get remove apt-humor |
| 06:06:20 | scanline | hehe! and you thought that only happens in Minix! |
| 06:06:25 | StarKruzr | sorry |
| 06:06:29 | scanline | ha |
| 06:06:48 | StarKruzr | echo /dev/null > /dev/humor |
| 06:06:55 | StarKruzr | okay I'm done now |
| 06:08:10 | cory | heh |
| 06:08:12 | Lenolium | wonders how many in this channel saw the true humor in scanline's command. |
| 06:08:29 | cory | minix used to think the hard drive device was a modem? |
| 06:08:47 | collord | wants you to fill us in then... |
| 06:08:55 | scanline | It's the reason Linus switched to using Linux fulltime instead of Minix |
| 06:09:10 | scanline | he accidentally used his minix partition instead of the modem device in a terminal emulator |
| 06:09:11 | StarKruzr | read that somewhere a while back |
| 06:09:27 | StarKruzr | yeah I remember that story |
| 06:09:41 | cory | hehe |
| 06:09:54 | collord | and he was mad that Minix let him torch his hdd, or he just needed a new unix box? |
| 06:12:45 | scanline | it gave him an excuse to use Linux fulltime i suppose |
| 06:12:47 | Lenolium | Actually, the story I heard was that Linus did that on his linux partiton, and started implementing permission controls in ext that night. |
| 06:12:55 | StarKruzr | hah |
| 06:13:37 | scanline | hehe. that sounds right |
| 06:14:51 | StarKruzr | scan I had totally forgotten about that story |
| 06:15:27 | cory | yeah, I'd heard of that before |
| 06:15:31 | cory | but I didn't know the string.. |
| 06:16:06 | scanline | ATDT should dial any hayes-compatible modem |
| 06:16:18 | scanline | tries to remember the last time he used a modem... |
| 06:17:15 | cory | heh |
| 06:17:23 | cory | I have a modem in my linux box and my ibook.. just in case.. |
| 06:17:32 | cory | and, uh, soon in my tuxscreen, hehe |
| 06:17:58 | StarKruzr | Every machine I use has a modem |
| 06:18:01 | scanline | I've used the modem in my ibook a few times |
| 06:18:09 | cory | me too. |
| 06:18:11 | StarKruzr | not EVERY hotel has broadband in the room after all ;) |
| 06:18:13 | scanline | seems like a waste of an IRQ on my desktop though |
| 06:18:19 | collord | i use my modem for faxing more than anything.. and that's not much... |
| 06:18:38 | cory | my irq's on my desktop are just fine at the moment |
| 06:20:01 | scanline | I have a bunch of weird cards in mine |
| 06:22:01 | StarKruzr | like what? |
| 06:22:50 | scanline | Matrox G450 dual-head, firewire, 2 ethernet, MPEG decoder, SB16 |
| 06:23:59 | StarKruzr | Ah, separate firewire controller |
| 06:24:10 | StarKruzr | you do capping from video cameras or something? |
| 06:24:17 | scanline | no, external HD |
| 06:24:21 | StarKruzr | ah, nice |
| 06:24:29 | scanline | yeah, i can use it with my ibook too |
| 06:24:40 | StarKruzr | What OS are you running? |
| 06:24:53 | scanline | Debian Linux |
| 06:24:57 | collord | what kind of speed is the firewire mass storage at now? (with linux) |
| 06:24:58 | StarKruzr | PPC? |
| 06:25:09 | StarKruzr | no, Celery 366, you said that on your site |
| 06:25:13 | scanline | yeah |
| 06:25:24 | StarKruzr | Wow, I didn't know firewire mass storage was supported on Linux |
| 06:25:25 | scanline | Slow by today's standards, but still great for coding |
| 06:25:33 | scanline | Yeah, it works pretty well |
| 06:25:34 | StarKruzr | has more reasons to switch now |
| 06:25:38 | StarKruzr | Eventually it will HAVE To happen |
| 06:25:49 | StarKruzr | I refuse to "subscribe" to my operating system |
| 06:25:55 | scanline | What OS do you use? |
| 06:26:10 | StarKruzr | Win2K. My tagline for it is "at least it isn't Win98." |
| 06:26:24 | collord | considers switch to OSX, if he can afford a machine to run it |
| 06:26:28 | scanline | heh. That's what most of the people in my dorm use |
| 06:26:33 | scanline | OSX is nice |
| 06:26:36 | StarKruzr | Win2K or Win98? |
| 06:26:45 | StarKruzr | Yeah, OSX 10.1 is pretty nice from what I hear |
| 06:26:49 | scanline | Well, now they're all running pirated copies of win2k |
| 06:27:02 | StarKruzr | I refuse to pay for Microsoft software out of principle |
| 06:27:08 | scanline | I have a little win95 partition i use occasionally to run Xilinx foundation |
| 06:27:20 | StarKruzr | I buy all the games I play, buy all the utilities I use |
| 06:27:41 | StarKruzr | I even donated to the Klient project, which is an EXCELLENT win32 IRC program |
| 06:28:09 | scanline | Hmm. I was trying to show my roommate how to use IRC. mIrc is weird |
| 06:28:24 | StarKruzr | I hate mIRC with the burning intensity of a thousand suns |
| 06:28:43 | StarKruzr | well |
| 06:28:46 | scanline | hehe |
| 06:28:52 | StarKruzr | maybe 1007 |
| 06:28:54 | StarKruzr | or 1008 |
| 06:28:55 | scanline | the UI is far inferior to X-Chat |
| 06:28:58 | StarKruzr | but a lot, nonetheless |
| 06:29:18 | StarKruzr | scanline: dear God yes |
| 06:29:42 | StarKruzr | Even the really old one I used to use at schoo |
| 06:29:43 | StarKruzr | l |
| 06:29:52 | StarKruzr | Is the latest release of X-Chat multiserver? |
| 06:30:10 | Lenolium | StarKruzr: It's been multiserver since I can remember. |
| 06:30:13 | scanline | I think so... haven't checked (lately I've only been on openprojects) |
| 06:30:54 | StarKruzr | okay |
| 06:31:10 | StarKruzr | We had one on the old dept server which is now sitting next to me on the floor which wasn't |
| 06:35:24 | StarKruzr | Hm |
| 06:35:31 | StarKruzr | Tim Riker just emailed me back |
| 06:35:47 | StarKruzr | but I'm not sure what I need from him in order to put Linux on my TS |
| 06:36:06 | scanline | you need a 12V power source and a null modem cable |
| 06:36:22 | StarKruzr | Well, I think I'm gonna have him unlock the flash |
| 06:36:32 | StarKruzr | Do I still need a 12V power source? |
| 06:36:34 | scanline | no |
| 06:36:37 | StarKruzr | k |
| 06:36:43 | StarKruzr | nullmodem cable then |
| 06:36:44 | scanline | it comes with a power cube |
| 06:36:44 | collord | how many unlocks have gone wrong? |
| 06:36:45 | StarKruzr | Have to dig that out |
| 06:36:59 | scanline | collord: none that I know of |
| 06:37:11 | StarKruzr | hm |
| 06:37:18 | StarKruzr | wonders if it's worth the $25 |
| 06:37:46 | scanline | It's easy to do yourself if you have a 12V supply |
| 06:37:50 | collord | nah... the people on irc will help you work up the nerve... |
| 06:38:45 | scanline | I like the 'chopsticks of death' approach where you hook some multimeter probes up to a big battery :) |
| 06:38:45 | Lenolium | StarKruzr: If you have a spare/dead fan, it's easy to get yourself the 12V, it's a bit tough getting them to the right points within 10s of entering the unlock command (it took me four hours) |
| 06:39:25 | scanline | It's easy if you enter the unlock command (but don't push enter) turn the tux upside-down, press enter, then apply the voltage |
| 06:41:21 | collord | have any of you dealt with JTAG? |
| 06:42:25 | Lenolium | luckily, I havn't |
| 06:49:32 | StarKruzr | lol |
| 06:49:36 | StarKruzr | Ah |
| 06:49:37 | scanline | I think GoRK put a JTAG on one of his tuxscreens |
| 06:49:44 | StarKruzr | Mr. Riker, I was just going to email you back :) |
| 06:50:11 | StarKruzr | What exactly IS a JTAG? |
| 06:50:48 | scanline | StarKruzr: Not sure what the acronym is for... It's a debugging port, you can use it to reflash your tux even if you clobber the bootloader |
| 06:51:01 | collord | Joint Test Action Group iirc |
| 06:51:16 | BZFlag | Joint Test Action Group, yes. |
| 06:51:30 | StarKruzr | so |
| 06:51:31 | collord | Just To Allay Greivousmistakes |
| 06:51:36 | StarKruzr | LOL |
| 06:51:37 | scanline | hehe |
| 06:51:46 | BZFlag | with jtag you can (at a minimum) access the internal flash from a remote device. |
| 06:51:59 | StarKruzr | BZFlag: What exactly do I need from you other than the TuxScreen if I want to get started hacking it? |
| 06:52:03 | scanline | StarKruzr: make sure to take the md5sum of your blob images :) |
| 06:52:35 | BZFlag | some jtag devices include debugging hooks too. the arm7tdmi has that, the strongarm family jtag is not that powerful. |
| 06:52:52 | BZFlag | StarKruzr: that's it, just the phone. |
| 06:53:00 | StarKruzr | cool |
| 06:53:13 | GoRK | hmm my daemon is about 5 minutes away from allowing me to pick up the telephone and dial out |
| 06:53:14 | BZFlag | you need a 12v power supply and steady hands to do the FlashUnlock, |
| 06:53:27 | BZFlag | I do that bit for $25 extra. |
| 06:53:29 | collord | anyone have any recommendations for other cheap ARM-based devices? |
| 06:53:38 | vosque | I used some thin-gauge wire and chopsticks |
| 06:53:41 | GoRK | collord: cerf cube is pretty cool |
| 06:53:43 | StarKruzr | BZFlag: Incoming PayPal. ;) $114USD? |
| 06:53:48 | BZFlag | and you need a DLed version of Inferno (free) to install blob over serial. |
| 06:53:50 | scanline | GoRK: nifty. It will be in CVS soon? |
| 06:53:55 | collord | GoRK: *cheap* |
| 06:53:59 | GoRK | scanline: yeah i guess i could put it in there |
| 06:54:06 | BZFlag | I'm not offering to do the blob/linux install on them yet, though I may in the furture. |
| 06:54:15 | scanline | seems like it should go in the tuxscreen cvs :) |
| 06:54:17 | BZFlag | StarKruzr: sounds good. ;-) |
| 06:54:42 | StarKruzr | well |
| 06:54:47 | StarKruzr | $99 + 15S&H? |
| 06:54:49 | GoRK | collord: hrm someone said that those polaroid photomax (ceiva) digital picture frames that compgeeks was selling for $68 were arm based... i ordered one to see if it was easily hackable |
| 06:55:03 | GoRK | collord: i could not find any info on the net |
| 06:55:17 | BZFlag | collord: I think the ipaq is the next best thing to a TuxScreen. more money, but also a large user base and good support from Compaq. |
| 06:55:24 | collord | GoRK: i'm working on that... any other? |
| 06:55:42 | GoRK | collord: working on what exactly? did you get one of those also? |
| 06:55:49 | StarKruzr | iPaqs aren't $99 though ;) |
| 06:56:14 | BZFlag | StarKruzr: fine. $15 shipping covers any site in the US. feel free to use usps.gov to calculate for where you live and include ammount and shipping type in the comments field. |
| 06:56:32 | collord | GoRK: that was me that dropped the compgeeks a few days ago... i've got 5 on order |
| 06:56:38 | BZFlag | GoRK: ceiva? |
| 06:56:55 | BZFlag | goes to look... |
| 06:57:13 | volkam | smirk |
| 06:57:16 | GoRK | BZFlag: ceiva is the company that will continue to market the rame... they run the service taht lets people "send" photos to others |
| 06:57:26 | GoRK | collord: cool well expect some help from me if i can assist |
| 06:57:38 | GoRK | rame/frame |
| 06:57:42 | collord | GoRK: they're Maverick 7212 which is ARM-based: beamaverick.com memory is the issue. |
| 06:57:58 | GoRK | collord: where did you find specs? |
| 06:58:46 | GoRK | collord: i know it has to have enough power to run pppd and decompress jpegs so i am optimistic regardless |
| 06:58:48 | collord | GoRK: don't have full specs. Digging through press releases... Wind River, Cirrus |
| 06:59:27 | GoRK | ahh. well i have high hopes for it.. i hope it will not be so difficult |
| 06:59:45 | StarKruzr | sales@tuxscreen.net, no? |
| 07:00:00 | GoRK | even if i can't hack it, it'd be a good present for my grandparents or something... $5/mo is nothing |
| 07:00:12 | GoRK | and i know they'd like it a bunch |
| 07:00:31 | collord | GoRK: yeah... i'm not sure on the photo format... they do conversion on their server so i wouldn't count on it being JPEG to the frame... |
| 07:01:11 | volkam | can you see people you call on tuxscreen |
| 07:01:12 | GoRK | collord: the PNG website says that the device supports PNG's via the conversion, but the storage on frame is JPEG |
| 07:01:32 | GoRK | collord: i dont know if that is official or just supposed |
| 07:01:35 | volkam | nm |
| 07:01:38 | volkam | theres not even a camera |
| 07:02:02 | StarKruzr | w00t, ordered |
| 07:02:11 | volkam | heh |
| 07:02:31 | collord | is the cerf board cheaper than the cerf cube? any other ARMs? |
| 07:03:39 | collord | if i'm spending money, it'll be on a bright star nanoengine |
| 07:03:45 | BZFlag | StarKruzr: yes, Sales@TuxScreen.net |
| 07:04:24 | BZFlag | GoRK: it looks like a similar panel to the Tux. get em one of those and netzero and poof, no monthy bill at all. |
| 07:04:59 | BedKruzr | night, folks |
| 07:05:05 | scanline | 'night |
| 07:05:23 | BZFlag | cerf board is much more than the cerf cube. |
| 07:05:31 | GoRK | BZFlag: if i can get in to hack it and at least change the URL to download photos.... %-\ |
| 07:05:34 | hypaStyle | so what is needed to load linux on the tuxscreen ? |
| 07:05:39 | GoRK | or server or whatever |
| 07:06:13 | BZFlag | hypaStyle: http://TuxScreen.net/wiki/view//InstallingLinux |
| 07:06:51 | hypaStyle | thx |
| 07:07:26 | BZFlag | GoRK: it would be a simple cron job on a tux and a directory on a web site to sync too. poof! a ceiva. would require ethernet or an external phone at present though. |
| 07:07:34 | collord | wonders if there some kind of way to sic Ethereal on a phone line |
| 07:07:52 | hypaStyle | could do hpna hehe |
| 07:08:15 | collord | means in the sense of snooping ppp and decoding whatever |
| 07:08:17 | GoRK | BZFlag: ahh you were talking about using a tuxscreen for it.. sorry |
| 07:08:24 | BZFlag | collord: yes. ;-) put a modem on it that goes to a PC, put another modem on the pc that actually goes out, and run a serial sniffer app. |
| 07:08:58 | BZFlag | [ceiva]-----[modem]-[PC]-[modem]---- |
| 07:09:07 | GoRK | BZFlag: the only problem is the display.. those little digital photo frames have nice bright full color displays |
| 07:09:42 | BZFlag | the tux screen is nice colors, but not as bright as many others. |
| 07:10:07 | GoRK | bzflat: also only 8 bit... |
| 07:10:20 | GoRK | err bzflag: sorry :) |
| 07:10:23 | collord | BZFlag: freshmeat tells me there are such things ;) I guess just let the modem listen |
| 07:10:32 | BZFlag | hypaStyle: need hpna hardware though. =/ |
| 07:10:51 | GoRK | collord: i think you have to know the phone number it's calling |
| 07:10:52 | hypaStyle | hrmm there might be pcmcia cards |
| 07:10:53 | collord | hpna hardware is $14/pr. on compgeeks.com |
| 07:11:09 | BZFlag | yes, that can be done too. a phone cable split with a modem in listen mode on the other line. |
| 07:11:33 | BZFlag | hypaStyle: perhaps. |
| 07:12:01 | BZFlag | has a Linux box with hpna at work, but is uses non-free drivers. |
| 07:12:03 | collord | GoRK: the ceiva screen must be 8-bit |
| 07:12:31 | BZFlag | collord: jpeg implies full color, not 8 bit |
| 07:12:34 | GoRK | collord: i have seen them in the stores. i am almost positive it is not 8 bit.. i could be wrong |
| 07:13:21 | collord | Wind River brags about the dithering software they wrote for the device... |
| 07:13:43 | GoRK | someone in here said that it should be possible to make the tuxscreen 16 bit.. that all the hardware should support it |
| 07:15:44 | BZFlag | collord: ahh, that does imply 8 bit, yes. |
| 07:16:13 | BZFlag | GoRK: perhaps not all. The panel supports it, but the panel driver (hw) would need to as well. |
| 07:16:26 | BZFlag | I have seen this panel on another system running 12 bit. |
| 07:16:54 | Lenolium | Isn't there a bandwidth issue at 16-bit? |
| 07:17:51 | GoRK | BZFlag: i though the sa1100 was the lcd controller? |
| 07:19:39 | Lenolium | I remember reading somewhere (don't know if it applies), that the sa1100 doesn't have enough bit-pushing power to run a 16-bit display at 640x480 |
| 07:23:56 | Lenolium | GoRK: So, did you start fresh on your daemon? or did you just hack the crap out of mine? |
| 07:24:00 | collord | the Ceiva is actually 4-bit color: 640x480 at thousands of colors. |
| 07:24:52 | collord | says amazon |
| 07:25:01 | scanline | collord: 4-bit? usually they mean 16-bit when they say 'thousands' |
| 07:25:13 | GoRK | Lenolium: i started fresh and right this second i just picked up the handset and made my first call using dtmf :) |
| 07:25:20 | BZFlag | GoRK: yes, sa1100 is the controller, but there is a bit of hw in the middle. |
| 07:25:22 | Lenolium | GoRK: Cool. |
| 07:25:46 | collord | puts math hat on |
| 07:26:01 | BZFlag | hmmm 4bit does not equal thousands of colors... it equals 16 colors. |
| 07:26:09 | GoRK | i called time & temp again |
| 07:26:30 | BZFlag | GoRK: aren't they sick of hearing from you yet? ;-) |
| 07:26:33 | BZFlag | ducks |
| 07:26:43 | GoRK | BZFlag: heh. sometimes i talk to them there |
| 07:27:04 | GoRK | they dont talk back :( |
| 07:28:08 | GoRK | Lenolium: right now im more focused on making it work as a telephone than making the dsp data available to other apps... as soon as i can get this functional, then I will add hooks for other apps to trap events or hook to event notifications, modify behaviors, etc |
| 07:30:11 | scanline | Any work on designing the interface used to call the daemon? |
| 07:30:48 | GoRK | i have some ideas but not really a lot... Actually I was goign to see what lenolium's work had become at that point |
| 07:31:04 | GoRK | brb.. gonna get some cereal |
| 07:31:04 | Lenolium | scanline: Me and GoRK are designing competing daemons, and then, (in two days or so) we would pick a winner. |
| 07:31:14 | scanline | hehe |
| 07:31:21 | collord | silly me, not 4 bits per channel. yes, virginia, 16 bits per pixel = 65k colors |
| 07:32:01 | scanline | collord: well, if it |
| 07:32:06 | BZFlag | GoRK: have you looked into tone dialing from the keypad and speaker phone? do they work yet? |
| 07:32:08 | scanline | stupid enter key... |
| 07:32:29 | scanline | collord: if it's really 4 bits per channel, it has 12 bits total so it would still qualify for 'thousands' |
| 07:32:42 | BZFlag | ah, but 4 bits per channel = 12 bit which I've seen often before. |
| 07:33:03 | scanline | The Tuxscreen's DAC is 12-bit IIRC |
| 07:33:13 | BZFlag | IIRC the ipaq 3600 are 12 bit data mode. |
| 07:33:27 | | 14 |
| 07:33:27 | BZFlag | ibot 2 ^ 12 |
| 07:33:40 | BZFlag | heh, nope, but thanx for playing.. |
| 07:33:46 | | 4096 |
| 07:33:46 | BZFlag | ibot: 2 ** 12 |
| 07:33:47 | | 4096 |
| 07:33:47 | scanline | ibot 2 ** 12 |
| 07:33:52 | BZFlag | better. |
| 07:33:54 | GoRK | BZFlag: tone dialing from the handset works. the speakerphone is kind of complicated |
| 07:33:57 | collord | but N bits per channel implies 3xN-bit truecolor, not 3xN-bits to reference a pallete. |
| 07:34:05 | scanline | C syntax i guess... |
| 07:34:10 | BZFlag | GoRK: cool! |
| 07:34:18 | GoRK | BZFlag: you have to download a ton of data into the speakerphone so it can do echo cancellation and stuff |
| 07:35:03 | GoRK | BZFlag: so far, i have not been able to successfully get that info in or out of the dsp.. i think there is something that is unique about that much data that has to go in the kernel driver |
| 07:36:04 | scanline | Didn't someone here have a logic analyzer set up on the DSP? we could see what inferno's sending it |
| 07:36:42 | BZFlag | scanline: see the -devel list. |
| 07:37:02 | GoRK | scanline: my idea for the "daemon" part of things... is to allow clients who connect to recieve all state change messages and events that occur in the dsp and also allow clients to request preempting certain functions... e.g. a client can connect and assume total control over the functions of the volume buttons for instance |
| 07:37:30 | GoRK | and they would also be allowed to send any commands they wanted to to the dsp |
| 07:38:08 | scanline | GoRK: Maybe each client (including the daemon's builtin functions) should have a 'priority level' determining what order they get called in, and they can opt whether to pass it on to other handlers |
| 07:38:12 | collord | i'm failing all tests. g'night all. |
| 07:38:13 | BZFlag | GoRK: have you reviewed the exiting telephony driver? have you read the Linux Journal article about it? |
| 07:38:26 | GoRK | BZFlag: no can you point me to it? |
| 07:38:36 | GoRK | is that article available online? |
| 07:39:20 | BZFlag | http://www2.linuxjournal.com/lj-issues/issue82/4468.html |
| 07:39:26 | BZFlag | and yes. ;-) |
| 07:39:40 | GoRK | scanline: it's not a question of what data to send the speakerphone; it's sending it that much that it has a problem with... something is funky. the inferno driver seems to have some special casing for doing that particular operation |
| 07:40:10 | scanline | hmm |
| 07:41:25 | BZFlag | the approach they took is to do everything in ioctl calls and just present the audio interface as the device. |
| 07:41:49 | BZFlag | that is what some telephony apps like openh323 will be expecting. |
| 07:42:23 | BZFlag | not sure a) if we can do that or b) if that's what we want even if we could. |
| 07:42:26 | GoRK | the only real problem with our device is that the audio will actually be going through a *different* device... |
| 07:42:45 | GoRK | the audio will 99% likely have to go through ucb1200-audio |
| 07:42:49 | BZFlag | but we should be familiar with it so that we know. |
| 07:43:26 | GoRK | do you know if there's anyway to simulate a character device that can accept ioctl's in userspace? |
| 07:43:44 | BZFlag | ok, so perhaps we will build a future device that is lust a layer on top of ucb1200-audio and the direct dsp control interface. |
| 07:44:16 | BZFlag | simulate a character device, yes. ioctls, hmm... not sure. |
| 07:44:22 | GoRK | well... i guess if we get a good phone daemon going in userland, we can move it into kernelspace and have it create a 'virtual' device that is some bizzare synthesis of the dsp/daemon and the ucb1200 device |
| 07:44:30 | BZFlag | s/lust/just/ =( |
| 07:44:32 | GoRK | i know you can simulate character devices with fifo's |
| 07:44:40 | GoRK | i have done it :) |
| 07:44:53 | BZFlag | heh ;-) |
| 07:44:57 | GoRK | i hate kernelspace. i do everything possible to engineer out of it |
| 07:45:03 | BZFlag | agreed |
| 07:45:10 | GoRK | but |
| 07:45:28 | GoRK | i know a kernel daemon most likely can operate a character device |
| 07:45:41 | GoRK | so it could work |
| 07:46:22 | GoRK | it would be highly ... unique .. for the tuxscreen but not a bad deal |
| 07:47:30 | GoRK | that's why i was looking at zaptel/zapata |
| 07:48:01 | GoRK | all you have to do is write glue to detect switchook, button presses, and tell it how to record and playback audio and it does everything else |
| 07:48:37 | GoRK | that's what mark spencer told me. he's the asterisk pbx guy. i know you sold him a tuxscreen at one point... |
| 07:52:18 | BZFlag | sound plausible, yes. |
| 07:52:56 | BZFlag | nite all. |
| 07:59:13 | | 3 |
| 07:59:13 | hypaStyle | ibot : 1 + 1 |
| 07:59:16 | | 3 |
| 07:59:16 | hypaStyle | ibot : 1 + 1 |
| 07:59:19 | | 5 |
| 07:59:19 | hypaStyle | ibot : 1 + 4 |
| 07:59:27 | hypaStyle | ibot : 1 * 0 |
| 07:59:40 | | 1 |
| 07:59:40 | hypaStyle | ibot : 1 ** 0 |
| 08:02:17 | | Lenolium: i'm not following you... |
| 08:02:17 | Lenolium | ibot: 7 / 0 |
| 08:02:57 | | fbzr fghss |
| 08:02:57 | scanline | ibot: rot13 some stuff |
| 08:04:28 | | I don't believe there really IS a GAS SHORTAGE.. I think it's all just a BIG HOAX on the part of the plastic sign salesmen -- to sell more numbers!! |
| 08:04:28 | scanline | ibot yow |
| 08:04:37 | scanline | hehe |
| 08:17:00 | hypaStyle | wtf is that |
| 08:17:00 | hypaStyle | hehe |
| 08:17:09 | scanline | try /msg ibot help |
| 08:17:11 | scanline | it's interesting |
| 08:17:58 | | Slashdot - Updated 2001-10-22 07:25:00 | Microsoft's Future | Technology and Society | GNU Emacs 21 | Security Issues with Windows 2000 Datacenter? |
| 08:17:58 | hypaStyle | ibot : slashdot |
| 08:18:20 | hypaStyle | x is at http://www.xxx.com |
| 08:18:27 | | x is used because it exists, many toolkits exist for it and people know it |
| 08:18:27 | hypaStyle | ibot : where is x |
| 08:18:42 | hypaStyle | sweetx is at http://www.xxx.com |
| 08:18:47 | | no idea, hypastyle |
| 08:18:47 | hypaStyle | ibot : where is sweeetx |
| 08:18:56 | | OK, hypaStyle. |
| 08:18:56 | hypaStyle | ibot : sweetx is at http://www.xxx.com |
| 08:19:01 | | hypastyle: bugger all, i dunno |
| 08:19:01 | hypaStyle | ibot : where is sweeetx |
| 08:19:07 | scanline | ha |
| 08:19:13 | hypaStyle | what progam is that ? |
| 08:19:33 | scanline | a CTCP VERSION says it's "infobot" |
| 08:19:37 | Lenolium | Looks like infobot |
| 08:19:58 | | hypaStyle: sorry... |
| 08:19:58 | hypaStyle | ibot : infobot |
| 08:20:00 | | hmmm... tuxscreen is http://TuxScreen.net/ order at http://tuxscreen.net/wiki/view/Users/TimRiker and http://tuxscreen.net/scrapbook/images.html or #TuxScreen |
| 08:20:00 | scanline | ibot: where is tuxscreen |
| 08:20:14 | | tuxscreen is http://TuxScreen.net/ order at http://tuxscreen.net/wiki/view/Users/TimRiker and http://tuxscreen.net/scrapbook/images.html or #TuxScreen |
| 08:20:14 | hypaStyle | ibot : what is tuxscreen |
| 08:20:23 | | ibot is, like, an infobot owned by Tangent with some assistance from BZFlag or stupid |
| 08:20:23 | scanline | ibot: what is ibot |
| 08:20:29 | scanline | ah |
| 08:20:50 | | OK, hypaStyle. |
| 08:20:50 | hypaStyle | ibot : lbot is stupid |
| 08:20:56 | | hmmm... lbot is stupid |
| 08:20:56 | hypaStyle | ibot : what is lbot |
| 08:21:01 | | scanline: wish i knew |
| 08:21:01 | scanline | ibot: what is stupid |
| 08:21:09 | | i heard BZFlag was TimRiker or a 3D internet multiplayer multiplatform (linux, irix, win32, etc) opensource opengl Battlezone capture the flag game at http://BZFlag.org/ debian arm linux contains a version that will run on the ipaq at like 6 fps |
| 08:21:09 | scanline | ibot: what is BZFlag |
| 08:21:09 | | OK, hypaStyle. |
| 08:21:09 | hypaStyle | ibot : Iam is stupid |
| 08:21:15 | | Iam is probably stupid |
| 08:21:15 | hypaStyle | ibot : what is Iam |
| 08:21:42 | scanline | haha |
| 08:22:05 | | I haven't seen 'ibot', scanline |
| 08:22:05 | scanline | ibot: seen ibot |
| 08:22:30 | | hypaStyle: i'm not following you... |
| 08:22:30 | hypaStyle | ibot : weather |
| 08:22:42 | scanline | ibot: help |
| 08:22:50 | hypaStyle | ibot : metar KAG |
| 08:22:50 | | hypaStyle: I can't find any observations for KKAG. |
| 08:22:59 | | ibot is nothing but a porous accumulation of toad-spotted dog balls. |
| 08:22:59 | scanline | ibot: insult ibot |
| 08:23:00 | hypaStyle | ibot : metar KAGC |
| 08:23:00 | | hypaStyle: KAGC 220853Z 22005KT 10SM BKN075 15/08 A3006 RMK AO2 SLP179 T01500083 55005 |
| 08:23:30 | | hypaStyle: sorry... |
| 08:23:30 | hypaStyle | ibot : traceroute apple.com |
| 08:23:41 | hypaStyle | ibot : internic yahoo,com |
| 08:23:43 | hypaStyle | ibot : internic yahoo.com |
| 08:23:51 | | scanline: what? |
| 08:23:51 | scanline | ibot: /bin/sh |
| 08:24:01 | | clerks can be found at http://www.imdb.com/M/title-substring?title=clerks&type=fuzzy |
| 08:24:01 | hypaStyle | ibot : imdb for clerks |
| 08:24:14 | | sex in the city can be found at http://www.imdb.com/M/title-substring?title=sex+in+the+city&type=fuzzy |
| 08:24:14 | hypaStyle | ibot : imdb for sex in the city |
| 08:24:56 | | hypaStyle: huh? |
| 08:24:56 | hypaStyle | ibot : '/bin/sh' |
| 08:25:25 | | hypaStyle: excuse me? |
| 08:25:25 | hypaStyle | ibot : goodnight |
| 08:25:30 | | hypaStyle: i'm not following you... |
| 08:25:30 | hypaStyle | ibot : bed |
| 08:25:43 | | i don't know, hypastyle |
| 08:25:43 | hypaStyle | ibot : wtf |
| 08:25:47 | | scanline: i'm not following you... |
| 08:25:47 | scanline | ibot: perl -e 'System("/bin/sh");' |
| 08:26:33 | | blob is probably the Boot Loader OBject |
| 08:26:33 | scanline | ibot: what is blob |
| 09:24:31 | GoRK | hmm something about changing the volume freaks the DSP out quite a lot |
| 09:24:50 | scanline | strange |
| 09:25:02 | | scanline: wish i knew |
| 09:25:02 | scanline | ibot: what is the DSP? |
| 09:25:31 | GoRK | well once i make the phone ring when it gets a ring im gonna let this thing out |
| 09:26:07 | scanline | cool |
| 09:26:29 | scanline | you could make the ring configurable: if the normal one is annoying, have it happy_tone instead :) |
| 09:26:46 | GoRK | i know .. i am going to but just getting the basics down |
| 09:26:57 | scanline | hehe. I just think it would be funny to customize the ring |
| 09:27:34 | scanline | isn't it possible for this thing to play sampled sound? You could have it play an mp3 file when a call comes in! |
| 09:39:13 | GoRK | scanline: it may be.. definately not yet |
| 09:39:44 | scanline | The sound hardware is in the CPU, right? |
| 09:40:05 | GoRK | no its in ucb1200 |
| 09:46:33 | scanline | do we have an audio driver for the ucb1200 yet? |
| 09:46:38 | GoRK | yep |
| 09:47:02 | scanline | so all we'd need to do us program the DSP to route the audio correctly |
| 09:47:34 | GoRK | yep and do some hardware modification to wire it up to the ucb1200 |
| 09:47:58 | scanline | the audio output on the ucb1200 isn't connected? |
| 09:49:13 | GoRK | i dont think so, no |
| 10:00:54 | GoRK | nite |
| 10:01:20 | scanline | 'night |
| 10:08:26 | Lenolium | Ugh... |
| 10:08:59 | Lenolium | I have to be to Client1 in 6 hours.... and I am still working at Client2's site. |
| 10:09:12 | scanline | ugh |
| 10:09:47 | scanline | I have to be at Class1 in 3 hours, and I'm still finishing Class1's lab |
| 10:12:21 | scanline | almost done... |
| 10:12:43 | scanline | then maybe I'll stay up and code. If I take a 'little nap' I'll sleep right through class |
| 10:13:20 | scanline | Lenolium: what kind of client? |
| 10:14:44 | Lenolium | Client's who pay me money. (I'm a consultant) |
| 10:15:17 | scanline | what kind of consultant? |
| 10:15:21 | Lenolium | Client1 I do system administration, Client2 I do web application development |
| 10:15:36 | scanline | ok |
| 10:15:42 | Lenolium | Client1 has a virus, Client2 has a rapidly-approaching deadline. |
| 10:15:51 | scanline | ugh |
| 10:20:18 | Lenolium | a virus on the radio broadcasting station systems. |
| 10:20:26 | scanline | what OS? |
| 10:20:33 | Lenolium | Win NT |
| 10:20:46 | Lenolium | Runs this really crappy software. |
| 10:20:49 | scanline | could be worse... |
| 10:20:56 | scanline | could be Win ME |
| 10:21:05 | Lenolium | yeah... |
| 10:32:52 | Lenolium | Mornin SpaceCoaster |
| 10:33:00 | scanline | good morning |
| 10:33:20 | scanline | yay, done with my lab, time to debug... |
| 10:35:55 | Lenolium | GAAR. |
| 10:36:45 | SpaceCoaster | Lenolium: howdy |
| 10:49:27 | Lenolium | SpaceCoaster: Mornin. |
| 10:56:24 | Lenolium | *plong* |
| 11:12:43 | SpaceCoaster | Lenolium: GoRK had trouble with the ioctl reset did you? |
| 12:08:11 | scanline | is away: food |
| 12:40:15 | scanline | is back (gone 00:32:04) |
| 12:43:21 | scanline | is away: class |
| 12:57:35 | Niddix | Anyone actually here? |
| 12:58:45 | prpplague | morning all |
| 12:58:55 | Niddix | Quite group. |
| 12:58:59 | Niddix | Quiet even. |
| 13:00:56 | prpplague | Niddix: shhh, everyone is sleeping |
| 13:01:05 | mallum | HELLOO |
| 13:01:13 | Niddix | I guess. |
| 13:01:14 | mallum | WHO WOKE ME ???? |
| 13:01:32 | Niddix | I was hoping to find Tim Riker in here. |
| 13:01:43 | prpplague | Niddix: now you've done it |
| 13:02:34 | Niddix | *points at prpplague* He did it. Not me.. |
| 13:02:49 | SpaceCoaster | yawn, stretch, grrr |
| 13:05:35 | Niddix | Either of you actually have a tuxscreen? |
| 13:05:39 | SpaceCoaster | Niddix: tim is in the middle bit of the country, he also works late, try around 1 or 2 EST |
| 13:06:02 | SpaceCoaster | indeedy |
| 13:06:17 | Niddix | I was thinking of trying to use one as an X10 controller. |
| 13:06:37 | SpaceCoaster | me too, practically perfect, with the touchscreen and all. |
| 13:06:53 | mallum | Niddix: I've been awake for ages ;) |
| 13:06:57 | Niddix | yeah.. I wanted to see if Tim still had any available. |
| 13:07:17 | SpaceCoaster | 800 is last count, see the wiki... |
| 13:07:24 | SpaceCoaster | ^is^ish |
| 13:07:33 | Niddix | I wasn't sure how up to date that was. |
| 13:08:03 | SpaceCoaster | pretty good, he says that he is selling about 30 a week |
| 13:08:36 | Niddix | Nod.. the project got slashdotted yesterday. I would imagine or hope that those numbers might go up. |
| 13:08:49 | Niddix | Well hope for it.. not for me.. since I want one ;) |
| 13:10:01 | Niddix | Does anyone know what these things were originally built to be? |
| 13:10:54 | SpaceCoaster | again see the wiki history, they are fancy phones with an isp connection. |
| 13:11:33 | nelson | Niddix: the project got slashdotted yesterday? In which story? |
| 13:12:41 | mallum | nelson: I think it was the geek gadget X-mas presents one |
| 13:12:53 | SpaceCoaster | signal11: ...the great spacecoaster... damn that song |
| 13:15:04 | Niddix | It was part of a reply to the article... Hackable Christmas presents. |
| 13:15:19 | mallum | ah yes, that was it |
| 13:15:46 | signal11 | spacecoaster heh |
| 13:15:54 | Niddix | http://slashdot.org/askslashdot/01/10/18/1823219.shtml |
| 13:15:57 | SpaceCoaster | signal11: hi ya |
| 13:16:01 | signal11 | playing with that mp3 is like playing with a hand grenade |
| 13:16:21 | signal11 | whats new |
| 13:16:48 | signal11 | i have the tux screen in the living room now. but the stylus is really not the best pointing device for that |
| 13:17:07 | signal11 | i might need to hook up a track ball or something to the console port |
| 13:17:13 | SpaceCoaster | Niddix: there was a complete article about a month ago http://slashdot.org/article.pl?sid=01/09/05/1413232&mode=thread |
| 13:17:29 | Niddix | Oh.. I must have missed that one. |
| 13:17:59 | SpaceCoaster | And photos of tims living room after the SlashDotting on the wiki |
| 13:21:38 | prpplague | if any of you guys are considering buying a frontpath progear, don't! |
| 13:27:51 | CosmicPenguin | Good morning! |
| 13:29:53 | SpaceCoaster | CosmicPenguin: how was vegas? |
| 13:30:33 | CosmicPenguin | It was really good. If any one gets a chance to see the Circe de Soeil at Treasure Island, I highly reccommend it.... |
| 13:30:58 | SpaceCoaster | CosmicPenguin: are you a millionaire now? |
| 13:31:15 | CosmicPenguin | SpaceCoaster: Nope.... But I did come back even, so thats always good news.... |
| 13:31:43 | signal11 | cosmicpenguin did you ever watch the great space coaster? it was a kids show in the 80's |
| 13:31:56 | SpaceCoaster | CosmicPenguin: you came back ahead, we implemented some of the stuff in the spec and it works |
| 13:32:12 | CosmicPenguin | SpaceCoaster: Rock on. Find any bugs in the spec? |
| 13:32:28 | SpaceCoaster | signal11: that was one of the programs that didn't make it across to Britain. |
| 13:32:50 | CosmicPenguin | SpaceCoaster: If I can get through my backlog here, I'll check our your code. |
| 13:33:04 | SpaceCoaster | CosmicPenguin: yes, if you send us the latest source then we will update it and return it. |
| 13:33:28 | CosmicPenguin | I don't know what GoRK has accomplished yet, but I am hoping that we can discuss the userland side of all of this.... |
| 13:33:56 | CosmicPenguin | I was thinking about a library of routines that we can all use.... I am sure that all the GUIs will want their own deamon.... |
| 13:34:08 | CosmicPenguin | SpaceCoaster: Do you mind StarOffice mangled HTML? |
| 13:34:41 | SpaceCoaster | CosmicPenguin: not at all, we use StarOffice too, so native is fine |
| 13:34:54 | CosmicPenguin | Watch your e-mail |
| 13:35:51 | CosmicPenguin | signal11: You been watching the LKML this week? |
| 13:36:01 | signal11 | cosmicpenguin no, whats new |
| 13:36:06 | SpaceCoaster | Tim is quite keen to do Linux Telephony stuff but I think that we are at too early a stage to move much in that direction. |
| 13:36:12 | CosmicPenguin | signal11: I' |
| 13:36:23 | CosmicPenguin | signal11: I'm just wondering if I need to read all 1211 messages.... :( |
| 13:36:29 | signal11 | ack that's a lot |
| 13:36:33 | signal11 | i wish they'd bring back the digest |
| 13:36:44 | CosmicPenguin | SpaceCoaster: I agree. We need to get it to pick up the phone before we do anything....... |
| 13:36:50 | CosmicPenguin | SpaceCoaster: What commands have we tried so far? |
| 13:37:27 | SpaceCoaster | nah GoRK has all that sorted, I haven't tried anything really just beeps and flashes, concentrating on the driver. |
| 13:37:45 | CosmicPenguin | SpaceCoaster: Ok, thats fine.... |
| 13:38:27 | CosmicPenguin | SpaceCoaster: I like the glossary.... :) |
| 13:38:27 | SpaceCoaster | GoRK: a good page about the SPI was added to the Wiki by a hardware guy, I've added our usage but please add what you find. |
| 13:39:04 | SpaceCoaster | CosmicPenguin: thanks, it was to put my own thoughts in place. |
| 13:39:04 | CosmicPenguin | SpaceCoaster: The flash memory has a whole different interface other than the SPI, so don't worry about that for now.... |
| 13:39:14 | CosmicPenguin | <---- obviously reading code |
| 13:39:29 | SpaceCoaster | CosmicPenguin: too late, i'm worrying about it, looks easy |
| 13:39:39 | CosmicPenguin | :) |
| 13:39:58 | SpaceCoaster | it uses the SPI as it was intended to be used via the SSP using FIFOs and all |
| 13:40:22 | SpaceCoaster | I want to run JFFS2 on it by the end of the week! |
| 13:40:24 | CosmicPenguin | SpaceCoaster: Yep.... You need to turn off the alternate functions bits.... |
| 13:40:30 | CosmicPenguin | SpaceCoaster: Ugh.... |
| 13:41:17 | CosmicPenguin | SpaceCoaster: Why a 10ms delay after writing a message? |
| 13:42:50 | SpaceCoaster | Looking at that now, RTS/CTS is used to handshake reads but nothing indicates when it is safe to write. |
| 13:43:36 | SpaceCoaster | If we don't find a handshake I can use a timer to pump the data out without delaying() in the kernel. |
| 13:44:20 | CosmicPenguin | SpaceCoaster: From here, it looks like the RTSN controls everything |
| 13:45:04 | SpaceCoaster | yeah but it's an input, I guess we might toggle CTS and see what happens |
| 13:45:29 | CosmicPenguin | SpaceCoaster: I would assume that the DSP tells *US* when its clear to write |
| 13:46:11 | SpaceCoaster | maybe or maybe we ask and it says yes when it's ok |
| 13:48:21 | SpaceCoaster | maybe assert CTS wait for RTS, send data, clear CTS. Which is what happens in reading. Now I understand this a bit better it should be traceable. |
| 13:48:53 | CosmicPenguin | SpaceCoaster: Ok, ok, I see where you are coming from...... |
| 13:49:24 | CosmicPenguin | It looks like you do have to sleep 10ms after each command.... bastards! |
| 13:50:35 | SpaceCoaster | CosmicPenguin: if that is the case then a timer to pump the data out will do the job. 100 commands a second is fine for me :-) |
| 13:52:59 | SpaceCoaster | The timer will also work on the initial delay on open (800ms). The first write won't progress until the timer has expired. This will allow the app to do stuff after open. |
| 13:53:44 | SpaceCoaster | CosmicPenguin: some of the commands require a 1000ms delay. |
| 13:54:00 | CosmicPenguin | SpaceCoaster: Where do you get that? |
| 13:54:32 | CosmicPenguin | SpaceCoaster: Some of the serial flash stuff does, but IMHO thats a whole nother driver |
| 13:56:59 | SpaceCoaster | maybe, but it would be simple in wheaties.c and complex apart, they would have to interact heavily, lets get them talking and then slice and dice. |
| 14:00:31 | CosmicPenguin | SpaceCoaster: Ok, ok.... Thats why I didn't even document the serial flash, because I figured that the DSP was way more important |
| 14:01:18 | CosmicPenguin | SpaceCoaster: anyway, the code looks good. There are a few places where you should be using mdelay() instead of udelay(), and I agree that we shouldn't init the DSP on every open, but it looks good |
| 14:02:45 | SpaceCoaster | indeed |
| 14:03:07 | CosmicPenguin | I would also like to see some script parsable output in the device file instead of digits, but thats a huge religious dicussion that I am sure will be passed back and forth for months.... |
| 14:03:45 | SpaceCoaster | what do you mean? |
| 14:05:08 | CosmicPenguin | SpaceCoaster: Oh, you know.... the ongoing wars regarding what belongs in the kernel and what belongs in the user space.... |
| 14:06:16 | CosmicPenguin | I mean, I would like to see the device driver output something like: |
| 14:06:22 | CosmicPenguin | RING LINE1 |
| 14:06:48 | SpaceCoaster | It's ok as soon as someone suggest moving it into the kernel I'll develope an app that requires it to be in userspace. "Tone Dialing must be in the kernel!", "but i want to do dialing shortcuts!" etc, etc. |
| 14:07:23 | SpaceCoaster | A decoded output could be provided via a minor device, so that's no problem. |
| 14:07:54 | SpaceCoaster | Thanks for the mdelay() I'll put it in, before I work out how to eliminate it! |
| 14:08:58 | SpaceCoaster | "I want to do call logging", "I want to do least cost routing"... |
| 14:11:36 | SpaceCoaster | "I want to disable international and 900 calls", "I want to implement 'follow me'" |
| 14:11:43 | SpaceCoaster | ...end rant |
| 14:17:50 | signal11 | is there any kind of driver standard from the telephony projects? |
| 14:19:35 | mallum | SpaceCoaster: do you think it would be possible to make the TUX work with callerID in the uk |
| 14:19:46 | mallum | knows nothing about callerID BTW |
| 14:27:16 | signal11 | heh: http://www.mnftiu.cc/mnftiu.cc/war.html |
| 14:30:17 | signal11 | warning, not politically correct |
| 14:35:56 | SpaceCoaster | mallum: don't know, probably not the DSP is recognising the line twiddles and sending back the digits, they were planning a Euro version but I think it didn't make it |
| 14:36:23 | CosmicPenguin | I am assuming CID is very different in europe? |
| 14:37:20 | SpaceCoaster | very different in Britain, British Telecom are like Microsoft, they adhere to the standards by writing their own. |
| 14:37:31 | mallum | heh |
| 14:37:53 | SpaceCoaster | mallum: is it true? |
| 14:38:13 | boris | except bt are really pro linux |
| 14:38:28 | boris | they use it everywhere afaik |
| 14:38:33 | CosmicPenguin | The Bell's here are definately *not* pro-linux |
| 14:38:59 | mallum | SpaceCoaster: what is true ? that BT is like M$ ? |
| 14:39:52 | SpaceCoaster | mallum: yeah |
| 14:40:19 | mallum | SpaceCoaster: I wouldn't know - Id imagine so - BT are pretty stupid |
| 14:45:41 | SpaceCoaster | mallum: I now have the rinky dinkiest telecom provider in all the world, only about 10,000 customer, ADSL in three days from order, janet in accounts sorts out any problems, it's brilliant |
| 14:48:31 | mallum | SpaceCoaster: you run an isp ? |
| 14:50:06 | SpaceCoaster | mallum: I wish, no my local phone company in New Hampshire. |
| 14:53:08 | mallum | SpaceCoaster: nice. I have to use cableLondon ( now telewest ) , the cable modem was great to begin with but now is continually up and down |
| 14:57:32 | SpaceCoaster | mallum: we used ISDN in Edinburgh, BT made the billing so complicated that they had to do it by hand every month, in the end they undercharged us by about 300 GBP a year, which was nice |
| 14:57:58 | mallum | cool, always good to get one over on bt |
| 14:58:46 | scanline | is back (gone 02:15:25) |
| 14:58:51 | SpaceCoaster | are they continual moving the numbers around to increase prices whilst claiming they are decreasing them? |
| 14:59:47 | mallum | probably |
| 15:24:13 | BZFlag | I just got an email from a german company http://www.memphis-components.de/ they have inventory on the 32M sodimm. Don't have a price yet. |
| 16:20:49 | pattieja | is away: lunch |
| 16:43:31 | SpaceCoaster | BZFlag: I just got mail from Thorsten Wronski @ memphis, I think he's looking for you, I have forwarded it to tim@rikers.org |
| 17:02:24 | scanline | is away: classs |
| 17:02:27 | scanline | is away: class |
| 17:12:55 | BZFlag | he found me. |
| 17:13:18 | BZFlag | SpaceCoaster: he claims to have cheaper 32M sodimms. could be a deal, we'll see. |
| 17:13:49 | AlHazWork | shipping from germay might even things out |
| 17:14:18 | AlHazWork | BZFlag: how's that thrift store VCR working out for you? |
| 17:15:45 | BZFlag | AlHazWork: heh, never even plugged it in. |
| 17:17:10 | AlHazWork | BZFlag: the one i picked up works pretty well, aside from the fact that it doesn't open the door when you hit eject, so the tape just smacks against the door and feeds back in unless you hold it open. Mostly bought it for parts, it's identical to the one I already had |
| 17:17:43 | BZFlag | cool. |
| 17:17:53 | BZFlag | headed into the office. cya |
| 17:23:29 | CosmicPenguin | has anyone else had issues building the new Emacs from CVS? |
| 17:26:49 | pattieja | is back (gone 01:06:00) |
| 17:27:54 | prpplague | what is the /proc entry that shows the currently 802.11b signal values? |
| 17:28:22 | AlHazWork | depends on the driver |
| 17:28:35 | prpplague | ahh, wavelan? |
| 17:28:48 | AlHazWork | dunno, don't have one, sorry . . . |
| 17:28:49 | CosmicPenguin | .. /proc/net/wireless should work for hermes |
| 17:28:57 | CosmicPenguin | And I would assume wavelan as well |
| 17:31:28 | prpplague | thanks |
| 17:35:10 | Lenolium | *pling* |
| 17:40:53 | Lenolium | Alright... back to work on wheatiesd |
| 17:50:39 | SpaceCoaster | CosmicPenguin: managed to remove the 10ms delay, just smallish delays left? |
| 17:52:20 | SpaceCoaster | CosmicPenguin: I think that we can transfer 1 message every 200-300 us, which is good enough. |
| 17:55:06 | Lenolium | SpaceCoaster: Anything different in the kernel driver since the last update you sent me? |
| 17:57:17 | SpaceCoaster | Lenolium: nothing significant just improved throughput, the delays can be take out now that we are using spinlocks and interrupts |
| 17:58:40 | CosmicPenguin | Lenolium: Would you mind putting all the wheaties stuff in a generic library as opposed to a deamon? |
| 17:59:02 | CosmicPenguin | SpaceCoaster: That sounds good to me... after all, the DSP is only delivering at 500ms anyway |
| 17:59:23 | SpaceCoaster | 500ms? |
| 17:59:28 | CosmicPenguin | Sorry, us |
| 17:59:49 | SpaceCoaster | how do you know that? |
| 18:00:08 | BZFlag | something needs to keep pinging the dsp though as apps come and go. |
| 18:00:39 | CosmicPenguin | c-n-p: * set bit 15 in GRCR (Drop CTS) |
| 18:00:39 | CosmicPenguin | If the next word of the message does not arrive in 500usec, then the message is done |
| 18:01:26 | CosmicPenguin | BZFlag: True.... personally, I would like to see more of that in the kernel.... I think we need to have a few battles regarding that before I feel comfortable with the userland stuff.... |
| 18:01:43 | Lenolium | Kernel drivers should be very simple. |
| 18:01:44 | Russ|werk | it would make more since to me to have the kernel do a reset on start, and then ping the DSP |
| 18:02:02 | SpaceCoaster | ahh, yes this is true, i do that, it's when reading a message |
| 18:02:10 | Lenolium | And, if the userland gets busted, you want the DSP watchdog to init, so you can still dial, even if the userland is all dead. |
| 18:02:15 | CosmicPenguin | It doesn't need to be kept alive.... |
| 18:02:26 | CosmicPenguin | the DSP I mean.... |
| 18:03:34 | BZFlag | CosmicPenguin: I'd like to have tone dial right from bootup. |
| 18:03:37 | Russ|werk | have you played with audio routes? |
| 18:03:57 | Russ|werk | the ucb should at least be able to route to line1 (modem operation) |
| 18:03:57 | CosmicPenguin | BZFlag: So would I... I also would like to have a perl script that could run the DSP if needed |
| 18:04:07 | Russ|werk | and hopefully handset and speakerphone |
| 18:04:07 | prpplague | hmmm, anyone know where i can find the details about the /proc/net/wireless entry? |
| 18:04:18 | Lenolium | Russ|werk: Yup, those are working. |
| 18:04:25 | BZFlag | CosmicPenguin: a deamon could still be addressed from perl, no? |
| 18:05:07 | SpaceCoaster | Or over the network |
| 18:05:21 | CosmicPenguin | BZFlag: Not as easily as parsing /dev/wheaties..... |
| 18:05:51 | CosmicPenguin | RING LINE1 |
| 18:05:53 | CosmicPenguin | RING LINE1 |
| 18:06:00 | BZFlag | CosmicPenguin: hmmm... the deamon could just build a /tmp/weaties with the same interface for now. |
| 18:06:13 | SpaceCoaster | CosmicPenguin: I though you meant shell scripts, perl can munge anything, more easily than C (or anything else). |
| 18:06:16 | CosmicPenguin | CID1 8015551234 COSMICPENGUIN |
| 18:06:32 | SpaceCoaster | ^though^thought |
| 18:06:34 | CosmicPenguin | SpaceCoaster: <insert your favorite scripting here> |
| 18:06:41 | Russ|werk | Lenolium: so you can play music on the speaker? |
| 18:06:56 | Lenolium | Russ|werk: Not so far. |
| 18:07:02 | Lenolium | Russ|werk: We can make it beep a lot though. :) |
| 18:07:14 | CosmicPenguin | Russ|werk: We're not sure how to get user sound into the DSP..... |
| 18:07:56 | SpaceCoaster | At this time I would like to stick with things which can't be done from userspace, interrupts, accurate timed loops etc, is this ok? |
| 18:10:20 | Lenolium | CosmicPenguin: The wheatiesd will have a simple client application that will send commands to the DSP, and give you back the reply. |
| 18:10:21 | SpaceCoaster | Russ|werk: thanks for the O'Reilly ref it got me going on the right track. |
| 18:10:46 | CosmicPenguin | Lenolium: Thats fine... but I would rather that be in a library instead of a user app..... |
| 18:11:32 | CosmicPenguin | I would rather not have two deamons running just to pop a window on the screen when the phone starts ringing... |
| 18:12:11 | Russ|werk | I would think it would be a daemon |
| 18:12:23 | Russ|werk | clients would connect maybe with unix sockets |
| 18:12:46 | CosmicPenguin | Why?? |
| 18:12:52 | Lenolium | Russ|werk: Yup, it will be able to connect with clients via unix sockets, or optionally via tcp/ip |
| 18:13:06 | CosmicPenguin | Do we talk serial via a deamon, or do we open up /dev/ttyS0 |
| 18:13:48 | Lenolium | CosmicPenguin: Yes, but this is closer to /dev/sound, if you want multiple devices connected to /dev/sound, you need a daemon. |
| 18:14:12 | Russ|werk | that way, you don't need one app that does everything |
| 18:14:29 | Lenolium | Russ|werk: Bingo! Thank you. |
| 18:15:05 | CosmicPenguin | Lenolium: I see no reason why you can't let mutiple apps connect anyway |
| 18:15:39 | SpaceCoaster | BZFlag: if the DTMF dialing was in the kernel how would you do things like shortcuts? |
| 18:16:04 | CosmicPenguin | I don't want to have to stick 20K of IPC code onto a simple app that dials a number |
| 18:16:30 | Russ|werk | thats not all it does though |
| 18:16:33 | Lenolium | Sockets ain't exactly difficult. :) |
| 18:16:57 | Russ|werk | rings, audio paths, leds, caller id blocks, time updates, etc |
| 18:17:28 | Lenolium | I'm gonna step out of this argument, I'm pretty cranky, and it looks like Russ understands my reasoning very well. |
| 18:18:38 | CosmicPenguin | Russ|werk: True, but why not stick all that code in a library? |
| 18:19:28 | SpaceCoaster | Lenolium: i'm cranky today too, must be the weather |
| 18:20:28 | Russ|werk | CosmicPenguin: ok...so you have a library of functions |
| 18:20:30 | Russ|werk | then what |
| 18:20:59 | Russ|werk | you make a monolithic application that uses the functions? |
| 18:21:29 | CosmicPenguin | Russ|werk: Doesn't matter.... we were just headed down a patch that was very one sided, and I wanted to insert a little genericness into the situation |
| 18:22:26 | CosmicPenguin | Lets just get the thing written, and we can complain about it later |
| 18:22:27 | Russ|werk | unix sockets aren't the only way to communicate with the daemon |
| 18:22:41 | Russ|werk | but are a simple way |
| 18:24:07 | CosmicPenguin | Russ|werk: Theres alot of complexity in there..... |
| 18:33:14 | SpaceCoaster | CosmicPenguin: yay the DSP works we have something to argue about!! |
| 18:33:52 | SpaceCoaster | CosmicPenguin: it was much better with the touchscreen I went around grinning for three days! |
| 18:34:24 | Niddix | I should stop watching.. I'm going to go nuts waiting for Tim to ship mine. |
| 18:35:38 | SpaceCoaster | Niddix: He ships pretty quickly after you pay him, have you tried that :-) |
| 18:36:03 | Niddix | I payed him today.. (I'm not a very patient person) |
| 18:36:09 | Niddix | instant gratification and all that. |
| 18:36:17 | SpaceCoaster | aaahh |
| 18:37:18 | Lenolium | Niddix: I understand that, I went down to his office and picked them up directly from him. |
| 18:38:10 | Niddix | I'm in Florida.. not really an option. |
| 18:38:37 | Niddix | I am considering asking him to do overnight instead of priority. |
| 18:39:13 | CosmicPenguin | SpaceCoaster: Are you working on the changes to the spec now? |
| 18:43:19 | Lenolium | Ouch, the cost on an overnight of one of those would be pretty huge. |
| 18:43:43 | GoRK | Niddix: phones are only shipped two days a week |
| 18:45:10 | SpaceCoaster | CosmicPenguin: talking to me again eh |
| 18:45:41 | Niddix | Yeah.. but 1 day from tommorrow is better than 3 ;) Anyhow.. I'll just have to wait. $35 for shipping is a bit much. |
| 18:45:49 | SpaceCoaster | GoRK: talk to CosmicPenguin, cheer him up, tell him how great our DSP work is! |
| 18:45:52 | CosmicPenguin | SpaceCoaster: Was I not? |
| 18:46:01 | GoRK | SpaceCoaster: is he in the dumps? |
| 18:46:07 | GoRK | CosmicPenguin: are you in the dumps? |
| 18:46:18 | SpaceCoaster | CosmicPenguin: yeah giving everyone grief today! |
| 18:46:25 | GoRK | cosmicpengiun: i wrote a phone daemon yesterday that lets you use the phone :-/ err yeah |
| 18:46:31 | CosmicPenguin | I'm just angry that everybody assumes that if you arn't using a NFS mount and X then you shouldn't have any say in how applications are written |
| 18:46:57 | nanoprobe | Is ther an email i can order a tuxscreen from? |
| 18:47:46 | GoRK | SpaceCoaster: you have to compile that against the new version of wheaties.h that I sent along with it. There are some bugs and missing stuffs in the old one |
| 18:47:56 | Niddix | nanoprobe: go to http://www.tuxscreen.net/wiki/view/Users/TimRiker |
| 18:48:12 | CosmicPenguin | GoRK: You have a new patch ready to go? |
| 18:48:23 | GoRK | CosmicPenguin: nfs and x are no fun. that is way too easy |
| 18:50:13 | GoRK | CosmicPenguin: well i wrote a small little app that uses the dsp driver to make the tux behave like a telephone (with dtmf and volume and LED's and things) |
| 18:50:29 | CosmicPenguin | GoRK: Cool..... |
| 18:51:00 | nanoprobe | Is it possible to order a tuxscreen to sweden then? |
| 18:51:09 | cory | heh, apple just called me up trying to get me to buy an applecare protection plan.. |
| 18:51:25 | SpaceCoaster | CosmicPenguin: GoRK speaks fluent DSP, last week he mostly spoke in 2 digit hex strings, say something GoRK? |
| 18:51:56 | CosmicPenguin | GoRK: Was it graphical? |
| 18:52:06 | GoRK | a6 02 a7 02 |
| 18:52:20 | GoRK | CosmicPenguin: no it's going to be a daemon |
| 18:52:30 | CosmicPenguin | GoRK: And thus my problem |
| 18:52:42 | GoRK | ? about graphics? |
| 18:53:02 | CosmicPenguin | Sure...... 20K is a whole app for me... I would rather not waste it by trying to talk to a deamon..... |
| 18:53:06 | GoRK | no a graphical app will attach to it via unix socket to recieve status messages and send events to it |
| 18:53:26 | Russ|werk | CosmicPenguin: why would unix sockets cost you 20k? |
| 18:53:45 | CosmicPenguin | Ok.... lets discuss unix sockets for a moment.... |
| 18:54:03 | Russ|werk | aren't they allready in uclibc? |
| 18:54:03 | GoRK | well the problem is that the daemon has to control operations of the phone as well.. the idea is that this will eventually become a *kernel* daemon and may provide a /dev/phone0 device .. |
| 18:54:05 | CosmicPenguin | Just for giggles.... how many clients would you expect would be connected to the deamon at anone one time |
| 18:54:26 | CosmicPenguin | It can't be datagram, because you have unsolicited events.... |
| 18:54:33 | Russ|werk | 2 or 3 |
| 18:54:44 | GoRK | CosmicPenguin: then just take the code and put graphics around it.. no big deal |
| 18:54:46 | CosmicPenguin | Russ|werk: I would see it be much higher.... |
| 18:55:01 | GoRK | CosmicPenguin: right now there's no input/output except from the dsp at all anyway |
| 18:55:12 | Russ|werk | CosmicPenguin: you would request hardware, use it, release it |
| 18:55:19 | Russ|werk | CosmicPenguin: why? |
| 18:55:36 | Russ|werk | CosmicPenguin: why would calling a function that uses sockets add size? |
| 18:55:36 | CosmicPenguin | Russ|werk: But if I wanted an application to pop up when the phone rings, I would need to stay connected |
| 18:56:02 | CosmicPenguin | Russ|werk: Because of the various managment queues and other bullshit that needs to be attached to a non threaded application that needs to manage multiple connections.... |
| 18:56:25 | CosmicPenguin | Now, I have no problem with that...... |
| 18:56:29 | GoRK | CosmicPenguin: so change the daemon itself. |
| 18:56:52 | Russ|werk | CosmicPenguin: right, so you'd just connect, and request ring notification |
| 18:57:02 | CosmicPenguin | Russ|werk: And? |
| 18:57:40 | Russ|werk | when a ring comes through, the daemon looks on its linked list of who want ring notification |
| 18:57:46 | Russ|werk | and sends a packet to each one |
| 18:57:57 | GoRK | and then wait for a ring. it's much better than trying to have your graphical application trying to do the dsp watchdogging and whatnot |
| 18:58:13 | CosmicPenguin | GoRK: Why? Whats the difference? |
| 18:59:23 | CosmicPenguin | Anyway, its not worth arguing about... we we actually get some code, then we can evaulate it at that point |
| 19:00:00 | GoRK | CosmicPenguin: there isnt much. if you want to make it graphical, put in some nano-x display routies. if i were going to write a graphical app i'd write it for picogui anyway and there would still be a problem |
| 19:00:13 | GoRK | cosmicpengiun: i can email it to you |
| 19:00:37 | CosmicPenguin | GoRK: True... but I asked for a library and people jumped down my throat... thats what I'm trying to prevent here... If I wanted intolerance for other viewpoints, I would hang out in handhelds.org |
| 19:00:58 | Russ|werk | I'm just confused as to what the library would do |
| 19:01:23 | Russ|werk | its like saying I want a circus, why do I want a circus |
| 19:01:39 | CosmicPenguin | Anyway, thats over... Gork, feel free to e-mail me anything you want |
| 19:01:44 | Lenolium | How would a library help you if you wanted a window to pop up when a call came in? Wouldn't you have to sit around all day on /dev/wheaties poll() ing it anyway? |
| 19:01:56 | GoRK | i thought about a library... the only problem is that all you really need are the #defines to determine a more readable method to send/recieve bytes from the dsp |
| 19:02:14 | CosmicPenguin | GoRK: Thats another battle.... :) |
| 19:02:42 | GoRK | cosmicpengiun: if you want this code, i'll email it to you. I was just interested in picking up my phone and being able to dial. |
| 19:03:05 | CosmicPenguin | GoRK: Send out anything you've got... code is good |
| 19:03:06 | GoRK | CosmicPenguin: i figured i'd save the rest for another day... ie. callbacks/graphics/whatever |
| 19:03:57 | GoRK | ok ill just send this to the list |
| 19:04:15 | Russ|werk | you'd probably want the daemon doing a good deal of abstraction |
| 19:04:37 | CosmicPenguin | Russ|werk: Yep |
| 19:04:41 | SpaceCoaster | GoRK: you must have lots of updates for CosmicPenguins spec? |
| 19:04:41 | GoRK | if you can make it into a library, that's great. if you want to make it into a standalone app, go for it |
| 19:05:33 | CosmicPenguin | Nah.... a deamon should cover it.... |
| 19:05:51 | GoRK | SpaceCoaster: yeah, a couple.... i also found out that tone_click played through the handset seems to not work so well .. it really screws up the audio.. i took it out of wheaties.h |
| 19:06:45 | CosmicPenguin | GoRK: Which tone was that? |
| 19:07:06 | GoRK | 0x0a or something.. i forget. i already took it out :) |
| 19:07:26 | CosmicPenguin | That would be medium beep.... odd.... |
| 19:07:33 | CosmicPenguin | Does it work on the speaker? |
| 19:07:48 | GoRK | no hang on lemme look |
| 19:08:05 | GoRK | 0x0b |
| 19:08:21 | GoRK | it was not 'official' it was just one higher than 0x0a and it made a little click sound |
| 19:08:22 | CosmicPenguin | GoRK: Where did you get *that* tone????? :0 |
| 19:08:52 | GoRK | i can't find the tone that gets played in the failsafe mode right after you push a key (to confirm keypush for pulse dialing) |
| 19:09:48 | CosmicPenguin | Hmm.... I don't know if these tones are the true tones for the POTS, or if they are just for the local sounds.... |
| 19:10:06 | GoRK | CosmicPenguin: which ones? |
| 19:10:12 | CosmicPenguin | 0x42 sounds |
| 19:11:15 | GoRK | i think they are just for local sounds... but if you ask for really high numbered tones (off the charts) the dsp makes some weird modem-like tones. Perhaps those are actually used in modem handshaking |
| 19:11:35 | CosmicPenguin | GoRK: Wierd.... |
| 19:11:47 | CosmicPenguin | I wouldn't think that would be on this particular command..... |
| 19:12:43 | GoRK | i dunno. it's probably just tyring to access a tone from some memory that doesnt contain one and thus the bizzare noises |
| 19:13:48 | GoRK | SpaceCoaster: did you ever get it to compile? |
| 19:17:18 | Lenolium | GoRK: You took the click out of the .h? blast, I guess my daemon will have to stop being annoying. |
| 19:18:17 | GoRK | Lenolium: yeah it's not a good sound. if you play it to the handset as a feedback tone... well first you can't even hear it on the handset and second it kills handset audio for some reason. playing through the speaker seems OK, but since it's not documented, i took it out :-/ |
| 19:19:07 | CosmicPenguin | GoRK: It should definately *not* be considered official.... |
| 19:19:26 | GoRK | i figured it might not have been a click sound at all but maybe just pulled voltage high or low on the speaker causing a nice little click |
| 19:19:50 | GoRK | anyway i posted the phone program to the mailing list. it should be rolling in for folks now |
| 19:20:33 | SpaceCoaster | GoRK: not properly did you send me an update/ |
| 19:21:19 | GoRK | SpaceCoaster: well the old wheaties.h i sent you was incomplete. i should have included the newer one with that email i sent last night.. anyway the one i just posted to the ML contains the newest versions of both files |
| 19:21:29 | SpaceCoaster | cool |
| 19:21:47 | CosmicPenguin | GoRK: Looking good..... |
| 19:22:26 | GoRK | CosmicPenguin: really it's just a skeleton app anyway. take out the debug and you could make it easily into a graphical app, library, daemon, anything... |
| 19:22:53 | hypaStyle | Gork: do you have the source posted somewhere ? |
| 19:23:11 | GoRK | SpaceCoaster: did you notice that inferno's driver does some special casing for reading speakerphone param/states data? |
| 19:23:13 | CosmicPenguin | GoRK: No, its a good start.... It appears that the winds of change want a deamon, so who am I to argue with the community? |
| 19:23:58 | GoRK | CosmicPenguin: well i just figured writing a daemon would be a good start since it could move into kernelspace and provide a proper /dev/phone device |
| 19:24:13 | CosmicPenguin | GoRK: Thats what I want.... :) |
| 19:24:35 | GoRK | heh well it's going to take a kernel daemon imo to get /dev/phone |
| 19:24:39 | SpaceCoaster | yes, long delays etc. |
| 19:24:41 | CosmicPenguin | I wanna see human parsable messages coming out of /dev/wheaties.... |
| 19:25:08 | CosmicPenguin | But for the moment, we'll go with what we've got..... |
| 19:25:15 | CosmicPenguin | GoRK: You want me to post this along with the spec? |
| 19:25:18 | GoRK | they are totally human parseable. i can read them :) |
| 19:25:54 | GoRK | CosmicPenguin: well i was gonna put it in tuxscreen cvs once i got a couple more features (like ringing) implemented and it becomes useful |
| 19:25:56 | CosmicPenguin | GoRK: I'm talking normal humans here.... I'd bet *you* could just use the phone by holding on to the bare wires and talking |
| 19:26:47 | GoRK | my friend wanted to hear happy tone the other day so i used buzby to crank out a long string of hex that would offhook the phone on handset, set volumes, and play happy tone down the line :) |
| 19:26:59 | CosmicPenguin | GoRK: Ok, that works... |
| 19:28:05 | CosmicPenguin | GoRK: I sent the spec source file to SpaceCoaster, so if you have any bugs / enhancements will you let one of us know? |
| 19:28:28 | SpaceCoaster | GoRK: would a kernel fallback which enabled DTMF dialling and keep alives suffice, it would drop out if a superior daemon program attached itself |
| 19:28:44 | GoRK | yeah ill have to go back over it |
| 19:29:00 | SpaceCoaster | CosmicPenguin: just sent it back |
| 19:30:27 | GoRK | SpaceCoaster: *shrug* i would just think that it'd still be better in a userspace app for now... i dunno. ill code for whatever interface i get :) |
| 19:31:04 | CosmicPenguin | SpaceCoaster: Ok, I'll stick that up in a second.... |
| 19:31:15 | SpaceCoaster | CosmicPenguin: cool |
| 19:32:14 | prpplague | man these progear units suck |
| 19:32:57 | Lenolium | Grrf. /dev entries should not mangle the data. I should look the same as it came out of the device. I havn't noticed /dev/psaux spitting back human-readable location values lately. |
| 19:33:01 | CosmicPenguin | Posted..... |
| 19:33:30 | CosmicPenguin | Lenolium: The kernel should provide the appropriate information for the widest audience.... |
| 19:33:46 | SpaceCoaster | GoRK: I was just trying to find some middle ground. I can see BZFlags point that when the phone boots it should work, whatever distribution etc is installed. |
| 19:34:03 | GoRK | yeah i am a /dev/purist also ... unless there's some reason to do it (ie. touchscreen calibration) they shouldnt do anything to the data.. if they do, they also need to provide a raw device that doesnt |
| 19:34:14 | CosmicPenguin | SpaceCoaster: I think thats the biggest thing to worry about... |
| 19:34:21 | GoRK | SpaceCoaster: if we want to mangle the data in the kernel, lets write up a kernel daemon that presents a /dev/phone device :) |
| 19:34:37 | Lenolium | It's way easier to read bits, and translate internally, than to read strings, and translate that into something that your program understands. |
| 19:34:38 | CosmicPenguin | GoRK: Thats the best idea yet.... |
| 19:34:52 | CosmicPenguin | Lenolium: not in shell scripts its not |
| 19:34:57 | SpaceCoaster | Hmmm what apps use the phone device? |
| 19:35:05 | CosmicPenguin | SpaceCoaster: What do you mean? |
| 19:35:15 | GoRK | shell scripts on busybox ash ar slow as my palm III |
| 19:35:29 | Lenolium | GoRK: Actually, a /dev/dsp with some funny ioctl()'s would probably fit better. :) |
| 19:35:44 | GoRK | Lenolium: that is what /dev/phone is anyway |
| 19:35:50 | SpaceCoaster | CosmicPenguin: is the /dev/phone a Linux Telephony thing, if so what does that buy us? |
| 19:35:52 | Russ|werk | I would just think a serial like device with a reset ioctl |
| 19:35:54 | Lenolium | GoRK: Ahh.. -k- |
| 19:35:55 | GoRK | it's an audio device with ioctls that do phone stuff |
| 19:36:07 | Lenolium | Russ|werk: Luckily, that is what we have right now. |
| 19:36:16 | CosmicPenguin | SpaceCoaster: I have no idea if there are any sorts of telephony precidents here.... |
| 19:36:39 | GoRK | SpaceCoaster: it buys us some apps that will work with it already. ie. the entire openh323 suite |
| 19:36:41 | CosmicPenguin | GoRK: It alot more complicated than that.... mainly because you can have unsolicited events |
| 19:37:04 | SpaceCoaster | GoRK: cool |
| 19:37:04 | CosmicPenguin | Gork, SpaceCoaster: If a specification already exists, then we should obey that.... |
| 19:37:07 | Lenolium | There is a pre-existing telephony API, but IIRC the devices still vary too widely to have much common between them. |
| 19:37:35 | CosmicPenguin | Then we should standardize to something like Gnophone.... |
| 19:37:40 | GoRK | the only spec is linux telephony device... there's also zapata/zaptel we could work with |
| 19:37:59 | CosmicPenguin | Or sorry, Asterisk....... |
| 19:38:06 | GoRK | gnophone has input channels. it doesnt care what the device looks like as long as you write an input channel for it.. right now it only supports soundcards |
| 19:38:09 | SpaceCoaster | CosmicPenguin: yes if it delivers benefits, like VOIP etc and doesn't cost too much in resources |
| 19:38:21 | GoRK | asterisk is the same way. the channel devices can work with any type of device |
| 19:38:35 | Lenolium | Going for a pure telephony device would be cool. I'm just afraid of how huge the kernel driver will get to get there. So, I like the current one for now, but a full-blown telephony driver would be cool later on. |
| 19:38:52 | CosmicPenguin | GoRK: Of course, thats a mute point, since we can't get arbitrary data from the DSP... |
| 19:39:04 | SpaceCoaster | Lenolium: could you point me to the applications that we would use? |
| 19:39:11 | CosmicPenguin | But if there is an existing standard for communicating with telephony devices, we should use that |
| 19:39:16 | GoRK | there are channel drivers for internet phone/linejack cards that technically use the telephony devices. The problem is that quicknet cards are so buggy that you can't just treat them as generic telephony devices |
| 19:39:59 | Lenolium | GoRK: Ahh.. so it would gain us nothing to turn it into a telephony device? |
| 19:40:50 | GoRK | well.. not really. it'd be easier to write a channel driver for gnophone that talks to whatever audio device we have to use and then directly to /dev/wheaties for things like keypad and switchook detection |
| 19:41:52 | GoRK | the main interfaces for asterisk are tormenta T1 cards (which are incedentally public domain hardware devices) use zaptel/zaptata |
| 19:42:15 | GoRK | and there's a chan_tormenta.so in asterisk |
| 19:45:35 | CosmicPenguin | SpaceCoaster: Still waiting..... :) |
| 19:45:58 | SpaceCoaster | hhmm, |
| 19:47:54 | GoRK | gnophone would require quite a bit of retooling anyway to work with the windowing systems that we can actually fit into flash |
| 19:48:08 | CosmicPenguin | GoRK: Nahh..... :) |
| 19:48:46 | SpaceCoaster | Hmm i sent it, I didn't change the Subject, it's re: DSP Driver Status, my mailq is empty and the logs say it's gone! |
| 19:49:19 | SpaceCoaster | GoRK: do we have any info about the speaker phone parameters? |
| 19:50:06 | GoRK | SpaceCoaster: there is state data in the inferno userspace stuff. ie. the raw hex state data that gets downloaded. that may be all we need. i dont know though. |
| 19:50:37 | SpaceCoaster | CosmicPenguin, GoRK: could you give me a pointer to the telephony, kernel interface, sounds interesting! |
| 19:50:41 | CosmicPenguin | GoRK: Did you try to download that? |
| 19:50:44 | GoRK | SpaceCoaster: there's also the special case stuff in the inferno dsp driver that makes the device wait longer when sending out state/param data |
| 19:51:15 | GoRK | CosmicPenguin: yeah i have tried. i know it must do something about echo cancellation since bringing up the speakerphone interface now results in a lot of feedback |
| 19:51:23 | SpaceCoaster | yeah, I could snooze for long messages, I'll look at it. |
| 19:52:22 | Lenolium | Ok, theoretically, I now have the daemon done. |
| 19:52:25 | CosmicPenguin | Apparently the data needs to be filtered on input |
| 19:52:40 | CosmicPenguin | Lenolium: How different is it from GoRK's work? |
| 19:53:55 | Lenolium | CosmicPenguin: Mine does the bare minimum with the phone, it will only keep it alive, it relies on everything going through external applications. You can connect to the daemon and tell it what events you are interested in, and it will notify you when they come up. |
| 19:54:07 | Lenolium | (you can also send data to the DSP via the daemon) |
| 19:54:19 | CosmicPenguin | So basincally its a kernel driver that lives in user space |
| 19:54:21 | Lenolium | Oh, and this is also totally untested. But hey, it compiles :) |
| 19:54:56 | GoRK | the kernel driver should not have the job of keeping the dsp alive. if your userspace phone app crashes, the watchdog timer does need to expire |
| 19:55:12 | CosmicPenguin | Does the DSP need to be kept alive? |
| 19:55:19 | SpaceCoaster | YES!!!! |
| 19:55:28 | CosmicPenguin | No joke? |
| 19:55:32 | GoRK | yes. you have to send it sanity checks every 1 second and if it responds that it needs a reset you have to reset it |
| 19:55:43 | CosmicPenguin | Wow..... |
| 19:55:51 | CosmicPenguin | That should be in the spec.... how did I miss that? |
| 19:55:57 | SpaceCoaster | You missed a long discussion, GoRK discovered it, GPSFan found a watchdog timer on the DSP |
| 19:55:59 | GoRK | it's not in the spec :) |
| 19:56:22 | CosmicPenguin | Yeah, but its in the inferno code |
| 19:56:46 | GoRK | well it's sort of in the inferno code |
| 19:57:04 | GoRK | i think the apps written in limbo actually do the work (which we dont have sources to) |
| 19:57:22 | CosmicPenguin | They definately define a task for it.... |
| 19:57:38 | CosmicPenguin | Well, anyway, will someone give me a blub for the spec.... |
| 19:57:51 | Lenolium | Oh, wait. This won't do, there is no way for you to unbind stuff you are interested in. |
| 19:58:23 | GoRK | the periodic task also checks if it needs to reset the clock and looks at battery state. that's all i thought it initialy did |
| 19:58:33 | CosmicPenguin | Same here... |
| 19:59:26 | SpaceCoaster | CosmicPenguin: the rational for the keep alives is that if the kernel poops, the daemon dies the UI hangs you can still call 911 |
| 19:59:34 | GoRK | CosmicPenguin: The DSP has a watchdog timer that expires after about 3 seconds. Inferno sends sanity check commands every one second. If the response to the sanity check command is 03, the dsp must be reset |
| 20:00:21 | GoRK | well if the kernel or anything poops. I think the watchdogging should be the responsibility of the last guy on the chain... if it was in the kernel and userspace died, it'd still be kept alive and unable to dial anything |
| 20:01:18 | GoRK | it seems to me that in inferno, there's not a big distinction between kernelspace/userspace |
| 20:01:37 | GoRK | either that or the periodic task exists in userland |
| 20:01:56 | CosmicPenguin | GoRK: Added that to the SPEC under the description for SANITY_CMD |
| 20:02:01 | Lenolium | GoRK: I think the periodic task is purely userland. |
| 20:02:04 | GoRK | CosmicPenguin: cool |
| 20:02:29 | GoRK | Lenolium: i would think that it's the only thing that would make sense |
| 20:02:49 | CosmicPenguin | Any other big things that I should know about? :) |
| 20:03:05 | CosmicPenguin | GoRK: Its an RTOS... its all one big process with lots of threads.... |
| 20:03:27 | GoRK | CosmicPenguin: ahh. ok well that makes sense then. |
| 20:04:01 | GoRK | CosmicPenguin: not until we get speakerphone param/state data working... i belive you have to do those two commands very carefully, and that it would be appropriate to put the steps in the spec |
| 20:04:17 | SpaceCoaster | Yeah they didn't have this discussion, it was all just another task. |
| 20:04:49 | CosmicPenguin | GoRK: I agree... though I think that the Inferno folks wern't too sure about it either.... |
| 20:06:24 | GoRK | did anyone else get my little phone program to work? |
| 20:06:45 | CosmicPenguin | Damnit... That SANITY check was in the periodic task.... why the hell did I not see that? <bangs self over head with large book> |
| 20:07:37 | SpaceCoaster | CosmicPenguin: how long were you looking at this stuff for? |
| 20:07:57 | CosmicPenguin | SpaceCoaster: A while... but obviously not long enough |
| 20:08:13 | SpaceCoaster | CosmicPenguin: days, weeks, months or years? |
| 20:08:24 | CosmicPenguin | SpaceCoaster: Bout a week.... |
| 20:08:41 | CosmicPenguin | SpaceCoaster, GoRK: Did you see any issues with the brightness / contrast as you used the DSP? |
| 20:08:59 | GoRK | CosmicPenguin: to be honest, i hadnt even tried |
| 20:09:14 | CosmicPenguin | It seems that there may be some corruption there...... |
| 20:09:14 | SpaceCoaster | They don't conflict at all, do they? |
| 20:09:23 | as_ | Evnin all. |
| 20:09:57 | SpaceCoaster | CosmicPenguin: why would you expect any? |
| 20:10:13 | CosmicPenguin | *shrugs* Who knows? |
| 20:11:07 | SpaceCoaster | true! |
| 20:11:22 | SpaceCoaster | wait, we do! |
| 20:11:33 | CosmicPenguin | Do we? |
| 20:14:39 | SpaceCoaster | Good grief man, what happened to the old positive penguin. |
| 20:15:13 | GoRK | posipenguin |
| 20:15:24 | SpaceCoaster | penguin++ |
| 20:15:54 | GoRK | what are the issues? i thought you had to reset the dsp to get the controls to work? |
| 20:15:57 | CosmicPenguin | Sorry, I shouldn't be *that* negative...... |
| 20:18:48 | Lenolium | Alright, now the daemon supports unbinding watches, and will do that on any lost connections by default. |
| 20:18:59 | Lenolium | Now, I just need to write a simple client app |
| 20:19:14 | GoRK | Lenolium: why dont you convert my phone program into a client? |
| 20:19:33 | Lenolium | GoRK: Because I'm really sleepy. :) (I only got three hours last night.) |
| 20:20:04 | Lenolium | GoRK: But if you want to do it, I would be more than happy to anwser any questions that may arise. :) |
| 20:28:17 | GoRK | how do i request to get all events? |
| 20:30:15 | Lenolium | Send a message to the daemon with a message of WHEATIESD_BIND, 0x00 |
| 20:30:29 | Lenolium | GoRK: Let me send you the latest daemon, and .h |
| 20:30:41 | GoRK | Lenolium: ok thanks |
| 20:34:46 | Lenolium | Sent. |
| 20:34:51 | GoRK | thanks |
| 20:35:21 | Lenolium | Np. |
| 20:35:41 | Lenolium | I'm gonna go cook some ramen, so if you want me for the next few minutes, type really loudly. ;) |
| 20:35:55 | GoRK | heh im not going to get to this till the evening |
| 20:36:04 | GoRK | so dont wait on me :) |
| 20:37:31 | prpplague | hmm ramen..., best meal you can buy for $.08 |
| 20:38:13 | signal11 | mmmm. one weeks' supply of sodium in every packet |
| 20:39:03 | AlHazWork | prpplague: I came across a website detailing the history of insta-ramen a few months ago |
| 20:39:24 | AlHazWork | prpplague: ironically, when it was introduced in japan, it didn't take off because it was /more expensive/ than fresh ramen |
| 20:40:13 | prpplague | that is interesting |
| 20:40:19 | GoRK | tonight, AUDIO (i hope) |
| 20:40:24 | GoRK | Russ|werk: are you still here? |
| 20:40:46 | CosmicPenguin | GoRK: That takes courage.... :) |
| 20:41:07 | AlHazWork | prpplague: aparantly, in japan, there were, and often still are, ramen stores & carts all over the place |
| 20:41:34 | AlHazWork | so nearly anyone could just walk down the street and buy some fresh made |
| 20:42:04 | AlHazWork | my friends who've been to japan assure me that insta-ramen sucks compared to the real thing |
| 20:42:06 | Lenolium | GoRK: How are you planning on doing audio? |
| 20:42:56 | GoRK | ucb1200 |
| 20:43:17 | GoRK | and attaching relays to the unused LED control pins |
| 20:43:18 | GoRK | of the dsp |
| 20:43:36 | GoRK | to switch the audio input off of the dsp and onto the ucb1200 |
| 20:43:52 | prpplague | AlHazWork: i always fix it with something else thought, tomato sauce, cream of chicken |
| 20:44:35 | GoRK | AlHaz: yes real ramen is nothing like that crap in the store |
| 20:45:09 | RobFlynn | Did someone say ramen? |
| 20:45:10 | Lenolium | I had a chance at real ramen once, I turned it down, since I didn't want to know how bad the ramen I was eating really was. :) |
| 20:45:16 | GoRK | but even in japan, people eat that crap too :) |
| 20:45:30 | Russ|werk | yes GoRK? |
| 20:45:43 | GoRK | Russ|werk: do you know anything about the ucb1200 audio driver in the kernel? |
| 20:45:47 | RobFlynn | When I lived in Irvine, CA, I worked about a mile from a place that makes Ramen :) |
| 20:45:56 | RobFlynn | It was a huge factory. It frightened me. |
| 20:46:05 | RobFlynn | I started thinking about how many college kids were being fed. ;-) |
| 20:46:28 | GoRK | Russ|werk: i probably asked you that before :-/ |
| 20:46:32 | Russ|werk | GoRK: it uses a different major/minor than /dev/dsp |
| 20:47:09 | GoRK | Russ|werk: so it is proprietary? but should it work with our chip without modification? |
| 20:47:47 | BZFlag | hmm... is the buildtree broken? I get: drwxr-xr-x 1 0 0 0 Oct 22 14:53 /etc//init.d/ |
| 20:47:47 | BZFlag | jffs2reader: /etc//init.d: No such file or directory |
| 20:47:49 | Russ|werk | I think the new ucb1x00 driver uses /dev/dsp though |
| 20:48:02 | Russ|werk | I have a two kernel patches now that don't boot quite right |
| 20:48:16 | GoRK | Russ|werk: for newer kernels? i suppose i should just wait on that |
| 20:48:24 | Lenolium | http://www.tuxscreen.net/wiki/view//MonGoose <-- Is that the picture of the actual mongoose board on our thing, or is it missing some chips (such as flash) |
| 20:48:26 | CosmicPenguin | RobFlynn: I thought that the UCB1200 audio takes a detour through the DSP |
| 20:48:31 | GoRK | since the whole ucb1200 stuff has changed... |
| 20:48:57 | RobFlynn | redirects that to Russ|werk |
| 20:48:58 | Russ|werk | the second one uses the new ucb1x00 driver (new backlight ioctl format, ts driver, audio driver, etc) |
| 20:48:58 | CosmicPenguin | Course, it shouldn't.... |
| 20:49:14 | GoRK | CosmicPenguin: afaik nobody has been able to show that the ucb1200 audio interface is connected to anything at all |
| 20:49:24 | Russ|werk | seems to have some mtd issues |
| 20:49:42 | BZFlag | Lenolium: where did that image come from? |
| 20:50:07 | Lenolium | BZFlag: I thought you were in charge of that stuff. |
| 20:50:07 | CosmicPenguin | GoRK: It should be connected through the MCP, right? |
| 20:50:17 | BZFlag | Lenolium: I didn't add that one. |
| 20:51:05 | Lenolium | BZFlag: Oh, well then, I don't know, it just showed up today. |
| 20:51:26 | GoRK | Lenolium: actually... i can tell you that it is the acual mongoose board from a shannon. i looked under mine with a dental mirror |
| 20:51:48 | Russ|werk | GoRK: to make the modem work, it would have to |
| 20:51:59 | SpaceCoaster | Lenolium: GPSFan took the picture and it does look like the chips are an option we didn't get. |
| 20:52:02 | GoRK | Lenolium: i think that GPSFan put it up. |
| 20:52:14 | BZFlag | KenMcguire put the page up, don't know where the picture came from |
| 20:52:30 | GoRK | im sure he took it since he's doing all those schematics |
| 20:52:31 | SpaceCoaster | KenMcguire==GPSFan |
| 20:53:02 | GoRK | i hope that this means he's doing mongoose/ucb1200 schemes next :))) |
| 20:53:05 | Lenolium | Hrm, so I guess our mongoose doesn't have any flash? Wait... is the main chip on the mongoose the wheaties? |
| 20:53:25 | GoRK | Russ|werk: yeah that is a good point.. perhaps that is where that other bizzare audio destination goes to |
| 20:53:25 | SpaceCoaster | DSP1653, mask programmed prom |
| 20:53:32 | SpaceCoaster | ^prom^rom |
| 20:53:54 | GoRK | Lenolium: perhaps the missing chips on mongoose are what that board uses when it is indeed in a digital answering machine or something |
| 20:54:16 | CosmicPenguin | GoRK: Thats the WH/CH destinations you are talking about? |
| 20:54:32 | CosmicPenguin | W could stand for Wheaties and C could stand for codec... :) |
| 20:54:32 | GoRK | CosmicPenguin: oh i know that CH goes to the handset |
| 20:54:50 | CosmicPenguin | GoRK: More bugs for the spec that you didn't tell me about??????? :) |
| 20:55:01 | GoRK | or that's the only way i can get a tone to play on the handset without it also going out on the wire |
| 20:55:21 | SpaceCoaster | CosmicPenguin: Gork knows it all, grill him |
| 20:55:49 | CosmicPenguin | GoRK: Any other bugs? |
| 20:56:02 | GoRK | CosmicPenguin: where is the current verison? ill go read it |
| 20:57:14 | Lenolium | GoRK: Now what is WH?.... |
| 20:57:20 | GoRK | dunno |
| 20:57:26 | GoRK | maybe ucb1200 |
| 20:58:33 | GoRK | if someone could point me to a program that can playback using our current version of the ucb1200 audio driver or give some visual indication of audio data coming IN, i can start looking at some of this stuff |
| 20:59:47 | prpplague | hmm, i think i'm going to pervet the buildroot for use with some of other embedded stuff... |
| 21:00:55 | GoRK | prpplague: buildroot-tux is actually perverted from another buildroot project... ftp://oss.lineo.com/ buildroot.tar.gz |
| 21:01:35 | prpplague | GoRK: sounds like its getting a little too perverted |
| 21:02:23 | GoRK | CosmicPenguin: ok on the LED's byte 3, the new call led is bit 5 and 6 not bit 7 and 8 |
| 21:02:30 | prpplague | GoRK: we've been using shell scripts to do something similiar, but our's isn't as nice as buildroot |
| 21:03:45 | prpplague | collord: fedex assures me that the ceiva will be in tomorrow |
| 21:04:10 | AlHazWork | that reminds me, I have to visit the ups dock tonight |
| 21:04:40 | GoRK | CosmicPenguin: you might want also to note that led byte 4 bit 7 and 8 control the relay which actually hooks the phone line to the dsp. you have to turn it ON when you want to take the phone off-hook on either handset or speakerphone. im not sure if it has to be on to detect incoming rings |
| 21:06:01 | prpplague | hmm, who is maintaining the buildroot on the tuxscreen site? |
| 21:06:23 | collord | prpplague: though that's the slowest fedex i've ever seen... great! |
| 21:06:58 | prpplague | collord: i had them send i commercial ground, its really cheap, but doesn't get a guarantee on delivery date |
| 21:07:00 | CosmicPenguin | GoRK: New call is 5 & 6, got it |
| 21:07:50 | GoRK | the ring responses are very strange |
| 21:07:52 | prpplague | BZFlag: you there? |
| 21:10:12 | prpplague | BZFlag: the buildroot version 0.4 on the tuxscreen.net site has a problem because it can't get the file mtd-snapshot-20010928.tar.bz2 doesn't appear to be available anymore |
| 21:11:54 | CosmicPenguin | GoRK: Ok, I got that in..... |
| 21:13:36 | CosmicPenguin | GoRK: I'm going to play with your app and see if I can't do some graphical things to it.... |
| 21:13:45 | CosmicPenguin | Course, you have the advantage because you have a widget set.... |
| 21:13:53 | GoRK | CosmicPenguin: cool |
| 21:14:07 | CosmicPenguin | And unfortunately, I can't get anyone to help me.... (ahem, signal11 and bowb.... I'm looking at you) |
| 21:14:53 | CosmicPenguin | GoRK: I assume that you will put it on PicoGUI soon as well? |
| 21:14:58 | Russ|werk | http://russ.dhs.org/files/tuxpatches/patch-2.4.10-ac11-rmk1-np1-tux1.gz <- b0rked patch |
| 21:15:04 | GoRK | i suppose so :) |
| 21:15:22 | Russ|werk | some mtd problem it seems |
| 21:15:36 | Russ|werk | but it has the new ucb1x00 driver, and backlight ioctls |
| 21:15:58 | BZFlag | Russ|werk: what's up with the mtd stuff in buildroot? |
| 21:16:04 | Russ|werk | dunno |
| 21:16:05 | GoRK | Russ|werk: will it boot at all? |
| 21:16:22 | Russ|werk | GoRK: it locks in checkroot.sh, I think you can play with it |
| 21:16:39 | prpplague | i'm just going to use my sources that i know are good for now |
| 21:16:46 | BZFlag | prpplague: cvs up -PdA that will fix your problem and give you mine. |
| 21:17:07 | SpaceCoaster | BZFlag: do you know an app that uses the Linux Telephony stuff which we could use on the TuxScreen? |
| 21:17:18 | GoRK | SpaceCoaster: openh323 |
| 21:17:24 | BZFlag | <Russ|werk> seems to have some mtd issues <- what were you refering to? |
| 21:17:36 | SpaceCoaster | GoRK: thanks |
| 21:18:03 | BZFlag | also check linuxtelephony.org for pointers. |
| 21:18:07 | Russ|werk | BZFlag: the above kernel patch |
| 21:18:21 | CosmicPenguin | GoRK: Did you try CID by chance? |
| 21:18:40 | BZFlag | Russ|werk: the latest buildroot builds a jffs2 image that jffs2reader will no read. |
| 21:18:47 | BZFlag | s/no/not/ |
| 21:18:54 | Russ|werk | BZFlag: this is unrelated |
| 21:19:00 | BZFlag | are you having mount trouble? |
| 21:19:34 | Russ|werk | remount trouble |
| 21:19:54 | CosmicPenguin | Lenolium: Where be your code? |
| 21:20:13 | Lenolium | CosmicPenguin: Currently, I'm emailing it out to people... let me stick it someplace people can get at it. |
| 21:20:22 | CosmicPenguin | Lenolium: Send it to the list |
| 21:21:06 | Lenolium | http://www.rawb.org/~rawb/TUX/ |
| 21:21:35 | Lenolium | After I get back from the bank, I'm gonna write a client app, to test it out a bit. Once it passes (no testing has been done yet), I'll post it to the list. |
| 21:24:38 | SpaceCoaster | GoRK: the openh323 stuff is written in C++ is that a problem? |
| 21:24:49 | GoRK | probably |
| 21:24:57 | GoRK | compile it static for testing... |
| 21:25:12 | BZFlag | SpaceCoaster: or fix the uclibc loader ;-) |
| 21:25:14 | GoRK | against glibc |
| 21:27:39 | CosmicPenguin | GoRK: What version is your arm compiler? |
| 21:27:59 | GoRK | 2.95.2 |
| 21:28:31 | SpaceCoaster | GoRK: I see what you meant the telephony clients all specify the telephony hardware that they work with, not very device independent. |
| 21:28:52 | CosmicPenguin | So is mine..... you need that damn nscd to do DNS |
| 21:29:06 | CosmicPenguin | Which is why the ViewML browser only works for IP addresses now.... |
| 21:29:08 | GoRK | openh323 is pretty quicknet-specific |
| 21:29:26 | SpaceCoaster | lets put one of those in then :-) |
| 21:30:08 | GoRK | i dont know if they are 3.3V, but we could |
| 21:31:03 | Russ|werk | CosmicPenguin: you get c++ into uclibc yet? |
| 21:31:25 | CosmicPenguin | Russ|werk: Nope.... andersee asked my company if we would be willing to sponsor it, but I think we passed.... |
| 21:31:55 | prpplague | BZFlag: thanks for the cvs offer, but i'm firewall handicaped |
| 21:32:44 | CosmicPenguin | Ok, this DMCA crap on the LKML is getting out of hand.... |
| 21:34:29 | mallum | hmm, wheaties stuff not yet in buildroot cvs ? |
| 21:37:50 | GoRK | whoot. the phone rings now |
| 21:38:06 | mallum | cool :) |
| 21:38:06 | CosmicPenguin | GoRK: When you call it, or just rings on demand? |
| 21:38:22 | GoRK | when you call it |
| 21:38:28 | SpaceCoaster | GoRK: once a second, keep the user alive :-) |
| 21:38:35 | GoRK | i am adding the logic now to terminate the tone in the middle of the ring |
| 21:38:44 | GoRK | SpaceCoaster: lol |
| 21:38:56 | SpaceCoaster | Hold on my phone rings fine, what gives? |
| 21:39:45 | CosmicPenguin | GoRK: Terminate the tone??? |
| 21:40:14 | GoRK | if the ring voltage stops halfway through the phone playing the ringtone, then we have to stop the bell in the middle |
| 21:40:36 | CosmicPenguin | GoRK: None of my phones at home do that |
| 21:41:16 | GoRK | CosmicPenguin: they do if you have custom rings that go like ... short short long |
| 21:41:52 | GoRK | where 'long' is a normal ring duration |
| 21:42:32 | SpaceCoaster | GoRK: Ahh, is this a custom ring that you have implemented? |
| 21:42:33 | CosmicPenguin | GoRK: Ok.... why not? |
| 21:42:47 | AlHazWork | right, you could have the phone ring different if caller id is unknown, etc |
| 21:43:17 | CosmicPenguin | GoRK: How do you know if the you lose the voltage? Is that another change for the DSP? |
| 21:43:17 | GoRK | SpaceCoaster: no. this is just to stop the ring playing if voltage doesnt last for the entire duration of the ring... it also provides a place for us to tack on the priority beeps based on callerid |
| 21:43:36 | CosmicPenguin | Or rather, the DSP spec |
| 21:43:41 | GoRK | CosmicPenguin: the dsp ring alterter bit is 0 |
| 21:43:54 | GoRK | it issues a response when the voltage goes away |
| 21:44:03 | CosmicPenguin | On what command? |
| 21:44:12 | GoRK | RSP_RING 0x10 |
| 21:44:45 | AlHazWork | I hope all this talk of ring tones doesn't bring back my suppressed memories of having worked for a company that wrote voicemail & autodialer software |
| 21:45:02 | GoRK | haha |
| 21:45:21 | CosmicPenguin | Ahhh..... so, whats up with the special 0x00 sent 10 seconds after? |
| 21:45:45 | AlHazWork | GoRK: the way i recall it, the entire telecom industry is just like the old tv show "Taxi", except it's not funny. |
| 21:45:49 | GoRK | CosmicPenguin: dunno. i just ignore it. i can't think of a use |
| 21:46:55 | GoRK | i have a good motivation to get ringing right the first time.. dont need to run up my cellphone bill testing this stupidity |
| 21:46:59 | GoRK | :) |
| 21:47:24 | GoRK | yep it cuts it off in the middle now.. whoot |
| 21:47:36 | GoRK | callerid is next, but i need to take a break for a minute |
| 21:47:42 | CosmicPenguin | GoRK: I've got a million minutes if you need to pass the baton... |
| 21:48:17 | CosmicPenguin | GoRK: Also, get ready to release some bleeding edge code around 5:30 MST or so, so I can take it home with me |
| 21:48:45 | GoRK | sure thing. i have a dinner engagement then anyway |
| 21:48:50 | GoRK | ok bbiab |
| 22:10:41 | GoRK | lala |
| 22:10:46 | Lenolium | is writing bubzy, with an interface to the new daemon. |
| 22:10:46 | GoRK | callerid... coming right along |
| 22:11:10 | Lenolium | GoRK: Are you just extending your current program? or are you running through the daemon? |
| 22:11:55 | mallum | hmm, my buildroot is not buildrootin ! getting wierd asm errors |
| 22:12:25 | mallum | arm-linux-gcc -D__ASSEMBLY__ -D__KERNEL__ -I/home/mallum/Projects/TUX/buildroot-tux/linux/include -mapcs-32 -march=armv4 -mno-fpu -c -o entry-armv.o entry-armv.S |
| 22:12:25 | mallum | entry-armv.S: Assembler messages: |
| 22:12:25 | mallum | entry-armv.S:781: Error: Internal_relocation (type 187) not fixed up (OFFSET_IMM) |
| 22:12:33 | mallum | ... lots more of the above .. |
| 22:12:41 | mallum | any ideas ? |
| 22:14:31 | CosmicPenguin | mallum: make mrproper and try again |
| 22:16:40 | mallum | CosmicPenguin: thx, what does make mrproper do over make clean ? |
| 22:17:15 | Russ|werk | lots |
| 22:17:59 | mallum | heh |
| 22:18:08 | mallum | it makes things really clean ? |
| 22:18:19 | CosmicPenguin | mallum: I've noticed that if I forget to add the skiff toolchain to my path before starting a build, it really annoys the kernel |
| 22:18:27 | CosmicPenguin | mrproper is like mrclean - it makes everything proper... :) |
| 22:19:05 | Russ|werk | puts it in a virginal state |
| 22:19:20 | mallum | CosmicPenguin: yes that what I did .. |
| 22:19:24 | Russ|werk | like you just untarred it |
| 22:19:29 | mallum | ok |
| 22:19:43 | mallum | why the 'mr' though ? |
| 22:20:10 | GoRK | are you dissing on mister proper? |
| 22:20:17 | CosmicPenguin | Its like Mr Clean...... |
| 22:20:47 | mallum | from the mister men ;) ? |
| 22:21:00 | CosmicPenguin | Damn.... mallum is across the pond... :) |
| 22:21:05 | CosmicPenguin | They don't have Mr Clean in England? |
| 22:21:17 | mallum | roger hardgreeves hacks kernels ? |
| 22:21:39 | mallum | we dont have mister clean AFAIK |
| 22:22:38 | GoRK | i think i figured out why the dsp is having a freakout on volume changes |
| 22:22:52 | GoRK | but ill have to wait till after dinner to fix it |
| 22:22:53 | AlHazWork | hometime |
| 22:25:28 | CosmicPenguin | Well, then.... things are moving well..... |
| 22:25:47 | CosmicPenguin | Now we just need to convince andersee to help us with the C++ and we will be set... |
| 22:26:00 | CosmicPenguin | We will be bored, because there will be nothing more to do, but we will be set... |
| 22:30:32 | mallum | hmmm, now busybox failed to patch.... maybe andersee updated busybox cvs and not the buildroot busybox patch ? |
| 22:30:49 | prpplague | mallum: ya my bb failed too |
| 22:32:38 | mallum | prpplague: hmmm, Ill play with it tommorow - bed time now ... |
| 22:33:20 | CosmicPenguin | Maybe thats why he didn't show up today |
| 22:33:34 | mallum | heh |
| 22:33:58 | mallum | night all + congrats on all the DSP success ! |
| 22:34:07 | CosmicPenguin | Either that or he's headed to Mexico with Linus.... ;) |
| 22:35:37 | prpplague | i'm out of here too |
| 22:35:38 | prpplague | later |
| 22:58:50 | cory | anyone ever installed ydl2.1 or something like it on mac? |
| 22:59:58 | art | exit |
| 23:00:00 | art | exit |
| 23:08:12 | Lenolium | Wow. I can lock up my tuxscreen. |
| 23:28:47 | Lenolium | Ahh.. cool, I did a stupid mistake. |
| 23:58:11 | Lenolium | Alright. My daemon works now. A bunch of stupid little bugs, that did some pretty random stuff. |
