| 00:03.08 | *** join/#edev rustye (n=rusty@pool-71-123-136-184.dllstx.fios.verizon.net) |
| 00:03.41 | *** part/#edev rustye (n=rusty@pool-71-123-136-184.dllstx.fios.verizon.net) |
| 00:09.03 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 00:12.39 | prpplague | ahh nothing like enjoying the good weather |
| 00:12.48 | prpplague | http://www.weather.com/weather/map/76053?from=36hour_map&mapdest=/maps/local/local/us_close_dfw_ultra_dfw_metro_dfw/1b/index_large.html |
| 00:13.24 | RobotGuy | sighs |
| 00:14.58 | CosmicPenguin | prpplague: ah, its a red evening |
| 00:15.00 | CosmicPenguin | I love those |
| 00:15.37 | prpplague | hehe |
| 00:15.55 | CosmicPenguin | ugh - pretty bad band of storms all across the midwest |
| 00:15.56 | prpplague | ~seen rustye |
| 00:15.58 | purl | i haven't seen 'rustye', prpplague |
| 00:16.24 | prpplague | CosmicPenguin indeed, lighting and thunder right now |
| 00:16.31 | prpplague | is sitting on the porch |
| 00:17.44 | CosmicPenguin | goes home |
| 00:17.59 | RobotGuy | ~seen rustycan |
| 00:18.00 | purl | RobotGuy: i haven't seen 'rustycan' |
| 00:18.53 | prpplague | RobotGuy thought maybe that was rusty |
| 00:19.05 | RobotGuy | Building a 3D renderer is intense. |
| 00:19.19 | RobotGuy | I fiigured he would use the same nick as he has on the forums. |
| 00:22.52 | RobotGuy | Have you seen Spiderman 3? |
| 00:26.27 | RobotGuy | http://dorkbotpdx.org/blog/feurig/benito_my_first_at90usb162_project |
| 00:30.17 | g1powermac_PB | alright, time to crack open the zipit. . . |
| 00:31.36 | RobotGuy | YAY, I got OGRE 1.4.7 dist sources built! |
| 00:31.40 | MonMotha | hah, nice to see somebody else that thinks the AVR hardware is good, but Atmel's code is terrible |
| 00:32.04 | RobotGuy | MonMotha: What would you use besides AVRs? |
| 00:32.32 | prpplague | pics, 8051, z80, hc11, 68000 |
| 00:32.55 | prpplague | 95 % of the cheap gadgets you see at the store are still 8051 based |
| 00:33.14 | ds2 | SiLabs 8051 isn't too bad |
| 00:33.21 | MonMotha | robbat2: I love AVR |
| 00:33.27 | MonMotha | I just don't like Atmel's supplied code |
| 00:33.31 | robbat2 | ? |
| 00:33.39 | MonMotha | that was supposed to be for RobotGuy, sorry |
| 00:33.42 | MonMotha | nick complete failed me |
| 00:33.58 | prpplague | robbat2 headache gone? |
| 00:34.01 | MonMotha | RobotGuy: my 1st preference for 8-bit is currently AVR, #2 is 8051 |
| 00:34.06 | ce_geek | So, tomorrow in class I've been tasked to lead a discussion about programming time sensitive things in an embedded environment with a real time OS. Things are different in that world as opposed to an 8-bit microprocessor. |
| 00:34.10 | robbat2 | prpplague, mostly |
| 00:34.20 | ce_geek | I've briefly asked about this before, just want to get a little better grip. |
| 00:34.31 | ds2 | don't for the MSP430's with the cheap tools |
| 00:34.31 | prpplague | prescribes two beers for robbat2 |
| 00:34.46 | prpplague | ce_geek time sensitive is relative |
| 00:34.47 | ce_geek | It seems that the following are useful: 1. Timers and interrupts 2. libraries that help with this stuff |
| 00:34.52 | ds2 | s/for/forget/ |
| 00:34.54 | MonMotha | RobotGuy: it's just that Atmel's code libraries are pretty terrible. Lots of wrappers so thin as to be pointless, code that is totally unreadable until you run it through "indent", etc. |
| 00:34.54 | RobotGuy | MonMotha: I want to get into using AVRs and AVR32's. Right now I use PICs and dsPICs. |
| 00:35.04 | MonMotha | ah yes, also MPS430 |
| 00:35.06 | prpplague | ce_geek most employers call out the need for rtos when standard OS is fine |
| 00:35.09 | ce_geek | prpplague: What if you want to bit-bang something with microsecond accuracy? |
| 00:35.10 | MonMotha | I don't have much experience with them, though |
| 00:35.18 | ds2 | *heart* ez430 USB thing |
| 00:35.26 | prpplague | ce_geek indeed that is sensitive |
| 00:35.28 | MonMotha | AVR's dev tools are mostly free, at least |
| 00:35.34 | ce_geek | prpplague: Linux on hammer = rtos? |
| 00:35.54 | RobotGuy | MonMotha: I dislike coders who don't indent properly. They would never be able to get a Python program of any complexity to work. |
| 00:35.54 | MonMotha | the only things that cost money are dev boards (which, of course, you don't strictly need) and their JTAG pod |
| 00:35.55 | prpplague | ce_geek you can do rtos on the hammer, there are plenty of patches for linux as rtos |
| 00:36.10 | MonMotha | RobotGuy: I'm talking downright misleading indent |
| 00:36.19 | RobotGuy | MonMotha: That's even WORSE |
| 00:36.28 | ce_geek | Okay, but as shipped, linux is not behaving as a "Real Time Operating System" on the hammer? |
| 00:36.36 | MonMotha | the USB code from Atmel for the 90USB64/128 was unreadable until I ran it through "indent" |
| 00:36.36 | prpplague | ce_geek that is correct |
| 00:36.53 | MonMotha | of course, it also built to like 60kB |
| 00:37.06 | MonMotha | I made some small modifications and got the flash code size down to like 12k |
| 00:37.06 | ce_geek | prpplague: Okay. I've been throwing around "RTOS" quite erroneously in class then. |
| 00:37.12 | MonMotha | with plenty of optimizatio still possible |
| 00:37.32 | MonMotha | I did that mostly by removing copy/pasted procedures and all the macro magic |
| 00:37.52 | ce_geek | prpplague: So if I needed microsecond accuracy for something happening on an I/O pin on the hammer, I would want to do whatever it took to put Linux in some sort of RTOS mode? |
| 00:38.08 | prpplague | ce_geek http://elinux.org/Realtime_Testing_Best_Practices |
| 00:38.11 | MonMotha | I should also note that the AT90USBKey (the basic dev/sample board for the 90USB128) is like $25 |
| 00:38.19 | RobotGuy | Crap, my chair has stopped working again - no more going out by myself for awile. :( |
| 00:38.50 | RobotGuy | MonMotha: I missed going to the local AVR32 seminar/class because I didn't have the $100 for it. |
| 00:39.52 | ce_geek | prpplague: Awesome, that's really helpful. |
| 00:39.53 | RobotGuy | I could have gotten a $300.00 programming/debug setup and a dev board. |
| 00:39.54 | MonMotha | AVR32 looks interesting |
| 00:40.00 | ds2 | AT90USBKey is a AVR32? |
| 00:40.08 | RobotGuy | ds2: No |
| 00:40.11 | MonMotha | no, it's 8-bit AVR |
| 00:40.17 | ds2 | ah |
| 00:40.23 | ds2 | but it has hardware USB support I take it? |
| 00:40.24 | MonMotha | the 90USB128 was their original USB device |
| 00:40.27 | MonMotha | well, in the 90 series |
| 00:40.33 | MonMotha | yes, it does |
| 00:40.43 | MonMotha | full/low speed, OTG supported |
| 00:40.50 | MonMotha | I've only ever bothered with the device side |
| 00:40.50 | ds2 | for 8 bit devices, I prefer them to be in 8pin or smaller packages |
| 00:41.01 | MonMotha | with USB support? there's not many otions |
| 00:41.13 | ds2 | is this one of those that has a host mode but is limited to what Atmel allows? |
| 00:41.27 | ds2 | there is software/bit bang USB on the AVRs |
| 00:41.28 | RobotGuy | Olimex has a 9260 and a 9261 board is supposed to come out next month, ARM9. |
| 00:41.37 | MonMotha | host mode looks pretty complete |
| 00:41.49 | MonMotha | several endpoints, and the host stack is mostly software so you can do what you want |
| 00:42.07 | ce_geek | prpplague: I wish I'd learned this stuff as an undergrad. Hopefully that will change as a result of this project |
| 00:42.17 | MonMotha | ds2: yes, low speed only, IIRC |
| 00:42.22 | MonMotha | and it's bit banged, so lotsa CPU usage |
| 00:42.25 | ce_geek | Does anyone here have any experience with the Micro-C OS? |
| 00:42.38 | ds2 | MonMotha: yeah but the 8 pin package is just so neat |
| 00:42.46 | ds2 | ATTiny13 |
| 00:42.47 | MonMotha | ds2: true |
| 00:42.55 | MonMotha | I didn't know it fit on a tiny13 |
| 00:43.11 | MonMotha | they do have some 8k 8-pin devices with A/D. Those look fun |
| 00:43.30 | MonMotha | I'm still waiting for an AVR with high-speed USB support |
| 00:43.44 | MonMotha | I wouldn't be surprised to see that on XMega since they are so DMA happy |
| 00:43.52 | ds2 | what would you use for high speed on such a small proc? |
| 00:44.14 | MonMotha | ds2: basically DMA'ing out to another device that doesn't speak USB and performing some board-level housekeeping |
| 00:44.21 | MonMotha | think video capture pod |
| 00:44.33 | ds2 | ah so like a faster FTDI 232 type application? |
| 00:44.38 | MonMotha | pretty much |
| 00:45.19 | MonMotha | I was going to take a video A/D and DMA onto USB |
| 00:45.30 | prpplague | i've been itching to use that libusb for avr |
| 00:45.39 | MonMotha | prpplague: I've done it. It's easy |
| 00:45.46 | SpeedEvil | ce_geek: On timing. It may be possible to use the PWM timers for that. |
| 00:45.48 | MonMotha | heck, I even did it in Windows and made it talk to a .NET app |
| 00:46.08 | SpeedEvil | ce_geek: many PWM units support high resolution timing of edges |
| 00:46.20 | MonMotha | ds2: 640x480x30FPSx16bpp is 450Mbps. Probably a bit much, but 512x480 might be doable |
| 00:46.35 | MonMotha | wiht proper fifos, you could buffer the blanked sections out |
| 00:46.37 | prpplague | ce_geek http://elinux.org/High_Resolution_Timers |
| 00:46.48 | prpplague | ce_geek http://elinux.org/High_Resolution_Timers |
| 00:46.52 | prpplague | oops duplicate |
| 00:47.07 | prpplague | ce_geek http://elinux.org/Real_Time |
| 00:47.12 | ds2 | I can settle for 320x240 |
| 00:47.16 | MonMotha | heh, Atmel's ARMs have slightly broken timers when used for edge timing measurement, FYI. |
| 00:47.27 | MonMotha | IIRC, you can do it on one input, but not the other |
| 00:47.49 | MonMotha | ds2: well, you'd need FIFOs for that (since you're dropping every other line), but that would work |
| 00:48.05 | SpeedEvil | mon" yeah - I havne't looked at that functionalitgy of the sc2410's PWM |
| 00:48.14 | MonMotha | it'll probably work on most of them |
| 00:48.18 | MonMotha | it works on AVR |
| 00:48.32 | SpeedEvil | MonMotha: or DMA |
| 00:48.33 | ds2 | MonMontha: but you are assuming a hardware demod ahead of it... if it can do direct NTSC capture.... |
| 00:48.51 | MonMotha | ds2: I was going to use an Analog Devices chip for that |
| 00:48.57 | ds2 | a quick app is to do frame captures of a auto backup cam, those are mostly NTSC it seems |
| 00:49.08 | ds2 | ah a dive into analog voodoo |
| 00:49.08 | MonMotha | also was going to chuck a TI ADC on there for grabbing VGA 480p |
| 00:49.15 | MonMotha | I had some parts picked out at one point |
| 00:49.25 | MonMotha | issue was effective bitrate of high-speed USB |
| 00:49.27 | SpeedEvil | ds2: you can cheat - and glue a usb1.1 webcam there. |
| 00:49.42 | ds2 | SpeedEvil: find one that is weatherproof for a decent price! |
| 00:50.08 | SpeedEvil | ds2: I solved that problem the silly way. |
| 00:50.16 | ds2 | so many fun things, so little time :( |
| 00:50.25 | MonMotha | you can do some cheating and convert to YUV, downres, and to 4:2:2 |
| 00:50.41 | SpeedEvil | ds2: camera inside car pointing out rear window |
| 00:50.46 | MonMotha | downres meaning discarding low bits of the sample, not decreasing frame size |
| 00:51.04 | ds2 | SpeedEvil: won't work on a pickup truck |
| 00:51.27 | SpeedEvil | ds2: no. I've done the 'webcam in a sealed box' before though. |
| 00:51.40 | ds2 | is there a chip that will do the YUV/4:2:2 conversion? |
| 00:51.55 | SpeedEvil | three solderable feedthroughs, inside a little metal box, and a soldered on glass. |
| 00:52.23 | MonMotha | ds2: built in to many video-oriented ADCs |
| 00:52.41 | ds2 | that sounds nice if only high speed analog wasn't so scary |
| 00:52.56 | MonMotha | heh, 480i/p isn't too bda |
| 00:52.58 | MonMotha | *bad |
| 00:53.02 | MonMotha | when you hit 1080p, that's scary |
| 00:53.47 | MonMotha | 240MBps at 16bits/pixel |
| 00:54.18 | MonMotha | 120MHz fundamental (edges will easily top 1GHz) |
| 00:54.49 | MonMotha | (for the analog signal) |
| 00:54.58 | SpeedEvil | On the USB side - you can get USB2 2MP cameras not too expensively. |
| 00:55.18 | SpeedEvil | the Creative optia AF - about $80/40 quid. 2MP 10FPS autofocus. |
| 00:55.35 | ds2 | it is the analog size that's trouble some. on the digital size things are more reasonable since noise is a little bit less of an issue |
| 00:56.12 | ds2 | SpeedEvil: I work in the $5, USB 1.1 webcam price range :) |
| 00:56.27 | MonMotha | ds2: well, completely different challenges |
| 00:56.51 | SpeedEvil | ds2: That's also nice in that it's a USB video class device, so works with linux. |
| 01:02.02 | ds2 | anyone played with the current MSP430 dongle thing with the CC2500? |
| 01:05.56 | *** join/#edev prpplague (n=dave123_@ppp-70-244-166-139.dsl.rcsntx.swbell.net) |
| 01:06.04 | prpplague | wheeee fun with lightning |
| 01:06.27 | prpplague | http://www.weather.com/weather/map/76053?from=36hour_map&mapdest=/maps/local/local/us_close_dfw_ultra_dfw_metro_dfw/1b/dopplerradar300mile_large.html |
| 01:07.28 | SpeedEvil | You? |
| 01:08.32 | prpplague | wow good one |
| 01:08.44 | SpeedEvil | (living in the area that is, not being the personification of the lightning god) |
| 01:09.10 | prpplague | lighting is really good this evening |
| 01:09.25 | SpeedEvil | ponders. |
| 01:09.34 | SpeedEvil | Don't think I've seen any this year. |
| 01:10.07 | prpplague | lighting? |
| 01:10.10 | SpeedEvil | yes |
| 01:10.32 | SpeedEvil | typically it's a summer/autumn thing here. |
| 01:10.39 | SpeedEvil | (Fife, Scotland) |
| 01:10.41 | prpplague | ah |
| 01:11.05 | SpeedEvil | sighs. |
| 01:11.25 | SpeedEvil | needs to do some sort of personal crap inventory system. |
| 01:11.27 | prpplague | nothing like sirens to scare the kids |
| 01:11.43 | SpeedEvil | Especially for electronics stuff. |
| 01:11.53 | MonMotha | SpeedEvil: BOMerMan! :) |
| 01:12.00 | prpplague | refrains from making joke about "crap inventor system" |
| 01:12.05 | ce_geek | Getting the paper off this really old plexi glass was like pulling teeth |
| 01:12.12 | ce_geek | But I've got the whole thing mounted nicely now: |
| 01:12.21 | ce_geek | http://www.eece.maine.edu/~rblanche/hammer_pix/ |
| 01:12.26 | solar | robbat2: if that headache goes away and you can get me into an ash login tonite that would be cool. |
| 01:12.57 | SpeedEvil | What I'd really like is for e-ink not to be fuckwits. |
| 01:13.05 | SpeedEvil | I want e-ink labels. |
| 01:13.08 | prpplague | ce_geek jeeze |
| 01:13.16 | ce_geek | prpplague: ya :-/ |
| 01:13.33 | SpeedEvil | A little handheld electrostatic field printer to write them, |
| 01:13.37 | ce_geek | prpplague: I think if I used some really nasty solvent it would have helped but that would have been hard to come by around here. |
| 01:13.48 | SpeedEvil | ce_geek: vegetable oil |
| 01:13.49 | prpplague | ce_geek oh cute setup |
| 01:13.56 | SpeedEvil | ce_geek: leave it on for a few hours |
| 01:13.57 | ce_geek | prpplague: thx |
| 01:14.13 | ce_geek | SpeedEvil: Will that help eat the adhesive too? |
| 01:14.17 | SpeedEvil | ce_geek: it penetrates through the label and adhesive, and makes the surface greasy |
| 01:14.31 | SpeedEvil | ce_geek: in most cases. It then peels right off. |
| 01:14.32 | ce_geek | SpeedEvil: If I do this again I'll remember that. Thanks. |
| 01:14.42 | SpeedEvil | Also - heat. |
| 01:14.53 | SpeedEvil | Many labels resistant to this peel off real easy if you apply 100C |
| 01:14.56 | ce_geek | SpeedEvil: Hopefully next time I'll have some stuff that hasn't been hiding in the storage room for about 15 years. The dust I wiped off of this was incredible. |
| 01:15.01 | SpeedEvil | :) |
| 01:15.14 | ce_geek | I couldn't find any stuff thick enough at the local stores so I went digging and found this |
| 01:15.37 | ce_geek | prpplague: yeah, I wanted to make something that would serve as an example of a setup they could use for the purpose of student labs. |
| 01:15.51 | ce_geek | prpplague: I have more ribbon cables to make so I can break out all the pins into the bread board |
| 01:15.52 | prpplague | looks good |
| 01:16.12 | ce_geek | prpplague: Almost going to want another bread board, going to eat up a good chunk of this one just bringing connections out with ribbon. |
| 01:16.50 | ce_geek | prpplague: for the purpose of students messing with stuff, I figured a bread board would be better than generations of people soldering things in the prototyping area. |
| 01:17.40 | prpplague | yea, we are looking at doing a combo breadboard carrier |
| 01:17.44 | SpeedEvil | ce_geek: sureelectronics on ebay have cheap solderless breadboards. $20 for 5 |
| 01:18.03 | SpeedEvil | ce_geek: seem to work OK - but I haven;'t really given them any thourough tests. |
| 01:18.27 | ce_geek | SpeedEvil: Cool. I grabbed this one from a pile in the storage room. They order like 50 of them at a time, I think the price they pay at circuit specialists with volume discount might be similar, not sure. |
| 01:19.04 | ce_geek | I remember when they switched to circuit specialists for components too... Oh man. The leads on the components were all like 30-AWG, it was bad. |
| 01:19.08 | SpeedEvil | Possibly. |
| 01:19.24 | ce_geek | Couldn't shove them in the sockets in the bread board almost. Too tiny |
| 01:19.55 | g1powermac_PB | alright, cracking open a zipit ain't easy |
| 01:20.05 | g1powermac_PB | but I did get it open without breaking anything, I hope |
| 01:20.08 | prpplague | thats what a dremel is for |
| 01:20.14 | SpeedEvil | Hammer! |
| 01:20.17 | g1powermac_PB | GPSFan: I got old model with the wires |
| 01:20.23 | SpeedEvil | However. |
| 01:20.32 | SpeedEvil | Being more serious, I like hot-glue |
| 01:20.34 | g1powermac_PB | SpeedEvil: heheh, that only works when you don't care and/or just pissed off ;-) |
| 01:20.50 | SpeedEvil | You hot-glue one side of the object onto a bit of metal, the other side onto another bit |
| 01:21.06 | SpeedEvil | then apply a lever |
| 01:21.36 | ce_geek | Clever |
| 01:21.40 | SpeedEvil | I'm buying some more of http://cgi.ebay.co.uk/MC34063-Based-Switching-Regulator-Adapter-Step-Down_W0QQitemZ230242851864QQihZ013QQcategoryZ58288QQssPageNameZWDVWQQrdZ1QQcmdZViewItem - SMPS - they are handy |
| 01:22.00 | g1powermac_PB | sigh, camera sucks |
| 01:22.10 | g1powermac_PB | this thing can't do macro images too well |
| 01:22.41 | ds2 | g1powermac_PB: you installing T0mW's board? |
| 01:22.44 | SpeedEvil | ce_geek: teh nice thing is that this doesn't mar the surface at all - heat it up, wipe it off, take tissue moistened by paint thinner, and clean it off. |
| 01:23.00 | g1powermac_PB | ds2: no, cracking open the zipit 2 |
| 01:23.08 | ds2 | ah newer stuff |
| 01:23.08 | SpeedEvil | ce_geek: this - http://cgi.ebay.co.uk/760-Point-Solderless-Breadboard-Include-Jumpwires-5pcs_W0QQitemZ230243830543QQihZ013QQcategoryZ50423QQssPageNameZWDVWQQrdZ1QQcmdZViewItem |
| 01:23.26 | ce_geek | SpeedEvil: sweet |
| 01:23.39 | g1powermac_PB | err |
| 01:23.41 | g1powermac_PB | stupid camera |
| 01:25.18 | SpeedEvil | found a source of .5mm gold wire. |
| 01:25.35 | SpeedEvil | I have a drawplate, so I'm half tempted to just buy some and try it. |
| 01:27.43 | *** join/#edev MothaWork (n=monmotha@c-69-245-153-193.hsd1.in.comcast.net) |
| 01:28.36 | ce_geek | drawplate? |
| 01:29.17 | prpplague | opps time to power down |
| 01:31.28 | SpeedEvil | oops - mischan |
| 01:31.52 | SpeedEvil | I'd been going on to someone about planning on a cheap wirebondy thing |
| 01:32.04 | SpeedEvil | for electronics prototyping |
| 01:32.16 | ce_geek | Woah we have a wire bonding machine here. |
| 01:32.25 | SpeedEvil | a drawplate is a thing with a hole in it, through which you draw wire through. |
| 01:32.25 | ce_geek | You have to get like permission from congress to use it. |
| 01:32.37 | SpeedEvil | it makes the wire the diameter of the hole |
| 01:32.38 | SpeedEvil | yeah |
| 01:32.45 | SpeedEvil | I'm aiming _much_ lower rent. |
| 01:33.10 | ce_geek | Vibrates, rght? |
| 01:33.33 | SpeedEvil | You apply pressure, then ultrasound - is one way |
| 01:33.43 | g1powermac_PB | well, no high res images of the mobo for prpplague |
| 01:33.47 | g1powermac_PB | need a better camera |
| 01:34.09 | SpeedEvil | g1powermac_PB: you're not too close are you? |
| 01:34.29 | SpeedEvil | Most cameras have 10-20cm minimum distance in macro mode |
| 01:34.38 | SpeedEvil | Also - make sure you've got plenty of light |
| 01:34.53 | g1powermac_PB | its a cheapo digicam from quite a few years ago |
| 01:35.06 | SpeedEvil | scanner can work too |
| 01:35.07 | g1powermac_PB | its only 3.2 megapixels |
| 01:36.52 | g1powermac_PB | if I had an dslr, these pics would've been awesome |
| 01:39.32 | g1powermac_PB | well, I'll upload these images anyway, even though they aren't great |
| 01:41.41 | g1powermac_PB | at least I know how to crack this sucker open, shouldn't be too hard next time |
| 01:45.27 | g1powermac_PB | ponders if he should blot out the MAC address and serial number on this thing. . . |
| 01:58.34 | RobotGuy | Wow, there are 33 GB SD cards now. |
| 01:58.38 | RobotGuy | 32 GB |
| 01:58.55 | SpeedEvil | 8G microSD too |
| 01:59.01 | g1powermac_PB | dang |
| 01:59.04 | RobotGuy | Yep. |
| 01:59.10 | g1powermac_PB | what about minisd? |
| 01:59.25 | RobotGuy | g1powermac_PB: I don't know about miniSD. |
| 01:59.48 | g1powermac_PB | k, prolly 16gb since its in between full sd and microsd |
| 02:00.55 | RobotGuy | I see some 2 GB miniSD |
| 02:04.03 | g1powermac_PB | bah, flickr is annoying me |
| 02:04.20 | SpeedEvil | Increase you refresh rate. |
| 02:04.32 | g1powermac_PB | no, no, flickr.com |
| 02:04.38 | g1powermac_PB | :-) |
| 02:04.49 | SpeedEvil | Why? |
| 02:05.04 | g1powermac_PB | keeps uploading this one pic out of order of the rest |
| 02:05.57 | RobotGuy | 2 GB microSD |
| 02:07.36 | g1powermac_PB | meh, its out of order, but screw it: http://www.flickr.com/photos/77471288@N00/sets/72157604702887753/ |
| 02:12.07 | solar | anybody ran across MPPE_MAX_KEY_LEN not being defined while cross compiling but is while native? |
| 02:34.47 | g1powermac_PB | hmm, had to turn off all the security to get my zipit to connect to my wifi here |
| 02:35.25 | g1powermac_PB | meh, its prolly something funky with the wep I use, really should get a router here that supports wpa |
| 02:59.47 | g1powermac_PB | gnight all |
| 03:58.23 | *** join/#edev linac (n=lin@222.76.82.137) |
| 04:01.51 | *** join/#edev DrZimmerman (n=theo@77-56-246-76.dclient.hispeed.ch) |
| 04:02.42 | robbat2 | ok, not sure if my brain is just slow today, but the stock env on the Nail I just brought up didn't have g_ether like the docs suggest |
| 04:03.15 | robbat2 | Linux Hammer 2.6.22 #5 Thu Nov 1 17:16:36 CDT 2007 armv4tl unknown |
| 04:05.39 | *** join/#edev steliosk (n=Stelios@athedsl-309648.home.otenet.gr) |
| 04:19.20 | Dr{Who} | was it compiled as a module? |
| 05:26.45 | *** join/#edev magnetron (n=magnetro@unaffiliated/magnetron) |
| 06:08.03 | *** join/#edev drath_ (i=vmaster@p5B07CA8F.dip.t-dialin.net) |
| 06:21.43 | *** join/#edev emil|k (n=emil@p548CFEFB.dip.t-dialin.net) |
| 07:03.36 | robbat2|na | Dr{Who}, no modules in that env that was already on the flash. I guess i'll compile a better env |
| 07:23.16 | *** join/#edev [AD]Turbo (n=turbo@host24-241-dynamic.22-79-r.retail.telecomitalia.it) |
| 07:24.12 | [AD]Turbo | hi all |
| 07:59.15 | *** join/#edev emil|k (n=emil@p548CFEFB.dip.t-dialin.net) |
| 08:02.49 | *** join/#edev hw_ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 08:14.02 | *** join/#edev harpal (n=harpal@121.246.75.165) |
| 08:42.27 | *** join/#edev lyakh (n=lyakh@p57BD232C.dip0.t-ipconnect.de) |
| 09:25.38 | *** join/#edev lyakh (n=lyakh@p57BD232C.dip0.t-ipconnect.de) |
| 10:13.18 | Redhatter | Hi... silly question... I've picked up an AT91SAM7X-EK development board, with the SAM-ICE USB JTAG device. I understand these devices aren't useful under Linux, but you can build a (short!) passive parallel port cable that does the same job. |
| 10:13.31 | Redhatter | Anyone know of a good schematic that'd be suitable for the AT91SAM7X? Are the pinouts the same? |
| 10:14.42 | Redhatter | I can code under Windows 2000 using gcc or IAR... but doing everything in Linux has a certain appeal. :-) |
| 10:15.04 | SpeedEvil | what does the JTAG device show up as on lsusb? |
| 10:15.10 | SpeedEvil | Or have you investigated that |
| 10:15.16 | Redhatter | SpeedEvil, One moment... haven't tried looking. |
| 10:17.03 | Redhatter | Bus 006 Device 003: ID 1366:0101 |
| 10:19.06 | Redhatter | Not very helpful... for a moment there I thought it wasn't showing up at all. |
| 10:25.50 | SpeedEvil | Can you see the chips it comprises? http://www.ixo.de/info/usb_jtag/ |
| 10:25.59 | SpeedEvil | (not directly related) |
| 10:33.27 | Redhatter | Not without taking it apart. Considering it's university property... I'm not keen to do this. |
| 10:33.38 | SpeedEvil | k |
| 10:34.00 | SpeedEvil | Sorry - I don't know direclty. |
| 10:34.21 | SpeedEvil | However - there are numerous JTAG - parallel port linux thingies on the net. |
| 10:52.03 | Redhatter | That's what I figured. However I'm not sure how compatible the pinouts these things use, are with the Atmel board I have here. |
| 10:54.35 | SpeedEvil | Does it have a standard 10 pin header? |
| 11:21.28 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 11:36.48 | *** join/#edev chouimat (n=dieu@r2351064.cidc.net) |
| 11:37.54 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 11:45.08 | *** join/#edev ali_as_ (n=as@ambix.plus.com) |
| 11:56.33 | *** join/#edev GPSFan (n=kenm@12.10.255.246) |
| 11:56.33 | *** mode/#edev [+o GPSFan] by ChanServ |
| 12:01.33 | *** join/#edev sjhill (n=sjhill@real.realitydiluted.com) |
| 12:01.33 | *** mode/#edev [+o sjhill] by ChanServ |
| 12:04.04 | *** join/#edev emil|k (i=emil@f054012177.adsl.alicedsl.de) |
| 12:10.11 | *** join/#edev slapin_nb (n=slapin@saris-243.ip.PeterStar.net) |
| 12:25.30 | *** join/#edev Appleman1234 (n=Appleman@CPE-121-222-170-32.qld.bigpond.net.au) |
| 13:11.01 | *** join/#edev chouimat|work (n=dieu@209.217.106.98) |
| 13:20.26 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 13:44.35 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 13:45.43 | *** join/#edev g1powermac (n=g1powerm@unaffiliated/g1powermac) |
| 14:12.18 | prpplague | wmat: ping |
| 14:21.04 | chouimat|work | morning |
| 14:21.19 | g1powermac | heya prpplague and chouimat|work |
| 14:22.26 | g1powermac | prpplague, I took some photos of my cracked open zipit, though they didn't turn out great due to bad digicam |
| 14:22.27 | prpplague | greetings |
| 14:22.36 | g1powermac | http://www.flickr.com/photos/77471288@N00/sets/72157604702887753/ |
| 14:22.39 | prpplague | g1powermac: ok |
| 14:22.41 | prpplague | looks |
| 14:24.03 | prpplague | g1powermac: yea too blurry to tell much |
| 14:24.16 | g1powermac | yea, unfortunately |
| 14:24.30 | g1powermac | dang digicam sucks at macro shots |
| 14:24.35 | g1powermac | even when using a tripod |
| 14:27.41 | g1powermac | these shots are pretty good though: http://www.drtune.com/zipit2/ |
| 14:32.42 | g1powermac | GPSFan, I'm going to try to wire up a usb cable for the zipit tonight and see what happens when I plug it in. . . |
| 14:33.13 | g1powermac | wonder if they programmed any functionality into the usb port with the default software |
| 14:34.09 | GPSFan | g1powermac: remember that it is a device interface, you may need to get the usb gaget driver module compiled and running. |
| 14:35.22 | g1powermac | well, I'm just going to see if the default software thats on it will respond to being plugged into a usb port on a computer |
| 14:35.27 | *** join/#edev CosmicPenguin (n=nobody@163.181.251.103) |
| 14:35.27 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 14:35.49 | g1powermac | since I don't have a minisd to sd adapter yet, I can't do any custom software on my zipit |
| 14:37.29 | GPSFan | ah, thats right, it took me a while to get some miniSD adapters. |
| 14:43.32 | SpeedEvil | passes g1powermac a soldering iron. |
| 14:44.54 | g1powermac | hehe, can't destroy the minisd card, it still has to work in a minisd slot :-) |
| 14:45.11 | GPSFan | g1powermac: looking at the z2 kernel config, it doesn't look like much usb is built into the kernel at all. |
| 14:45.22 | g1powermac | ahh |
| 14:45.24 | g1powermac | hmm |
| 14:45.47 | g1powermac | speaking of the kernel, what gcc/libc was used to compile it? |
| 14:46.53 | GPSFan | from the z2's dmesg: Linux version 2.6.21.1 (jkaisner@savior.csd.local) (gcc version 3.4.1 20040714 (TimeSys 3.4.1-7)) #162 Wed Nov 28 15:21:59 EST 2007 |
| 14:47.44 | g1powermac | k, so I guess I got to get that compiler version working in my buildroot to make kernel modules |
| 14:48.19 | GPSFan | not really, I used my buildroot's compiler to build the fb modules. |
| 14:48.33 | g1powermac | really? I thought you needed to use the same compiler |
| 14:50.16 | g1powermac | is thinking of bring unionfs/aufs to the zipit |
| 14:50.20 | GPSFan | I actually used the one from the Hammer buildroot. IIRC it is recommended that you should, but often one uses what one has, just to see if it works. |
| 14:50.20 | g1powermac | *bringing |
| 14:50.53 | g1powermac | it would really help overlaying stuff ontop of the default system at boot time |
| 14:51.13 | g1powermac | I've worked with unionfs back when I made my clusterix linux livecd |
| 14:51.43 | g1powermac | and was one of the first guys who brought it to linux livecds, which eventually led it to be used in knoppix |
| 14:52.09 | GPSFan | g1powermac: then you are the guy to do it... |
| 14:52.29 | g1powermac | it was quite fun back then, it was utterly buggy ;-) |
| 14:54.22 | g1powermac | but yea, the neat part with unionfs is that you can combine directories, say like if you wanted to use some libraries you made that you loaded on the sd card and you wanted to combine them with the libraries already on a read only device/cd |
| 14:54.37 | CosmicPenguin | the idea of unionfs is great |
| 14:54.39 | CosmicPenguin | the implementation...... |
| 14:55.05 | g1powermac | unionfs will combine the lib directories of both and then make it writable to the writable part of the union, which would be on the sd card |
| 14:55.28 | *** join/#edev lyakh (n=lyakh@p57BD232C.dip0.t-ipconnect.de) |
| 14:55.59 | g1powermac | normally the writing part takes place in ram on linux livecds, but in this case it can go to the sd card due to obvious ram constraints and the writable nature of sd cards |
| 14:56.06 | *** join/#edev MothaWork (n=monmotha@mail.forwardengineering.com) |
| 14:56.21 | GPSFan | when one has a combined dir, say /lib if you do an ls -la is there any indication where each file is really located? would it be a "good thing(tm)" if there was? or not? |
| 14:56.56 | g1powermac | theoretically, you'd even be able to save your entire session and/or any changes in this type system |
| 14:57.07 | g1powermac | to ls, I don't think there is any difference |
| 14:57.23 | g1powermac | however, its been awhile since I've worked with it, so things might have changed |
| 14:58.06 | GPSFan | to ls, probably, but maybe to a special form of ls. |
| 14:58.19 | CosmicPenguin | you would need to get down inside the layering |
| 14:58.22 | g1powermac | yea, I think unionfs has special tools to tell |
| 14:58.33 | CosmicPenguin | which would have to be reported through something other then the vfs |
| 14:58.37 | g1powermac | and as CosmicPenguin said, you could look inside the layering |
| 14:58.41 | CosmicPenguin | <PROTECTED> |
| 14:58.45 | GPSFan | CosmicPenguin: indeed, I wasn't really thinking that ls could do it. |
| 14:59.10 | CosmicPenguin | Thats why its a great idea, because you get this magic laying for essentially free |
| 14:59.20 | g1powermac | yup |
| 14:59.48 | GPSFan | how does unionfs handle identical (and identically named) files on the 2(or more) parts? |
| 14:59.52 | CosmicPenguin | I liked it because you could provide default configuration files for users, and then let them change them if they needed to |
| 14:59.53 | g1powermac | its wonderful for livecds and I think could be really beneficial in embedded systems with expansion cards |
| 15:00.07 | g1powermac | GPSFan, there's settings for that |
| 15:00.08 | CosmicPenguin | as opposed to autogenerating scripts or copying them about or whatever |
| 15:00.31 | GPSFan | doesn't know squat about unionfs |
| 15:00.49 | *** join/#edev ico2 (n=ico2@modem-1663.cougar.dialup.pol.co.uk) |
| 15:01.55 | g1powermac | I'm currently looking at aufs which is supposedly more stable than unionfs atm: http://aufs.sourceforge.net/ |
| 15:04.40 | g1powermac | ponders how he would build aufs within buildroot |
| 15:05.01 | g1powermac | specifically using the needed kernel version |
| 15:07.16 | GPSFan | g1powermac: don't build the kernel inside buildroot, build it outside, just point the Makefile at the buildroot toolchain. that's how I did it |
| 15:07.29 | g1powermac | ahh, k |
| 15:09.39 | g1powermac | realistically, I don't think I even need to build the kernel, I just need the kernel headers and some source to build aufs. . .I think |
| 15:12.26 | *** join/#edev landley (n=landley@adsl-68-93-84-18.dsl.hstntx.swbell.net) |
| 15:12.26 | *** mode/#edev [+o landley] by ChanServ |
| 15:12.27 | prpplague | looks like submitting arm patches has gotten harder :( |
| 15:12.32 | prpplague | landley: greetings |
| 15:13.21 | landley | prpplague: hiya. |
| 15:13.23 | landley | prpplague: arm patches got harder? |
| 15:14.40 | prpplague | http://lists.arm.linux.org.uk/lurker/message/20080424.101338.4384cbc2.en.html |
| 15:16.03 | landley | prpplague: except that's how it worked this time around anyway. |
| 15:16.12 | landley | Your patch _did_ get into his git tree for -rc1, didn't it? |
| 15:16.18 | prpplague | yea |
| 15:16.36 | landley | So he's saying he's going to make what he's been doing official policy. |
| 15:19.24 | prpplague | yea, just means that you have to babysit your patches to make sure they actually get applied |
| 15:19.42 | prpplague | landley: give notice yet? |
| 15:21.22 | chouimat|work | prpplague: interesting :) |
| 15:21.39 | CosmicPenguin | prpplague: what RMK is doing is standard kernel policy across the board |
| 15:24.53 | prpplague | CosmicPenguin: i agree, again my problem is the statement "Anything _not_ in my git tree, even if it's in the patch system or has |
| 15:24.53 | prpplague | been around on the mailing list, even if it's been around for months, |
| 15:24.53 | prpplague | will NOT be merged during that merge window. |
| 15:24.53 | prpplague | " |
| 15:25.17 | CosmicPenguin | thats what everybody does |
| 15:25.22 | CosmicPenguin | lm-sensors - same thing |
| 15:25.30 | CosmicPenguin | watchdog, same thing |
| 15:25.34 | CosmicPenguin | crypto same thing |
| 15:25.52 | prpplague | whats the point of posting it on the mailing or even on the patch system then? i don't see the point |
| 15:26.39 | CosmicPenguin | so that it can get code reviewed and the git maintainer can add it to the tree |
| 15:27.37 | prpplague | ok i understand code review on the list , but if you've posted it to the patch system and it sits there |
| 15:27.55 | g1powermac | GPSFan, is the z2kernel in the svn repo already patched up for the z2? |
| 15:28.30 | prpplague | CosmicPenguin: the hammer stuff had been signed off on for literally a month before i started sending email to RMK asking why it hadn't been applied |
| 15:29.00 | CosmicPenguin | So then whats your solution? |
| 15:29.13 | CosmicPenguin | You are bound and determined that the kernel process is broken, so I'm sure you have your own thoughts |
| 15:30.43 | prpplague | CosmicPenguin: 1) posted schedules 2) a deadline for maintainers to vote yea/nea on the patch 3) clear process to go over the head of the maintainer |
| 15:31.04 | GPSFan | g1powermac: don't know, I started with a vanilla kernel from kernel.org and added zipits patches. |
| 15:31.11 | g1powermac | k |
| 15:31.23 | CosmicPenguin | You're never going to get a posted schedule |
| 15:31.27 | CosmicPenguin | ever |
| 15:31.42 | prpplague | CosmicPenguin: yet they expect for you to "know" when a merge window is |
| 15:31.54 | CosmicPenguin | its a big boy operating system |
| 15:32.48 | prpplague | CosmicPenguin: even if it wasn't a posted schedule, a simple banner on the patch system or post on the mailing list "merge window has started" |
| 15:33.01 | prpplague | CosmicPenguin: i don't buy that argument |
| 15:33.21 | prpplague | i consider that elitist |
| 15:33.30 | CosmicPenguin | Yes - it is |
| 15:33.42 | CosmicPenguin | The merge window opens when the kernel is released |
| 15:33.51 | CosmicPenguin | for Linus |
| 15:34.14 | CosmicPenguin | the other merge window will open before that - nearly everybody can read the tea leaves and see when the -rc is maturing |
| 15:36.30 | CosmicPenguin | It is very difficult to be a kernel developer - I think thats why so many people burn out |
| 15:36.45 | prpplague | CosmicPenguin: is there a page or document that states that "the merge window opens wiehn the kernel is released" ? |
| 15:37.05 | CosmicPenguin | There has to be - I didn't make that up |
| 15:37.56 | ce_geek | Can anyone help me roughly identify circumstances where a RTOS in an embedded system is necessary? It seems like a system such as the hammer board can respond to things reliably within fractions of a second, but perhaps when microsecond accuracy is important that's when a RTOS is useful? |
| 15:38.46 | prpplague | ce_geek: you looked over the real time page on elinux? |
| 15:39.04 | prpplague | CosmicPenguin: if there is, i've yet to find it and i've looked very hard |
| 15:39.46 | prpplague | CosmicPenguin: even if the schedule is thrown out, you still have items 2 and 3 |
| 15:40.14 | ce_geek | prpplague: Looking through that stuff just now actually :-/. I'll just read more. Rick was just down and we were chatting about things in a real general sense, that's why I asked. |
| 15:40.26 | CosmicPenguin | I have no option on 2 and 3 because I've never had that problem |
| 15:40.28 | prpplague | CosmicPenguin: if you are going to volunteer to be a maintainer of a subsystem you should have a responsibility to look at submited posts in a reasonable amount of time and either give a yea/nea decision |
| 15:40.33 | CosmicPenguin | indeed |
| 15:40.38 | CosmicPenguin | but you can't force it |
| 15:40.54 | CosmicPenguin | You should ask for a new maintainer |
| 15:40.55 | prpplague | ce_geek: ahh ok |
| 15:41.07 | *** join/#edev ico2_ (n=ico2@modem-1170.cheetah.dialup.pol.co.uk) |
| 15:41.08 | prpplague | yea, not l ikely to happen |
| 15:41.19 | ce_geek | prpplague: I see there's some tables and stuff in here with test results, these will probably answer my question |
| 15:41.34 | landley | prpplague: to understand the kernel development process, you have to start with the assumption the maintainers are swamped. |
| 15:41.38 | ce_geek | prpplague: Yep, interrupt response times and such. |
| 15:41.41 | landley | The higher up the tree you go, the more swamped they are. |
| 15:41.51 | landley | So they push work down as far as they can. |
| 15:42.05 | ce_geek | prpplague: Looks like this stuff can get really, really deep. |
| 15:42.25 | landley | The merge window opens when the new .0 is released, and runs until the -rc1 is released. |
| 15:42.52 | landley | They try to do this 4 times a year, which is every 3 months, but it's got +- a month in there with each cycle. |
| 15:43.17 | landley | The theory is that you queue up patches during the entire 3 months, and they go into the next merge window. |
| 15:43.22 | prpplague | landley: i understand that completely, what i want is a far shake |
| 15:43.34 | landley | In practice, everybody tries to submit stuff _during_ the merge window, leading to the maintainers losing stuff in the noise. |
| 15:43.51 | prpplague | landley: when patches submitted 3 or 4 weeks after mine get applied in hours and mine is still sitting there |
| 15:43.58 | prpplague | landley: thats what i have a problem with |
| 15:44.10 | landley | Agreed. Point it out. |
| 15:44.40 | landley | The current kernel development process is a 4 level thing. |
| 15:44.42 | prpplague | ce_geek: indeed RTOS has a ton of ramifications |
| 15:44.54 | landley | Developer, maintainer, lieutenant, Linus. |
| 15:45.14 | ce_geek | prpplague: Yep. I think it'll be a good thing to talk about at least on a high level for our class. |
| 15:45.27 | landley | (I happen to know the details because of that patch penguin thing in 2002.) |
| 15:45.37 | landley | At each level, you're owed a response. |
| 15:45.53 | landley | Maintainer owes developer a response, lieutenant owes maintainer a response, and linus owes his lieutenants a response. |
| 15:46.16 | landley | The response may be "no", but it's still a response. |
| 15:46.25 | CosmicPenguin | yes - and if any part of the chain is broken, then that link of the chain should be replaced |
| 15:46.38 | prpplague | landley: yea, which is my point that the maintainer should have a responsibility to review a patch in a timely manner and if they fail to that, a clear procedure needs to be in place to contact the lieutenant for proper handling |
| 15:46.59 | landley | Taking the 3 month release schedule as a frame of reference: the best you can do is to submit right after the last merge window, aiming at the next merge window. |
| 15:47.08 | landley | That gives them the maximum possible time to respond to the thing. |
| 15:47.10 | CosmicPenguin | prpplague: have you tried contacting the lieutanant? |
| 15:47.33 | landley | There are four layers, which is three escalations. |
| 15:47.36 | prpplague | CosmicPenguin: i would assume thats RMK in my case |
| 15:47.47 | landley | So if they haven't responded in 1/3 of the time (1 month), you're probably justified in going over their heads to the next level. |
| 15:47.54 | landley | They did not _reject_ your patch. |
| 15:48.05 | CosmicPenguin | so who is your maintainer? |
| 15:48.08 | landley | That said, during that time you should probably try at least 3 retransmits before giving it up as a dead host. |
| 15:48.10 | prpplague | landley: bjdooks |
| 15:48.25 | prpplague | CosmicPenguin: bjdooks |
| 15:48.29 | CosmicPenguin | and what is he doing then? |
| 15:48.29 | landley | So probably about once a week, poke 'em again, and after a month go to the next level and say "the maintainer is dead, reboot him". |
| 15:48.30 | SpeedEvil | ce_geek: as a most basic definition - when you require faster response than a non-realtime OS can do with your available hardware peripherals. For example, to bit-bang USB2 requires a few nanosecond timing, but if you've got a USB2 controller, 1/100th of a second can be just fine. |
| 15:48.32 | CosmicPenguin | does he maintain a git tree? |
| 15:48.35 | CosmicPenguin | does he push to rmk? |
| 15:49.07 | CosmicPenguin | in other words, rmk is clearly pushing to linus - who is pushing to rmk? |
| 15:49.11 | prpplague | CosmicPenguin: he currently does not maintain a git tree, rmk is the one that applies the patch to his tree, but only after djdooks signs off |
| 15:49.26 | CosmicPenguin | and has djdooks signed off? |
| 15:49.56 | prpplague | CosmicPenguin: in this case for the hammer patch only after i sent a nasty email to him |
| 15:50.14 | CosmicPenguin | okay, then you should be contacting rmk and telling him that djdooks is giving you a hastle |
| 15:50.15 | prpplague | CosmicPenguin: but its been the same with all my other patches |
| 15:50.56 | prpplague | CosmicPenguin: i have, he's made it pretty clear that bjdooks needs to sign off |
| 15:51.11 | CosmicPenguin | so did you ask him why bjdooks is ignoring you? |
| 15:51.32 | CosmicPenguin | he must have a reason - he can't just be doing it to be a dick |
| 15:51.32 | prpplague | CosmicPenguin: yea, i got the standard "kernel developers are busy" |
| 15:51.53 | prpplague | CosmicPenguin: but that doesn't seem to bar bjdooks from approving his company's patches |
| 15:52.07 | CosmicPenguin | I would point that out clearly to rmk then |
| 15:52.14 | CosmicPenguin | if you feel you are being unfairly ignored, then you should take action |
| 15:52.20 | CosmicPenguin | and if rmk doesn't listen, then go higher |
| 15:53.07 | CosmicPenguin | The process itself isn't perfect, but its not broken either |
| 15:53.17 | CosmicPenguin | but any process is only as good as the individuals executing the process |
| 15:53.26 | CosmicPenguin | if you have legitimate concerns, then you should voice them |
| 15:53.50 | CosmicPenguin | But really, none of your items stand a very good chance to be implemented |
| 15:54.08 | CosmicPenguin | so you can either be bitter or a cause for change |
| 15:54.37 | ce_geek | SpeedEvil: Thanks |
| 15:54.54 | prpplague | or just don't bother with submitting unless its really important |
| 15:55.38 | CosmicPenguin | Further proof that you need to make people aware that your maintainers aren't listening to you |
| 16:03.55 | *** mode/#edev [+o prpplague] by ChanServ |
| 16:16.53 | *** join/#edev landley_ (n=landley@adsl-68-93-84-18.dsl.hstntx.swbell.net) |
| 16:17.34 | landley_ | prpplague: when you resubmit patches, include a timestamp of each previous submission and a link to that submission in the archive. |
| 16:17.37 | landley_ | Always CC the arm mailing list. |
| 16:17.57 | landley_ | If the average resubmit (with one week gap between them) is >2 after a few months, point this out to RMK. |
| 16:18.05 | landley_ | The two words you need here are "paper trail". |
| 16:18.19 | prpplague | landley_: yea, i think i'm gonna "blog" the next time just to keep track of it |
| 16:18.22 | *** mode/#edev [+o landley] by ChanServ |
| 16:18.39 | landley | blog? |
| 16:21.13 | g1powermac | GPSFan, which patch did you use btw? |
| 16:23.46 | g1powermac | I'm using the original one and its failing to patch |
| 16:24.15 | g1powermac | complaining about not finding build_tree/linux/arch/arm/Kconfig |
| 16:25.28 | prpplague | landley: just post a daily report of whats going on to a wiki page or something |
| 16:25.41 | landley | Cool. |
| 16:25.47 | landley | Let me know where that goes up, if you do. |
| 16:26.12 | landley | In fact, having a page on the elinux wiki with the active submissions and another with the historical submissions might be really nice. :) |
| 16:26.30 | landley | But a "hammer blog" would be nice too. Do an rss feed. :) |
| 16:26.51 | g1powermac | GPSFan, meh, nm, I tried the slimer patch instead, it worked fine |
| 16:29.48 | prpplague | landley: hehe |
| 17:05.32 | *** join/#edev dijenerate_ (n=dijenera@65.48.148.198) |
| 17:09.18 | *** join/#edev Redhatte1 (n=vk4fsjl@203.94.56.252) |
| 17:10.47 | prpplague | landley: hehe interesting, looks like luminary did the exact same thing we did on the nail board for the jtag and serial port |
| 17:20.00 | prpplague | lordy lordy |
| 17:20.03 | prpplague | so much to do |
| 17:20.14 | SpeedEvil | passes prpplague a spare hour. |
| 17:20.34 | prpplague | landley: what are you going to be working on after your soon departure from the great city of Houston? |
| 17:20.39 | prpplague | SpeedEvil: thanks |
| 17:21.16 | wmat | prpplague: pong |
| 17:21.47 | prpplague | wmat: hey, not nagging, just curious where Hammer stands as far as codesourcery support |
| 17:21.58 | prpplague | wmat: is it something that is still in the works? |
| 17:22.11 | prpplague | wmat: oh and while i got you, is there a way to delete pages on the wiki? |
| 17:22.25 | wmat | prpplague: i haven't started on it yet, there just seems to be so much other work to do |
| 17:22.33 | wmat | prpplague: good question, i'll check |
| 17:23.01 | prpplague | points wmat to prpplagues statement above ^ |
| 17:23.09 | prpplague | <prpplague> so much to do |
| 17:27.05 | wmat | prpplague: yes, there seems to be a delete tab along the top of a page for me |
| 17:27.14 | prpplague | checks |
| 17:28.09 | wmat | prpplague: you may need increased sys privs |
| 17:28.24 | prpplague | wmat: yea no delete tab |
| 17:28.40 | prpplague | wmat: possible me to get that? i'd like to keep up on spam and redirects |
| 17:29.06 | wmat | prpplague: yes |
| 17:30.33 | prpplague | wmat: well get to it then |
| 17:30.53 | wmat | prpplague: no sysop for you! |
| 17:31.01 | prpplague | hehe |
| 17:32.02 | prpplague | hmm, i need to finish the bsp for this new cpu board, do s3c24xx support in qemu, cortex-a8 support in openocd, and do some kernel/rootfs builds for the nail |
| 17:32.59 | landley | prpplague: what's luminary? |
| 17:33.18 | landley | prpplague: dunno yet. I've now been pointed at Ubuntu Mobile and Embedded as yet another time waster... |
| 17:33.27 | prpplague | landley: a small cortex-m3 board, pretty cute, but doesn't have mmu |
| 17:33.50 | prpplague | landley: let me give you a url |
| 17:34.21 | prpplague | landley: http://www.luminarymicro.com/products/ekk-lm3s811_evaluation_kit.html |
| 17:36.06 | SpeedEvil | How much? |
| 17:37.20 | landley | Their jtag doesn't go through the usb, that I can see. |
| 17:37.22 | prpplague | most places retail it for $49 |
| 17:37.35 | prpplague | landley: yea, they use the ft2232 same as ours |
| 17:37.52 | prpplague | landley: the jtag pins you are seeing are for use with external devices |
| 17:39.42 | prpplague | landley: i'm looking at doing a qemu machine for the hammer/nail |
| 17:39.45 | prpplague | landley: thoughts? |
| 17:39.49 | landley | Yay? |
| 17:40.07 | SpeedEvil | The ST micro Primer is interesting. |
| 17:40.10 | prpplague | landley: most of the code is there from the openmoko folks |
| 17:40.13 | SpeedEvil | LCD, accelerometer, USB, ... |
| 17:40.19 | landley | The lumiary page doesn't say how much memory or flash it has. |
| 17:40.30 | prpplague | landley: so its a matter of cleaning it up and adding hammer machine definition |
| 17:40.33 | landley | SpeedEvil: does it have a battery backed up clock? |
| 17:40.36 | landley | prpplague: cool. |
| 17:41.21 | SpeedEvil | landley: yes |
| 17:41.26 | SpeedEvil | landley: IIRC |
| 17:41.46 | prpplague | landley: 8k ram and 64k flash\ |
| 17:41.55 | landley | 8k? Kilobytes? |
| 17:41.58 | landley | No wonder it's so cheap... |
| 17:42.10 | prpplague | yea |
| 17:42.12 | SpeedEvil | Yeah - these chips sell for around the $5-7 mark |
| 17:42.30 | SpeedEvil | they're basically an upgrade from 8 bit ones with 10K RAM or so. |
| 17:42.30 | landley | The processor's fine. Linux runs on nommu, that's what the uClinux guys do. |
| 17:42.38 | landley | But you'll never boot Linux in 8k. :) |
| 17:42.47 | SpeedEvil | spots a challenge! |
| 17:43.09 | SpeedEvil | If it had a MMU, you could use external RAM bit-banged. |
| 17:43.27 | SpeedEvil | Even lacking a RAM controller |
| 17:43.34 | landley | SpeedEvil: it hasn't got an mmu. |
| 17:43.41 | SpeedEvil | I know. |
| 17:43.47 | prpplague | yea |
| 17:44.03 | prpplague | landley: i was mainly looking at it for the configuration of the jtag and such |
| 17:44.05 | landley | The kernel stack needs 4k for PID 1 and 4k for the interrupt stack. |
| 17:44.15 | *** join/#edev magnetron (n=magnetro@unaffiliated/magnetron) |
| 17:44.16 | prpplague | landley: since the cortex-m3 is very similiar in nature as the cortex-a8 |
| 17:44.21 | landley | There's literally no room left over for any data structures, even if you put the whole kernel in flash. |
| 17:44.52 | prpplague | landley: yea, i barely got linux running on the pixter-mm |
| 17:45.22 | *** join/#edev magnetron (n=magnetro@unaffiliated/magnetron) |
| 17:50.24 | g1powermac | GPSFan, interesting, can't get the standard linux kernel patched with slimer to compile with buildroot toolchain |
| 17:51.11 | g1powermac | http://www.pastebin.ca/996034 |
| 17:52.51 | landley | What's slimer? |
| 17:53.07 | g1powermac | its a zipit 2 specific patch |
| 17:54.44 | GPSFan | g1powermac: that's the initramfs problem, just add an empty dir labeled rfs. that's where you are supposed to put the code that goes in the initramfs |
| 17:55.52 | g1powermac | ahh, that worked :-) |
| 17:55.54 | g1powermac | thanks |
| 17:56.01 | g1powermac | never seen that before |
| 17:56.04 | GPSFan | g1powermac: since you are going to be building modules, it doesn't matter what is or isn't in /rfs |
| 17:57.02 | g1powermac | been trying to get aufs to compile, it just annoying me atm, so I wanted to test out building the kernel first |
| 17:57.25 | g1powermac | might just try building aufs from the kernel side and rip the module out |
| 17:58.17 | g1powermac | building it outside of the kernel source just doesn't seem to want to work |
| 18:00.28 | g1powermac | meh, it died again |
| 18:00.40 | g1powermac | hmm |
| 18:00.52 | g1powermac | its a zipit specific error |
| 18:01.35 | g1powermac | http://www.pastebin.ca/996043 |
| 18:03.28 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 18:04.48 | g1powermac | any idea what that means GPSFan? |
| 18:05.54 | RobotGuy | Check these out: http://www.sparkfun.com/commerce/product_info.php?products_id=8668 |
| 18:05.58 | RobotGuy | g1powermac: Greetings |
| 18:06.31 | GPSFan | g1powermac: hmm, something to do with the zipit2 logo. what is the file size of the patch you used to patch the kernel? |
| 18:07.38 | g1powermac | heya RobotGuy |
| 18:07.48 | g1powermac | GPSFan, hmm, think it was 4mb |
| 18:08.01 | GPSFan | exact byte count please. |
| 18:09.33 | g1powermac | 4056kb |
| 18:13.57 | GPSFan | 4056x1024=4153344 which is a bit smaller than the one I used. what is the date in it, and does your kernel tree have: logo_zipit_clut224.c in drivers/video/logo/ ? |
| 18:16.29 | g1powermac | yup it has it |
| 18:17.01 | g1powermac | exact byte count btw is 4148813 |
| 18:17.11 | g1powermac | according to du |
| 18:17.34 | *** join/#edev bbradley (n=bbradley@92-235-133-146.cable.ubr17.azte.blueyonder.co.uk) |
| 18:18.31 | GPSFan | date? |
| 18:23.56 | g1powermac | yesterday, 2008-04-23 22:38 |
| 18:27.34 | g1powermac | actually, that is prolly the date I downloaded the thing |
| 18:28.33 | g1powermac | I used the one off the zipit wiki dated by magon on 1/27/08 17:27:07 |
| 18:36.17 | Dr{Who} | landely you here? |
| 18:38.07 | prpplague | landley is never entirely "here" |
| 18:38.28 | Dr{Who} | if you get back ping me.. i have a patch for tinycc for a bug i found.. and i have quesitons about your mercurial repo and how it compares to your pre2 release. |
| 18:38.34 | Dr{Who} | hehe ya... |
| 18:39.53 | Dr{Who} | he is in good company then.... |
| 18:40.07 | landley | hmmm? |
| 18:40.14 | Dr{Who} | : c ) |
| 18:40.35 | Dr{Who} | ok your mercurial repo.. seems very different than your pre2 release.. |
| 18:40.41 | landley | Um, yay patch. Please email it to tinycc@landley.net, plus a test case that triggers it if you've got one. |
| 18:40.45 | Dr{Who} | just lots of changes since you built that snapshot? |
| 18:40.53 | landley | I've done some development since -pre2. :) |
| 18:40.54 | Dr{Who} | http://pastebin.com/d5f6409ee |
| 18:40.58 | Dr{Who} | k. |
| 18:41.07 | Dr{Who} | its a silly bug. but its a start. |
| 18:41.25 | Dr{Who} | kinda shocked it exists... honestly. |
| 18:42.17 | g1powermac | GPSFan, any idea? |
| 18:42.32 | g1powermac | I'm looking at the code now where it failed, I don't get why its erroring |
| 18:42.48 | Dr{Who} | if you dont mind.. tag it with my data so i can track it later.. |
| 18:43.05 | GPSFan | g1powermac: gust got back from lunch, looking now. |
| 18:43.16 | g1powermac | ahh, k, speaking of food, I could use some :-) |
| 18:45.29 | GPSFan | g1powermac: look in drivers/video/logo and pastebin an ls -la of that dir. |
| 18:46.53 | g1powermac | http://www.pastebin.ca/996097 |
| 18:47.19 | g1powermac | btw, just to let you know, the timestamps are off due to the clock being off on this vm |
| 18:48.31 | GPSFan | built-in.o should be ~19000 not ~6000 so something's hosed. |
| 18:50.44 | GPSFan | g1powermac: there is no logo_zipit_clut224.o. So maybe the logo isn't enabled in the .config |
| 18:54.34 | g1powermac | hmm |
| 18:55.26 | g1powermac | its enabled |
| 18:58.32 | g1powermac | just to let you know, this is just a stock kernel with the slimer patch, no aufs stuff yet |
| 19:08.53 | g1powermac | I assume you're not having the same issue over there. . . |
| 19:10.06 | RobotGuy | I think it is going to be pretty easy to add the Phidgets libraries to buildroot. |
| 19:16.15 | GPSFan | g1powermac: correct. my kernel builds to completion. I didn't use the slimmer patch though. |
| 19:16.34 | g1powermac | hmm |
| 19:17.00 | g1powermac | I couldn't even get the original patch to patch the kernel |
| 19:18.29 | GPSFan | odd indeed. |
| 19:19.02 | solar | prpplague: around by chance? |
| 19:19.06 | g1powermac | lemme try the other patch again |
| 19:19.59 | prpplague | solar: yep |
| 19:20.15 | prpplague | solar: just banging my head on the desk |
| 19:20.28 | prpplague | solar: i can stop for a momment though, hehe |
| 19:20.37 | g1powermac | yup, won't patch, lemme pastebin it |
| 19:20.48 | GPSFan | g1powermac: the first patch that they had wouldn't apply ckeanly, I previewd it and the supposedly fixed it. IIRC the next one they sent me applied ok. |
| 19:21.13 | g1powermac | oh |
| 19:21.16 | g1powermac | hmm |
| 19:21.29 | g1powermac | I'm using the one of their wiki |
| 19:21.30 | prpplague | solar: whats cookin today? |
| 19:21.32 | g1powermac | *off |
| 19:21.43 | g1powermac | lemme still pastebin the error |
| 19:22.19 | GPSFan | my "new patch" from them is dated jan 8 on many of the diff's inside. |
| 19:23.01 | g1powermac | http://www.pastebin.ca/996162 |
| 19:27.14 | g1powermac | maybe I should just get their z2kernel from the svn |
| 19:27.20 | RobotGuy | Does the sc32410 have the camera interface? |
| 19:27.28 | RobotGuy | s3c2410 that is |
| 19:27.28 | GPSFan | g1powermac: I have the same code in the patch I have. let me gram a clean kernel and try patching it. what are you using for a patch command. |
| 19:27.35 | prpplague | RobotGuy: no |
| 19:27.51 | g1powermac | patch -p1 < path/to/patch |
| 19:27.57 | RobotGuy | prpplague: Drats. Another reason for me to want the s3c2440 |
| 19:34.29 | GPSFan | g1powermac: I just succesfully patched a kernel with the patch I have dated jan 8 4155746 bytes. btw the dir for the kernel source needs to be linux. |
| 19:40.14 | g1powermac | ahh |
| 19:40.17 | GPSFan | g1powermac: just re-downloaded the patch from the zipit site and it is idednical to the one I used. |
| 19:40.17 | g1powermac | ok |
| 19:40.44 | GPSFan | if the dir ins't named linux bad-things(tm) will happen |
| 19:41.46 | g1powermac | hmm, why do I have this feeling that it also needs to be a dir called build_tree. . . |
| 19:42.10 | GPSFan | not for me |
| 19:42.28 | g1powermac | well, it still failed for me even when changing it to just linux |
| 19:43.38 | GPSFan | that patch created a build-tree dir for me |
| 19:44.18 | g1powermac | what command did you use to patch it? |
| 19:45.39 | GPSFan | patch -p1 < patchname the patch was in the same dir as the linux dir |
| 19:45.40 | landley | DrWho: could you /msg me with a name to attach to the commit message for your tinycc fix? |
| 19:45.52 | g1powermac | hmm, maybe that makes a difference |
| 19:46.03 | landley | DrWho: the bug exists because I broke options.c off from tcc.c fairly recently. |
| 19:46.39 | g1powermac | nope |
| 19:46.49 | g1powermac | weird |
| 19:47.27 | g1powermac | it keeps whining about build_tree/linux blah |
| 19:48.02 | GPSFan | g1powermac: http://www.pastebin.ca/996204 see where I tried other patch commands? |
| 19:48.03 | g1powermac | where is you kernel source located? |
| 19:48.51 | g1powermac | weird |
| 19:49.26 | GPSFan | http://www.pastebin.ca/996206 |
| 19:50.58 | g1powermac | whats in that build_tree dir? |
| 19:51.48 | GPSFan | the patch added it. |
| 19:52.18 | GPSFan | I think when I did one of the other patch commands |
| 19:52.37 | g1powermac | hmm |
| 19:52.49 | g1powermac | did you try just doing the p1 on a clean source? |
| 19:55.15 | GPSFan | here is a clean one: http://www.pastebin.ca/996216 |
| 19:55.27 | g1powermac | hah, I just got it to patch |
| 19:55.48 | robbat2 | prpplague, I don't know if solar pinged you yet this morning, but I got time to bring up the Nail board he got, and just oddly found that the g_ether stuff isn't built at all on there. |
| 19:56.01 | g1powermac | -p2 made it patch after playing with the others |
| 19:56.12 | robbat2 | prpplague, also, either the busybox rx tool or my terminal app (xc) is corrupting Xmodem data transfers |
| 19:56.59 | GPSFan | g1powermac: I think it may be because of how zipit created that patch, diff -u as opposed to diff -urn |
| 19:57.11 | Dr{Who} | landely.. oh i thought it existed on 0.9.24 too. |
| 19:57.45 | RobotGuy | Dr{Who}: Greetings. How are you doing with your Hammer? |
| 19:57.49 | g1powermac | yup, just tried it on a clean source, p2 does it here |
| 19:58.27 | Dr{Who} | its fine.. i am trying to get it running tinycc and able to compile so far not a lot of progress. |
| 19:58.47 | Dr{Who} | seems like 1 brick wall after another.. but that is new : c ) |
| 19:58.48 | RobotGuy | Dr{Who}: tinycc on Hammer for native builds? |
| 19:58.56 | Dr{Who} | yes |
| 19:59.48 | Dr{Who} | and just because i prefer C over anything else. |
| 20:01.23 | g1powermac | GPSFan, yup, its compiling now |
| 20:01.30 | Dr{Who} | this is odd binutils.mk under the toolchain build package has settings as if it could build bintuils for the target.. but i cant find any way to enable it? |
| 20:01.36 | GPSFan | g1powermac: ;>) |
| 20:01.53 | Dr{Who} | i want to get objdump onto the hammer. save time in debugging. |
| 20:02.11 | g1powermac | GPSFan, wonder why I needed p2 here but it worked with p1 over by you |
| 20:02.34 | GPSFan | g1powermac: I think it may be because of how zipit created that patch, diff -u as opposed to diff -urn |
| 20:02.44 | g1powermac | hmm |
| 20:03.41 | g1powermac | anyway, thanks for the help :-) |
| 20:05.16 | RobotGuy | Dr{Who}: I'm building a 2.6.25 kernel now. |
| 20:06.59 | GPSFan | g1powermac: np, you chech out this link that RobotGuy posted earlier. might it work with the z2? http://www.sparkfun.com/commerce/product_info.php?products_id=8668 |
| 20:08.24 | RobotGuy | I think that is a cool little camera and full data is available for it. |
| 20:08.56 | g1powermac | woah |
| 20:09.03 | GPSFan | RobotGuy: indeed, sparkfun needs a breakout board for it too. |
| 20:09.10 | g1powermac | its the holy grail of cameras I've been looking for ;-) |
| 20:09.11 | RobotGuy | GPSFan: Agreed! |
| 20:09.42 | GPSFan | g1powermac: they have a 640 x 480 one as well |
| 20:09.50 | GPSFan | ame price |
| 20:09.58 | GPSFan | s/ame/same |
| 20:10.03 | g1powermac | dude, this sucker should work, I'll have to look at it more closely |
| 20:11.19 | g1powermac | man, this completes my goal :-) |
| 20:11.29 | g1powermac | woot, the kernel compiled completely |
| 20:11.38 | GPSFan | g1powermac: ;>) |
| 20:12.14 | RobotGuy | A small breakout board that would match holes on http://www.lynxmotion.com/Product.aspx?productID=397&CategoryID=8 would be great. |
| 20:13.57 | GPSFan | RobotGuy: that should be easy to do "except" for the connector. almost 20 pins |
| 20:14.32 | RobotGuy | GPSFan: It could be done with a 2x10 connector. |
| 20:14.33 | GPSFan | in that small a housing it would be unfriendly to hobbyists |
| 20:14.53 | GPSFan | maybe |
| 20:15.01 | RobotGuy | GPSFan: 2x10 right angle connector should work fine. |
| 20:15.47 | RobotGuy | All that would be needed are appropriate length standoffs. |
| 20:16.40 | RobotGuy | Connector could come off the top of the breakout board, which would be ideal. |
| 20:17.28 | GPSFan | RobotGuy: suggest that to Sparkfun, they have lots of other breakoout boards, one geared towards robot-folks might be of interest. |
| 20:17.43 | RobotGuy | GPSFan: I think I will do that! |
| 20:18.19 | GPSFan | RobotGuy: is the s3c2440 camera interface compatible with that camera? |
| 20:18.30 | g1powermac | hah, awesome |
| 20:18.33 | g1powermac | aufs compiled |
| 20:21.09 | g1powermac | this should be interesting |
| 20:23.53 | prpplague | robbat2: i suspect the boss didn't reload the kernel and rootfs for the nail, as we pre-load all the hammer modules with the same kernel/rootfs |
| 20:24.11 | robbat2 | prpplague, not a problem, i'll have something cross-built here tmrw probably |
| 20:24.32 | prpplague | robbat2: give me a few and i'll get a kernel up for you |
| 20:24.32 | RobotGuy | GPSFan: I just submitted my suggestion to Spark Fun. |
| 20:24.43 | GPSFan | RobotGuy: ;>) |
| 20:24.47 | prpplague | robbat2: what kind of problem are you having with the file transfers |
| 20:24.52 | prpplague | robbat2: ?? |
| 20:25.01 | robbat2 | prpplague, solar had an ext image he wanted to loopmount on there |
| 20:25.06 | RobotGuy | GPSFan: I don't know if the 2440 camera interface is compatible yet, because I haven't checked the data for each. |
| 20:25.06 | robbat2 | since there wasn't any networking |
| 20:25.11 | robbat2 | i was trying to send it via Xmodem |
| 20:25.25 | robbat2 | busybox's 'rx' |
| 20:25.46 | robbat2 | just that the data on the receiving side seems to come out garbled |
| 20:26.07 | prpplague | robbat2: that sounds odd, i've not run across that before |
| 20:26.16 | robbat2 | what term app do you use? |
| 20:26.22 | prpplague | minicom |
| 20:26.27 | robbat2 | the possibility exists that xc's xmodem impl is broken |
| 20:26.34 | prpplague | but i've tested with a number of others, but not xc |
| 20:26.48 | robbat2 | one sec, i'll test with minicom |
| 20:30.33 | g1powermac | GPSFan, I have a good feeling that camera should work |
| 20:30.50 | g1powermac | seems the interfaces line up |
| 20:31.22 | GPSFan | g1powermac: !!!! |
| 20:34.44 | g1powermac | will have to get abit of clarification for drmikecrowe about one or two of the pins |
| 20:34.49 | g1powermac | but it seems good |
| 20:35.05 | g1powermac | s/for/from |
| 20:35.18 | robbat2 | weird. i think my minicom-foo has atrophied. it refuses to go into online mode |
| 20:36.13 | prpplague | robbat2: try starting it with minicom -s |
| 20:36.23 | prpplague | robbat2: do the setup |
| 20:36.53 | prpplague | robbat2: actually thats what landley reported with his setup |
| 20:37.42 | prpplague | robbat2: you are starting minicom with what command? |
| 20:38.01 | prpplague | robbat2: i can;'t seem to find any sources for xcomm |
| 20:40.11 | prpplague | robbat2: nm found some |
| 20:42.08 | prpplague | robbat2: "Note that you should replace your current rz and sz with included |
| 20:42.08 | prpplague | versions which are known to work with XC. Other versions of rz/sz |
| 20:42.08 | prpplague | which open the comm port directly should *NOT* be used" |
| 20:45.41 | RobotGuy | robbat2: You don't need online to console to the Hammer |
| 20:46.26 | RobotGuy | I blanked out all result codes and modem commands in my minicom setup. |
| 20:46.26 | GPSFan | g1powermac: how did you finally get qtopia to build? |
| 20:46.46 | g1powermac | GPSFan, oh yea, I thought I told you |
| 20:47.01 | GPSFan | is all ears |
| 20:47.15 | g1powermac | it compiled just fine :-) |
| 20:47.55 | landley | drwho: I had nothing to do with 0.9.24, other than that they sucked a lot of code out of my mercurial repository. :) |
| 20:49.47 | robbat2 | prpplague, still nothing after blanking the minicom init settings (I had some of them blanked already). I reasonably trust the xc Xmodem impl, because I've used it to upload firmware to Force10 switches before (which had firmware checksums that passed) |
| 20:49.53 | landley | prpplague: the term program in KDE (Konsole) has rz and sz support built in. :) |
| 20:50.04 | robbat2 | rz/sz are zmodem, I was doing xmodem |
| 20:50.08 | prpplague | landley: ahh |
| 20:50.20 | landley | In the pulldown menus: "edit->zmodem upload". |
| 20:50.21 | prpplague | robbat2: i still can't find xcomm source that compiles |
| 20:50.36 | prpplague | robbat2: you got source for that somewhere? |
| 20:50.36 | landley | It calls the command line utility, of course... |
| 20:50.49 | robbat2 | http://www.ibiblio.org/pub/Linux/apps/serialcomm/dialout/xc-4.3.2.tar.gz |
| 20:50.49 | landley | uses netcat -f to talk to the hammer. |
| 20:50.57 | robbat2 | there's a patch to add 115200 support |
| 20:51.34 | prpplague | robbat2: this does seem very odd that your minicom won;t talk , morton had the same problem as well as landley |
| 20:51.37 | ce_geek | prpplague: Damn. No luck so far with the DS1820. Time to bring out the O-scope to see what's going on. |
| 20:51.45 | RobotGuy | robbat2: Are you using buildroot? The minicom there works perfect with Hammer at 115200. |
| 20:51.51 | landley | Did andrew morton try to play with the hammer already? |
| 20:51.55 | prpplague | ce_geek: bummer |
| 20:51.58 | landley | I know David Mandala did... |
| 20:52.10 | landley | still needs to send akpm directions on how to play with his hammer. |
| 20:52.10 | prpplague | landley: only from what i know you said you were trying to use it at celf |
| 20:52.17 | landley | That was david mandala. |
| 20:52.21 | robbat2 | RobotGuy, just the stock minicom I installed quickly. I normally use xc, which gives me a terminal fine, just the xmodem stuff seemed broken |
| 20:52.32 | landley | akpm was keeping it in its box until he got time-saving instructions. |
| 20:52.42 | GPSFan | g1powermac: I knew that you had got it to compile, my question is how? |
| 20:52.46 | landley | minicom seems somewhat bit-rotted. |
| 20:52.54 | RobotGuy | robbat2: Oh, wait, ignore what I just said. |
| 20:52.58 | landley | I think it needs the DTR line enabled or it will NEVER talk to the serial port, or some such. |
| 20:53.09 | RobotGuy | robbat2: What distro are you using? |
| 20:53.16 | robbat2 | i'm a Gentoo guy |
| 20:53.32 | g1powermac | GPSFan, I did a change to the uclibc config according to this: http://www.busybox.net/lists/buildroot/2007-September/005338.html |
| 20:53.35 | robbat2 | previously did some porting to embedded mips gear |
| 20:53.45 | robbat2 | mycable.de XXS1500 box, fun little thing |
| 20:53.53 | RobotGuy | robbat2: I am surprised that minicom doesn't have 115200 support. |
| 20:54.15 | robbat2 | RobotGuy, err, I never said it didn't. |
| 20:54.19 | solar | prpplague: nevermind it looks like robbat2 found you about the things we ran across. |
| 20:54.34 | robbat2 | just that it doesn't seem to work where xc does |
| 20:54.58 | RobotGuy | I've never had problems using minicom with Hammer at 115200 |
| 20:55.00 | solar | just fixed xc in our tree btw. it did not cross-compile very well. |
| 20:55.02 | robbat2 | ah, /me stabs minicom flow control |
| 20:55.31 | prpplague | yea looking at xc now |
| 20:55.37 | GPSFan | g1powermac: yeah, it compiles fine for me under buildroot, but it jus doesn't build any executables, lots of .o's all compiled for arm, but there is something hosed in the buildroot make (still after people patching it yesterday & today). You built it out side buildroot. I'd like to know what you did. |
| 20:56.12 | robbat2 | prpplague, grab the patches from Gentoo for it, fixed the makefile, adds 115200 (one line), plus some #includes for newer gcc/libc |
| 20:56.42 | prpplague | gentoo.org? |
| 20:57.48 | prpplague | robbat2: sorry i'm not too familair with gentoo layout |
| 20:57.55 | robbat2 | http://sources.gentoo.org/viewcvs.py/gentoo-x86/net-dialup/xc/files/ |
| 20:58.07 | g1powermac | GPSFan, well, I built the toolchain with that modification, and then made a device profile for the z2 in qtopia: http://doc.trolltech.com/qtopia4.3/buildsystem/over-device-profiles-1.html |
| 20:58.38 | prpplague | robbat2: thanks |
| 20:58.40 | ce_geek | prpplague: Okay. Looks like digitemp isn't doing anything on TXD2 when I tell it to search for devices on /dev/ttySAC2 |
| 20:58.59 | GPSFan | g1powermac: k, could you paste your device profile? |
| 20:59.02 | prpplague | ce_geek: have you got ttySAC2 active? |
| 20:59.09 | prpplague | ce_geek: check it with devmem |
| 20:59.13 | g1powermac | I will |
| 20:59.18 | ce_geek | prpplague: How do I do that? :-/ |
| 20:59.26 | prpplague | hehe |
| 20:59.44 | ce_geek | Tricky? |
| 20:59.49 | GPSFan | g1powermac: thanks, if I can get it to build outside BR maybe I can figure out why it is'nt building inside. |
| 21:00.55 | prpplague | ce_geek: devmem 0x56000070 |
| 21:01.02 | prpplague | ce_geek: that should give you the current config |
| 21:01.12 | ce_geek | At what prompt do I do this? |
| 21:01.28 | prpplague | ce_geek: just at a standard shell prompt |
| 21:01.33 | ce_geek | I must not have devmem installed |
| 21:01.50 | prpplague | ce_geek: ahh you need to have that, good tool for debugging |
| 21:02.00 | RobotGuy | prpplague: The spam filter on the forums seems to be working really well. Could they restore the stay logged in feature now? |
| 21:02.13 | ce_geek | prpplague: Cool. What's the significance of the address you gave? |
| 21:02.17 | prpplague | ce_geek: let me ask, did you include the init for uart2 in your kernel? |
| 21:02.38 | ce_geek | prpplague: Of course not :) |
| 21:02.39 | prpplague | ce_geek: look at page 273 of the s3c2410 users manual pdf |
| 21:02.51 | ce_geek | prpplague: Looking now, thanks. |
| 21:03.03 | robbat2 | prpplague, ok, the minicom xmodem transferred it almost perfectly |
| 21:03.21 | robbat2 | it added 7 bytes onto the end, but chopping those off, the md5sum match on src+dst |
| 21:03.46 | prpplague | robbat2: oh you got minicom to work? |
| 21:03.50 | prpplague | robbat2: what did you have to do? |
| 21:03.55 | robbat2 | prpplague, muck with the flow control |
| 21:03.57 | RobotGuy | I got a 2.6.25 kernel to build. :D |
| 21:04.22 | prpplague | robbat2: hmm |
| 21:04.39 | solar | robbat2: was the end result supposed to be that the device lacks network abilities? |
| 21:05.05 | robbat2 | solar, yeah, g_ether wasn't built on here, one sec, i'm just about to loopmount that file you gave me |
| 21:05.11 | ce_geek | prpplague: I think I'm feeling better about getting this temp sensor talking on a web page being a sufficient deliverable for Rick..... :) |
| 21:05.11 | prpplague | solar: yea looks like the hammer module was just loaded with the default hammer kernel and not the nail board kernel |
| 21:05.51 | robbat2 | lol, ok, will need a new kernel. not even loop is built into this one |
| 21:05.55 | ce_geek | prpplague: Okay so I need devmem and I need some goodies in the kernel.... |
| 21:06.14 | ce_geek | prpplague: Any idea what it's literally called in the kernel menu? |
| 21:06.28 | prpplague | ce_geek: one sec too many items going |
| 21:06.54 | robbat2 | i gotta do some real work, i'll have more time tmrw |
| 21:06.57 | robbat2 | take care |
| 21:07.32 | ce_geek | prpplague: No problem take your time, I've got plenty. |
| 21:08.20 | prpplague | ce_geek: http://www.pastebin.ca/996309 |
| 21:08.50 | prpplague | ce_geek: you can edit your arch/arm/mach-s3c2410/mach-tct_hammer.c to include those commented section to turn on uart2 |
| 21:09.05 | prpplague | robbat2: i'll prep you a kernel with loop |
| 21:09.15 | prpplague | robbat2: is there any other items you want turned on? |
| 21:09.17 | robbat2 | and g_ether ;-) |
| 21:09.25 | prpplague | robbat2: yea that too |
| 21:10.12 | RobotGuy | Ooooooooooooh, my 2.6.25 kernel is 998 Kb. |
| 21:10.52 | solar | prpplague: when you update the kernel can you post the new kernel.config somewhere I/we can link to it please. thanks in advance |
| 21:11.18 | prpplague | solar: will do, i'll get that posted in the morning on the nail page |
| 21:11.31 | prpplague | solar: just out of curiousity, whats on the image you are going to loop mount? |
| 21:11.58 | prpplague | ce_geek: were you able to follow those mods? |
| 21:12.36 | RobotGuy | quickly edits arm/arc/mach-s3c2410/mach-tct-hammer.c to change .SIZE |
| 21:12.42 | ce_geek | I'm editing the kernel stuff now, going to compile in a sec. |
| 21:12.51 | ce_geek | prpplague: Is devmem included in the buildroot menu? |
| 21:12.53 | prpplague | ce_geek: oops wait i see a typo |
| 21:13.18 | prpplague | ce_geek: not currently its pretty small doesn't take but a second to compile |
| 21:14.03 | prpplague | ce_geek: that option in for the init should be TXD2 and RXD2 not TXD0 and RXD0 |
| 21:14.17 | prpplague | ce_geek: i update the pastebin |
| 21:14.26 | ce_geek | cool thx |
| 21:15.00 | prpplague | ce_geek: guess i should add devmem to the build |
| 21:15.03 | prpplague | makes a note |
| 21:16.27 | prpplague | ce_geek / solar / robbat2 i appreciate your patience , we are still trying to "learn" what kind of things people are asking for as far as documents and source |
| 21:16.51 | ds2 | read the source... don't need no stinkin documentation :P |
| 21:17.23 | robbat2 | source for code documentation, decent electrical diagrams and pinouts for the hardware |
| 21:17.59 | robbat2 | but i'm impressed by the docs thusfar |
| 21:18.30 | prpplague | robbat2: yea, all of the schematics and pinouts should be on your cd |
| 21:18.31 | robbat2 | including all the component datasheets on the CD was a nice move too prpplague++ |
| 21:18.39 | prpplague | robbat2: thanks |
| 21:18.45 | ce_geek | prpplague: Why not enable it by default? |
| 21:19.28 | prpplague | ce_geek: not everyone is going to use those as serial ports |
| 21:19.35 | prpplague | ce_geek: they might use them as gpio |
| 21:19.43 | prpplague | ce_geek: but i can add a menu option |
| 21:19.44 | ce_geek | prpplague; oh, right. yes. Good call. |
| 21:19.58 | ce_geek | prpplague; So it would come up in the make menuconfig? |
| 21:20.41 | prpplague | yea |
| 21:21.23 | prpplague | ce_geek: http://www.elinux.org/Hammer_Board_Software_System |
| 21:21.27 | prpplague | ce_geek: binary and source added |
| 21:21.31 | prpplague | ce_geek: for devmem |
| 21:21.52 | ce_geek | prpplague: Damn nice, kthx. |
| 21:22.11 | ce_geek | prpplague: Any chance it'll just *work* after I fix the kernel? |
| 21:22.19 | ce_geek | prpplague: Can it be temporarily fixed with devmem? |
| 21:22.40 | prpplague | ce_geek: hehe yea you probably could use devmem to tweak the registers |
| 21:22.47 | prpplague | ce_geek: devmem is very useful |
| 21:23.07 | prpplague | ce_geek: i usually use devmem to iron out alot of settings before i do an actual kernel driver |
| 21:23.25 | ce_geek | prpplague: You told me to get familiar with it, I have just been occupied with other more pressing issues. It's nice to be at a point where it'll be useful though. |
| 21:23.35 | prpplague | ce_geek: hehe np |
| 21:23.52 | ce_geek | prpplague; Cool. I'm hoping that this temp thing works and then I want to do stuff with blinking LEDs and such. |
| 21:23.55 | prpplague | ce_geek: if you want to send over devmem to your hammer, we can have a look at the register values |
| 21:24.03 | ce_geek | prpplague: Okay. one sec. |
| 21:25.37 | prpplague | ce_geek: not sure why that wiki always wants to cap file names |
| 21:25.51 | prpplague | ce_geek: so you'll need to change it to devmem and not Devmem |
| 21:25.58 | ce_geek | prpplague: Yup, pastebinning |
| 21:26.14 | ce_geek | http://www.pastebin.ca/996330 |
| 21:27.35 | ce_geek | prpplague; Okay so that's configuring "Port H" pins. |
| 21:27.49 | prpplague | ce_geek: ok so that indicates that GPH6 and GPH7 are configured for RTS and CTS |
| 21:28.05 | prpplague | ce_geek: to change that we do: devmem 0x56000070 w 0x0016aaaa |
| 21:28.26 | ce_geek | prpplague; Mind reminding me what RTS and CTS are for? |
| 21:28.54 | prpplague | ce_geek: rs-232 control lines, Ready To Send and Clear To Send |
| 21:28.59 | prpplague | ce_geek: basic flow control |
| 21:29.23 | ce_geek | prpplague; Oh yes yes yes yes those pins we usually ignored for the RS232 labs we did but discussed in class :) |
| 21:29.31 | prpplague | ce_geek: yea |
| 21:29.35 | ce_geek | prpplague: And hurray my TXD2 pin is doing things. *hooks up sensor* |
| 21:31.50 | RobotGuy | We need to put info on how to init the second serial port on the Wiki. |
| 21:32.10 | prpplague | RobotGuy: yea, just been swampped this week |
| 21:32.21 | prpplague | RobotGuy: i'm hoping to have some time next week |
| 21:32.23 | ce_geek | grumbles... no dice. |
| 21:32.24 | RobotGuy | prpplague: No problem, I can do it when I get the info. |
| 21:32.29 | ce_geek | Time to investigate further. |
| 21:32.46 | prpplague | ce_geek: double check you serial port settings with stty |
| 21:32.52 | prpplague | ce_geek: stty -F /dev/ttySAC2 |
| 21:33.02 | RobotGuy | I don't mind taking care of Wiki stuff. |
| 21:33.26 | ce_geek | http://www.pastebin.ca/index.php |
| 21:33.33 | ce_geek | ...awesome. http://www.pastebin.ca/996339 |
| 21:34.28 | prpplague | ce_geek: set the baud rate to 115200 and then back to 9600 |
| 21:34.43 | prpplague | ce_geek: just to make that after we turned on the serial port it got initialized |
| 21:34.51 | RobotGuy | ce_geek: Are you sure that paste is correct? |
| 21:34.51 | prpplague | ce_geek: stty -F /dev/ttySAC2 115200 |
| 21:35.04 | prpplague | ce_geek: stty -F /dev/ttySAC2 9600 |
| 21:35.41 | ce_geek | http://www.pastebin.ca/996340 |
| 21:36.02 | prpplague | ce_geek: try you digitemp again |
| 21:36.42 | RobotGuy | ce_geek: Is the rest of that input to the stty command? |
| 21:37.03 | ce_geek | RobotGuy: No, that's what it spits out |
| 21:37.13 | RobotGuy | ce_geek: Oh, OK |
| 21:37.30 | prpplague | ce_geek: what does digitemp say when you run it? |
| 21:37.32 | ce_geek | prpplague: Hrm. No such luck. Trying to see if the temp sensor talks back. |
| 21:38.01 | ce_geek | http://www.pastebin.ca/996346 |
| 21:38.06 | CosmicPenguin | Don't take any lip from a back-talkin' sensor |
| 21:38.57 | ce_geek | :) |
| 21:39.10 | prpplague | ce_geek: just for kicks, might swap txd and rxd lines |
| 21:39.19 | RobotGuy | throws a blanket over CosmicPenguin |
| 21:39.41 | prpplague | pulls up the schematic |
| 21:40.23 | ce_geek | prpplague, http://www.lecad.uni-lj.si/~leon/other/wlan/wrt54ow/1-wire-pnp.png |
| 21:40.37 | prpplague | looks |
| 21:41.08 | ce_geek | I'm not happy with the DC voltage at the emitter of my PNP, it's like a volt. |
| 21:41.19 | ce_geek | Must not be completely cut off. |
| 21:41.34 | ds2 | you got a emitter resistor? |
| 21:41.38 | ce_geek | 4.7M resistor from TXD2 to the base |
| 21:41.55 | ce_geek | ds2: yeah |
| 21:42.09 | prpplague | ce_geek: 4.7M resistor? |
| 21:42.18 | ce_geek | er 4.7k |
| 21:43.01 | prpplague | ce_geek: hmm that should work |
| 21:43.11 | solar | prpplague: iirc. It's just a rebuild of the base firmware more or less with a little extra cruft. mainly wanted to just test that the ELF's run etc. |
| 21:43.20 | ce_geek | oooh okay so I see the sensor barking back, it does send some stuff. |
| 21:43.27 | prpplague | ce_geek: only thing i can think might happen is its not totally initialized, i'd try with the gpio's set properly in the kernel |
| 21:43.38 | ce_geek | Maybe it's time to short RX and TX and see if that works. |
| 21:43.45 | prpplague | solar: ahh ok\ |
| 21:44.09 | prpplague | ce_geek: simple loopback |
| 21:44.12 | ce_geek | yea |
| 21:44.30 | ce_geek | Hrm I wonder..... not sure who's sending what actually. Might need to get another probe to verify that. |
| 21:47.32 | ce_geek | re-flashing |
| 21:48.23 | ce_geek | er Jtaging or whatever the appropriate verb is |
| 21:48.42 | prpplague | ce_geek: hmm, i ordered some of those maxim chips, but they probably won;t be here untill monday |
| 21:48.58 | prpplague | ce_geek: you are using the circuit with a single transistor , correct? |
| 21:49.05 | ce_geek | prpplague: yeah. |
| 21:49.10 | ce_geek | prpplague: What's this about the maxim chips? |
| 21:49.22 | *** join/#edev CosmicPenguin (n=nobody@163.181.251.103) |
| 21:49.22 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 21:49.26 | ce_geek | prpplague: (I'm not using one....) |
| 21:49.55 | prpplague | ce_geek: oh which one are you using? |
| 21:51.07 | prpplague | ce_geek: i thought you were using the DS1820 |
| 21:51.27 | ce_geek | prpplague: Yeah |
| 21:51.34 | ce_geek | prpplague: Oh I thought you were talking about level shifting things from maxim |
| 21:51.47 | prpplague | ce_geek: hehe, np |
| 21:52.04 | ce_geek | I see now, maxim makes this thing or sells it or something. I associated "Dallas" with the manufacturer. |
| 21:52.19 | prpplague | ce_geek: maxim purchase dallas a few years back |
| 21:52.22 | ce_geek | ah |
| 21:53.01 | ds2 | the DS1820 seems expensive |
| 21:53.28 | ce_geek | ds2: A bit. :-/ |
| 21:53.46 | g1powermac | hmm, for that camera on sparkfun, it takes a 2.8v supply, yet the zipit stuff is 3.3v, what do you guys thing I need to do to interface the two? |
| 21:53.47 | ds2 | ce_geek: might be cheaper to get a MSP430 and use it as a slave |
| 21:54.46 | prpplague | ce_geek: which transistor are you using ? |
| 21:55.07 | ce_geek | 2n3904 |
| 21:55.28 | ds2 | ce_geek: how far away is your sensor? |
| 21:55.31 | ce_geek | ds2: Just doing a project for a class with parts I had in my goodie box. Good to know for future use though. |
| 21:55.38 | ce_geek | ds2: Inches? |
| 21:55.48 | ds2 | ah |
| 21:56.25 | ds2 | the DS1820 are great for home automation type uses (sensor on a RJ45 connected to a 1Wire bus through out the house |
| 21:56.34 | ds2 | but the $$$ |
| 21:57.38 | ce_geek | ds2: yeah :( |
| 21:58.38 | ce_geek | Awe spew. No dice. |
| 21:58.43 | *** join/#edev Redhatter (n=vk4fsjl@2001:388:f000:0:0:0:0:279) |
| 21:58.56 | ce_geek | Does more interesting things on the scope when I give digitemp the right tty though :) |
| 21:59.35 | MonMotha | 1-wire also doesn't exactly bus well over long distances |
| 21:59.39 | MonMotha | it's kinda sensitive to capacitance |
| 22:01.58 | ce_geek | prpplague: I'm going to go get a different PNP from the parts cabinet. Mabe a 2n3906 (?) or something. We have a few different ones. |
| 22:02.13 | ce_geek | prpplague; Should be saturated, doesn't seem to be. |
| 22:02.31 | ds2 | MonMontha: I have it running over some pretty decent distances (donno the numbers) |
| 22:02.56 | prpplague | ce_geek: might i make a suggestion? |
| 22:03.29 | prpplague | ce_geek: you might try to divide and conquer on this |
| 22:03.57 | prpplague | ce_geek: get the second uart working first, i.e connect it up to your pc and do some tests |
| 22:04.16 | prpplague | ce_geek: then get your sensor working on your pc |
| 22:04.28 | ds2 | ce_geek: are you using the DS/Maxim reference code or the Linux kernel driver? |
| 22:04.48 | prpplague | ce_geek: its long and would have to use some level shifters, but it would be more definitive |
| 22:04.56 | ce_geek | I'm an idiot |
| 22:05.03 | prpplague | ds2: he's using the digitemp app |
| 22:05.05 | prpplague | ce_geek: uh oh |
| 22:05.07 | ce_geek | A complete idiot. |
| 22:05.15 | *** join/#edev landley_ (n=landley@adsl-68-93-84-18.dsl.hstntx.swbell.net) |
| 22:05.15 | prpplague | ce_geek: you use the wrong part? |
| 22:05.17 | ce_geek | The 3n2904 is an NPN |
| 22:05.26 | ds2 | that one should just "work" |
| 22:05.27 | ds2 | DOH |
| 22:05.39 | prpplague | ~lart ce_geek for using the wrong part |
| 22:05.39 | purl | pushes the wall down onto ce_geek whilst whistling innocently for using the wrong part |
| 22:05.41 | g1powermac | GPSFan, the other camera is nice too, even though it has less of a resolution, you get 30fps |
| 22:06.30 | g1powermac | wonders if he should get one of each |
| 22:06.57 | g1powermac | though I'm not sure what to do with the 3.3v to 2.8v difference. . . |
| 22:07.22 | ce_geek | prpplague: BINGO! |
| 22:07.35 | ce_geek | does a little jig |
| 22:07.58 | ds2 | g1powermac: it may just work...that's around one diode difference |
| 22:08.06 | ce_geek | http://www.pastebin.ca/996393 |
| 22:09.07 | g1powermac | ds2, yea |
| 22:09.21 | ce_geek | Wow, even the one I thought i cooked earlier (reverse biased....) works! |
| 22:11.43 | ce_geek | prpplague: Thanks for your help guy, this is cool. |
| 22:12.21 | g1powermac | hmm, according to the datasheet, it might tolerate 3.3v, though it is above the recommended top level of 3v |
| 22:13.21 | prpplague | hey hey got it working? |
| 22:13.35 | prpplague | ce_geek: good deal |
| 22:13.35 | ce_geek | prpplague: Yup! |
| 22:13.49 | ds2 | g1powermac: what about drive all the lines through a 1N914 + a resistor to pull down? |
| 22:14.25 | prpplague | ce_geek: looks good |
| 22:14.38 | ce_geek | http://www.pastebin.ca/996398 |
| 22:14.40 | g1powermac | hmm, might work |
| 22:14.51 | ce_geek | prpplague: (Temperature output this time....) |
| 22:15.18 | prpplague | ce_geek: do me a good paste so i can email the boss |
| 22:15.20 | prpplague | hehe |
| 22:15.46 | ce_geek | prpplague: What do you want included in the paste? |
| 22:16.00 | ce_geek | oh one without junk int it? :) |
| 22:16.14 | prpplague | ce_geek: oh i thought you were going to paste one with some temp values |
| 22:16.39 | ce_geek | prpplague: does this work? http://www.pastebin.ca/996401 |
| 22:16.52 | g1powermac | ds2, I could prolly just send all the lines coming out of the zipit through diodes that I have loads of too |
| 22:16.54 | prpplague | ce_geek: lovely |
| 22:17.03 | ce_geek | prpplague: Excellent. |
| 22:17.26 | ds2 | g1powermac: shouldn't you also have a pull down since the diode would prevent it from driving it low |
| 22:17.29 | g1powermac | and I think the camera will get a high enough voltage on the io to signal a high level to the zipit |
| 22:17.34 | ce_geek | prpplague: I think it's beer:30 now, but tonight or tomorrow I'll document the crap out of it and then figure out how to do a decent web deal. There are a few options for that, I thought I saw one with graphs. |
| 22:17.46 | g1powermac | hmm, true |
| 22:17.50 | ds2 | or toss in one of those LVCX or like chips |
| 22:18.14 | prpplague | ce_geek: good deal |
| 22:18.17 | prpplague | ce_geek: nice work |
| 22:18.23 | ce_geek | prpplague: Digitemp's page shows a graphic. Wonder if I can do that. |
| 22:18.30 | ce_geek | prpplague: thx. And thanks for your help. |
| 22:18.37 | prpplague | ce_geek: if you have time tomorrow, i'd like to go over some of the items you did today |
| 22:19.03 | ce_geek | Excellent. There are some steps I'd like to thoroughly document that I'll probably want your feedback on too. |
| 22:19.27 | prpplague | i have a meeting at 1:30CST tomorrow, but the rest of the day should be ok |
| 22:19.35 | prpplague | heads home for some beer |
| 22:23.12 | g1powermac | GPSFan, your connectors will be dropped off at the post office tonight or tomorrow morning, finally was able to make the label |
| 22:39.48 | *** mode/#edev [+o landley] by ChanServ |
| 22:42.59 | GPSFan | g1powermac: thanks. I think that for the camera, you may need to drive the camera input lines through a LVX or similar device, but there are not too many of those. The video data 0-7 outputs may be good enough for the zipit as is. |
| 22:56.07 | landley | I'm having a horrible time with this whole "becoming unemployed" thing. |
| 22:56.12 | landley | I'm _trying_ to become unemployed. |
| 22:56.15 | landley | My consulting company won't let me. |
| 22:56.23 | landley | needs time to do open source projects. |
| 22:56.28 | landley | About a spare _year_, at this point. |
| 22:56.48 | landley | Instead they wave a really cool company at me within biking distance of home (in Austin) that needs, specifically, an embedded Linux expert. |
| 22:56.49 | landley | Right. |
| 22:56.56 | landley | Meeting 'em for lunch on monday. Won't think about it now. |
| 22:57.05 | landley | Must... Finish... OLS... Papaer... |
| 23:00.09 | *** join/#edev dijenerate (n=dijenera@65.48.152.24) |
| 23:01.27 | g1powermac | GPSFan, hmm |
| 23:01.53 | g1powermac | GPSFan, I doubt the logic high voltage on the zipit reaches 3.3v |
| 23:02.17 | g1powermac | it may all just work together except maybe one diode on the power |
| 23:02.30 | ds2 | g1powermac: you will run into problems in worse case situations |
| 23:02.44 | g1powermac | ds2, yea, I know, unfortunately |
| 23:02.45 | ds2 | threshold for CMOS is defined as 1/3 and 2/3 Vcc |
| 23:02.47 | GPSFan | g1powermac: been looking at the datasheet from sparkfun. it's nice as far as it goes, but it's missing the definations of all the i2c registers, |
| 23:03.00 | ds2 | plus you loose noise margin |
| 23:03.04 | g1powermac | meh, don't need i2c |
| 23:04.05 | GPSFan | g1powermac: oh yes you do, lots of control functions are set via the i2c. I don't know if it will just come up in rgb or yuv mode as the default with no command inputs. |
| 23:04.34 | g1powermac | hrmm, should be able to control it all over the data interface |
| 23:04.49 | GPSFan | the data interface is output only. |
| 23:05.23 | g1powermac | bah, crap, you're right. . . |
| 23:06.04 | GPSFan | looks for another datasheet |
| 23:06.26 | g1powermac | needs to actually sit down and read the datasheet instead of just looking at pinouts. . . |
| 23:10.07 | Dr{Who} | landley.... cloning.. man i wish it were possible.. btw.. sorry about that bug comment.. i just presumed it was in the main original branch.. i see how it could happen now. I ment i didnt understand how it could happen on the main branch as it would have been seen by now. |
| 23:11.44 | *** join/#edev linac (n=lin@121.41.33.27) |
| 23:13.36 | landley | drwho: I test my toys rather a lot, but I just haven't had time recently to play with it. |
| 23:13.42 | landley | tinycc is starved for attention. :( |
| 23:13.52 | g1powermac | GPSFan, there's definitely enough of the same datasheet out on the intarwebs |
| 23:14.09 | Dr{Who} | ya i am alomost ready to commit all my spare time to it.. if we can have a few people on it we can get er done. |
| 23:14.18 | Dr{Who} | i think from my pov.. this is a good project. |
| 23:14.27 | GPSFan | g1powermac: as far as I can find, they are all the same as what we have got. |
| 23:14.35 | g1powermac | yea |
| 23:14.37 | Dr{Who} | it may take a year but its worth it. |
| 23:15.37 | Dr{Who} | i did find another old c compiler that is active i think you also noted it on one of your posts |
| 23:16.03 | Dr{Who} | it was based upon the original c compiler as i recall. |
| 23:16.09 | g1powermac | GPSFan, I wonder if the sparkfun guys know more about the cameras |
| 23:16.20 | landley | drwho: pcc. |
| 23:16.27 | ds2 | Hmmm |
| 23:16.28 | landley | The openbsd and netbsd guys are turning it into something modern. |
| 23:16.34 | Dr{Who} | ya that sounds right. |
| 23:16.41 | SpeedEvil | g1powermac: is this the $19 sparkfun camera? |
| 23:16.47 | landley | http://lwn.net/Articles/255558/ |
| 23:16.55 | g1powermac | SpeedEvil, no |
| 23:16.56 | ds2 | hmmmmm landley and Dr{Who} are not the same person... |
| 23:17.00 | g1powermac | http://www.sparkfun.com/commerce/product_info.php?products_id=8667 |
| 23:17.10 | g1powermac | http://www.sparkfun.com/commerce/product_info.php?products_id=8668 |
| 23:17.14 | landley | See also http://lwn.net/Articles/269101/ |
| 23:17.18 | GPSFan | g1powermac: they say that they have a reliable source of parts, you'd think they would like to have good docs too. |
| 23:17.21 | landley | ds2: what? |
| 23:17.53 | SpeedEvil | g1powermac: one caveat with those - the readout rate has to be over xMHz |
| 23:17.53 | landley | ds2: I did mail DrWho my old hamme test kit when I got a nail board. Perhaps that confused you? :) |
| 23:17.54 | Dr{Who} | no landly is way cooler than i.... |
| 23:17.54 | ds2 | landley: for some reason I thought you just go by 2 different nicks as I have never seen both of you talk at the same time |
| 23:18.01 | landley | Nope. |
| 23:18.08 | SpeedEvil | g1powermac: you've got to read out the entire array at near video rates - typically. |
| 23:18.12 | g1powermac | GPSFan, yea, its kinda hard to use the camera with no clue what all those acronyms for the registers mean |
| 23:18.41 | g1powermac | SpeedEvil, hmm, well, I think the pxa270 in the zipit could handle it |
| 23:18.46 | GPSFan | g1powermac: the blurb says the devices have a "standard data+i2c" interface, maybe there is a defination of the reg values somewhere else. |
| 23:18.56 | g1powermac | GPSFan, maybe |
| 23:19.00 | SpeedEvil | g1powermac: can it do DMA in? |
| 23:19.18 | g1powermac | SpeedEvil, not sure, but it has its own dedicated camera interface |
| 23:19.19 | SpeedEvil | g1powermac: the data cannot be gotten over the I2C port BTW |
| 23:19.23 | SpeedEvil | Ah. |
| 23:21.26 | Dr{Who} | made a few mods to the configs and the makefile for bintools to make it build a cross version of objdump.. |
| 23:22.03 | Dr{Who} | landely. i see you mention several times your frustration in debugging. |
| 23:22.20 | Dr{Who} | i have debugging with gdbserver but with no symbols its a bit useless. |
| 23:22.29 | Dr{Who} | hence how i found the bug. |
| 23:23.38 | Dr{Who} | heh silly some symbols one finds.. ''been_there_done_that'' |
| 23:24.25 | roxfan | symbols are for wussies |
| 23:24.32 | *** join/#edev prpplague (n=dave123_@ppp-70-244-164-200.dsl.rcsntx.swbell.net) |
| 23:25.46 | Dr{Who} | ok same thing.. seems the segfault occures in .init every time.. |
| 23:26.36 | prpplague | Dr{Who} hammer board? |
| 23:27.04 | *** mode/#edev [+o prpplague] by ChanServ |
| 23:27.06 | landley | DrWho: I think I've already got cross objdump. :) |
| 23:27.16 | landley | Didn't have to modify anything... |
| 23:27.35 | prpplague | wonders if landley forgot about his dev notes |
| 23:27.56 | landley | prpplague: no, I've got a big penguicon write-up blocking it up. That's all. |
| 23:28.11 | prpplague | hehe |
| 23:28.16 | prpplague | np just being naggy |
| 23:28.39 | g1powermac | GPSFan, I'd ask the sparkfun people |
| 23:29.05 | prpplague | landley looks like the openedhand folks along with the openmoko folks are willing to push the s3c24xx stuff for qemu into the main tree |
| 23:29.21 | prpplague | landley should be able to add my hammer/nail stuff in too |
| 23:30.22 | prpplague | too much to do |
| 23:31.12 | Dr{Who} | ya |
| 23:31.15 | Dr{Who} | hammer board. |
| 23:31.38 | prpplague | Dr{Who} you using tinycc ? |
| 23:31.56 | landley | prpplague: join the club. :) |
| 23:32.10 | prpplague | landley only if its a nudie bar club |
| 23:32.53 | prpplague | welcome to the "Naked Penguin - strip club and linux development consulting" |
| 23:33.25 | prpplague | wow there is a post from jeff sutherland on LAK |
| 23:33.33 | prpplague | i haven't heard from him in years |
| 23:33.47 | g1powermac | prpplague, hmm, have you ever worked with these types of cameras like the ones on sparkfun: http://www.sparkfun.com/commerce/product_info.php?products_id=8667 |
| 23:34.06 | g1powermac | the datasheet for that sucker doesn't describe what all the registers do |
| 23:35.00 | prpplague | http://www.elinux.org/Sparkfun_Camera and http://www.elinux.org/Pixter_Camera |
| 23:35.14 | prpplague | i actually ordered some of those from sparkfun a couple days ago |
| 23:35.22 | prpplague | we want to test them with some 2440 boards |
| 23:35.24 | Dr{Who} | yes |
| 23:35.32 | g1powermac | hah, there we go |
| 23:35.41 | Dr{Who} | in working on a hello world built with tinycc |
| 23:35.54 | prpplague | Dr{Who} ahh lovely |
| 23:36.07 | landley | drwho: I think the problem is an unaligned access. |
| 23:36.07 | prpplague | Dr{Who} i hope you are going to doc it on the wiki |
| 23:36.20 | landley | I got it arm-tinycc to work on an arm host under qemu. |
| 23:36.22 | g1powermac | prpplague, wait a sec, is that the $19 one thats on a breakout board? |
| 23:36.46 | prpplague | g1powermac yea, similiar i/o though |
| 23:37.04 | prpplague | you can do YUV or RGB |
| 23:37.42 | Dr{Who} | prpplague.. im going to feed any work i do back to landley and keep the svn updated. |
| 23:37.48 | landley | drwho: but there's a lot of *(int *)ptr going on in tcc. ELF linking follows a linked list, for example. |
| 23:37.54 | landley | There's an svn? |
| 23:38.00 | g1powermac | prpplague, hmm, interesting |
| 23:38.04 | Dr{Who} | well your mercurial.. |
| 23:38.08 | landley | Yay mercurial. :) |
| 23:38.10 | prpplague | Dr{Who} but no docs on the wiki? |
| 23:38.25 | landley | My main objection to the old tcc mainline was it's still (2008) in CVS. |
| 23:38.27 | landley | That's just silly. |
| 23:38.29 | g1powermac | prpplague, so in theory this cheaper camera should work on the zipit with the info we have? |
| 23:38.51 | prpplague | uh, i'd have to look |
| 23:39.03 | prpplague | no sure if that one has all the data over i2c or what |
| 23:39.05 | prpplague | one sec |
| 23:39.11 | SpeedEvil | None send the data over I2C |
| 23:39.16 | Dr{Who} | I will add docs as soon as i have any : c ) so far its not worth posting a build procedure as landley has most of it on his blog. |
| 23:39.24 | g1powermac | prpplague, no the data is over the data bus |
| 23:39.26 | SpeedEvil | From the dozens of datasheets I've read. |
| 23:39.32 | g1powermac | I2C just does controls |
| 23:39.34 | SpeedEvil | yeah. |
| 23:40.07 | prpplague | g1powermac no not unless the zipitv2 has camera interface lines |
| 23:40.08 | landley | needs a week just to spend on tinycc... |
| 23:40.17 | g1powermac | prpplague, it does |
| 23:40.57 | prpplague | and they are brought out? |
| 23:41.03 | g1powermac | yup to the expansion port |
| 23:41.24 | prpplague | ahh well its a standard camera interface so yea it probably will work |
| 23:41.31 | g1powermac | neato |
| 23:41.47 | g1powermac | wonder if I should go for the bigger camer |
| 23:41.49 | g1powermac | *camera |
| 23:42.05 | CosmicPenguin | bigger is always better |
| 23:42.12 | g1powermac | got this one: http://www.sparkfun.com/commerce/product_info.php?products_id=8667 or this one http://www.sparkfun.com/commerce/product_info.php?products_id=8668 |
| 23:42.26 | SpeedEvil | Good luck. :) |
| 23:42.48 | prpplague | landley won;t you have it soon when you quit? |
| 23:43.01 | prpplague | landley or have you already got some head hunters after you? |
| 23:43.38 | g1powermac | the bigger camera I'll only get 15fps |
| 23:43.47 | g1powermac | but the smaller one I'll get 30fps |
| 23:44.03 | g1powermac | otoh, the bigger will do jpg |
| 23:44.26 | g1powermac | which is what I kinda want, I want to be able to do still images |
| 23:44.29 | prpplague | g1powermac now if you really want to stir up some stuff, add a camera to your zipitv2 and include the zebra image decoder |
| 23:44.59 | prpplague | g1powermac http://zebra.sourceforge.net/ |
| 23:45.24 | g1powermac | woah |
| 23:45.31 | prpplague | _that_ would shake things up |
| 23:45.34 | g1powermac | multipurpose |
| 23:45.35 | g1powermac | dang |
| 23:45.45 | g1powermac | that would be pretty freakin cool |
| 23:46.12 | prpplague | for good decodes you really need atleast 640x480 |
| 23:46.48 | g1powermac | so that 1.3 megapixel camera would be better |
| 23:46.51 | prpplague | http://www.nbc5i.com/news/15984283/detail.html |
| 23:47.01 | prpplague | about 2 blocks from my house |
| 23:47.11 | prpplague | we have like 20 helicopters flying around |
| 23:47.30 | g1powermac | ahh, lovely |
| 23:48.28 | g1powermac | alright, I might order me a camera. . . |
| 23:48.55 | prpplague | might i suggest you order atleast 2 and with 3 being optimal |
| 23:50.51 | prpplague | hmm whats the foot print for that |
| 23:52.37 | landley | prpplague: I have headhunters after me. |
| 23:54.11 | g1powermac | prpplague, why so many? |
| 23:55.19 | prpplague | g1powermac usually you want one to tinker with and a backup/reference |
| 23:55.31 | g1powermac | hmm, that is true |
| 23:56.01 | prpplague | i can't tell from the datasheets as the pitch of the pads |
| 23:56.23 | prpplague | gonna need something to solder it to |
