| 00:09.58 | *** join/#edev Sienna (n=Sienna@pool-71-246-77-216.bltmmd.east.verizon.net) |
| 00:36.14 | *** join/#edev prpplague (n=dave123_@ppp-70-244-160-94.dsl.rcsntx.swbell.net) |
| 00:39.56 | ce_geek | prpplague: What's the history with flyback? |
| 00:46.19 | *** join/#edev chouimat (n=dieu@r2351064.cidc.net) |
| 00:46.31 | *** join/#edev prpplague^2 (n=dave123_@ppp-70-244-160-94.dsl.rcsntx.swbell.net) |
| 00:47.08 | agaffney | ce_geek: does "CANUCK CANUCK CANUCK" mean anything to you? :P |
| 00:47.30 | _trine | he's ill |
| 00:47.42 | _trine | he means nothing by it |
| 00:47.57 | ce_geek | Haha ok |
| 00:48.21 | _trine | canook I think is a reference to Canadians |
| 00:48.23 | agaffney | yeah, but it's still annoying |
| 00:48.28 | agaffney | it i |
| 00:48.28 | agaffney | s |
| 00:49.09 | _trine | as 'they' say it takes all kinds to make up a world |
| 00:49.12 | SpeedEvil | He's ill, goes on excessively about his illnesses, and canadians. |
| 00:49.21 | ce_geek | haha |
| 00:49.39 | _trine | and there for the grace of God go I |
| 00:50.00 | SpeedEvil | There is a difference between being ill, and mentioning it every third line. |
| 01:23.23 | *** join/#edev prpplague (n=dave123_@ppp-70-244-160-94.dsl.rcsntx.swbell.net) |
| 01:26.26 | Sienna | prpplague, Tom_L in another channel tells me you are a Hammer dev? |
| 01:34.18 | SpeedEvil | He's at least the public face of the hammer maker on freenode. |
| 01:34.33 | *** join/#edev linac (n=lin@121.41.48.131) |
| 01:34.50 | Sienna | hello linac |
| 01:36.08 | linac | Sienna, hello |
| 01:46.31 | GPSFan | Sienna: prpplague has had some issues with his router and often (of late) pops in/out on a random basis when he is really not on line. he is usually on during normal (CDT) business hours. There are a bunch of us that have hammers and mignt be able to answer questions in his absense... ask away... |
| 01:46.48 | Sienna | I am looking for a custom carrier board |
| 01:47.06 | RobotGuy | Here you are too. :) |
| 01:47.14 | GPSFan | custom as in more thatn the TCT hammer carrier... |
| 01:47.40 | RobotGuy | It's a good board with a generous prototype area |
| 01:47.56 | Sienna | custom as in much smaller then that, with 2.5V to 3.3V, and 3.3V to 5V logic shifters |
| 01:48.03 | GPSFan | since the hammer is a 40 pin DIP, it's fairly easy to prototype something using wirewrap wire, |
| 01:48.08 | SpeedEvil | yeah |
| 01:48.11 | RobotGuy | There is going to be a Hammer Bot Board soon. :D |
| 01:48.21 | Sienna | RobotGuy, I want something the size of the Lynx boards :P |
| 01:48.30 | SpeedEvil | 2.5-3.3 input from a 2.5V device isn't an issue |
| 01:48.30 | Sienna | who is making the Hammer bot board!!! |
| 01:48.36 | RobotGuy | Sienna: That's what I am talking about - small like those. |
| 01:48.45 | RobotGuy | TinCanTool is making it. |
| 01:48.45 | GPSFan | for something more commercial several of the quick turn PCB houses could turn something around fairly quickly |
| 01:48.59 | SpeedEvil | 3.3-5V output also isn't a problem |
| 01:49.12 | Sienna | except that i don't know the first thing about logic or board design |
| 01:49.12 | SpeedEvil | Sienna: how many I's and O's were you thinking of? |
| 01:49.46 | Sienna | um, i am looking at SFE's 2.5V SPI triple axis accel, some 3.3V gyros, and 5V I2C devices |
| 01:49.58 | Sienna | and of course a 5V serial to a SSC32 |
| 01:50.15 | SpeedEvil | sfe's accel is analog out isn't it? |
| 01:50.31 | Sienna | they have one, $45, that reports over SPI |
| 01:50.33 | SpeedEvil | I don't recall which one they use |
| 01:50.34 | SpeedEvil | ah |
| 01:50.45 | GPSFan | Sienna: RobotGuy & SpeedEvil are the folks you want to talk to... |
| 01:51.00 | SpeedEvil | Sienna: look at the SPI thingy datasheet, and see if the inputs are 3.3V tollerant |
| 01:51.07 | RobotGuy | Hammer has two UARTS. |
| 01:51.16 | Sienna | they are... but all there ratings are for 2.5V |
| 01:51.19 | RobotGuy | 32 MB RAM/16 MB Flash |
| 01:51.26 | Sienna | so i didnt' want to stress if i didn't have too |
| 01:51.50 | SpeedEvil | Sienna: see if it says in the datasheet that you can A) run it on 3.3V, or b) when run on 2.5V that it can take 3.3V inputs |
| 01:51.51 | Sienna | wait, so can i not use the i2c, SPI, and uarts at teh same time> |
| 01:52.10 | SpeedEvil | Sienna: see the pinout diagram on the wiki |
| 01:52.23 | Sienna | SpeedEvil, yes, it says you can run the device at 3.3V. 3.6V is maximum rating |
| 01:52.39 | RobotGuy | I have tried to arrange stuff on the Wiki so it is easiest to find. |
| 01:52.40 | Sienna | its just you might have to adjust reported values if input != 2.5V |
| 01:53.01 | SpeedEvil | Sienna: so why not do that? |
| 01:53.09 | Sienna | i saw the pinout... and i could get to five different serial standards at once on it |
| 01:53.16 | SpeedEvil | http://www.elinux.org/Hammer_Board has pinouts |
| 01:54.03 | RobotGuy | I have been debating whether to add my WiFi stuff to the Wiki or not. |
| 01:54.13 | SpeedEvil | RobotGuy: go for it |
| 01:54.52 | RobotGuy | SpeedEvil: Well, it is not reliable. |
| 01:54.53 | Sienna | is that the 'wiki'? |
| 01:55.10 | SpeedEvil | Sienna: yes |
| 01:55.18 | RobotGuy | It;s the Documentation like from the main www.tincantools.com site |
| 01:55.18 | SpeedEvil | RobotGuy: so say that |
| 01:55.54 | RobotGuy | SpeedEvil: I could do that. Maybe I will add it, since I don't know when we will get a 2.6.24 kernel. |
| 01:55.57 | SpeedEvil | Sienna: you mean this one? |
| 01:56.07 | Sienna | is the hammer 5V input tolerant? (i.e., if I use a 5V SRF-08, will it work?) |
| 01:56.08 | SpeedEvil | http://www.sparkfun.com/commerce/product_info.php?products_id=8637 |
| 01:56.16 | SpeedEvil | Sienna: no |
| 01:56.55 | RobotGuy | Isnt't there a chip you can connect to the I/Os and then connect 5V stuff to? |
| 01:56.59 | Sienna | http://www.sparkfun.com/commerce/product_info.php?products_id=758 the one i was looking at |
| 01:57.03 | SpeedEvil | RobotGuy: yes |
| 01:57.12 | SpeedEvil | RobotGuy: hundreds |
| 01:57.49 | RobotGuy | I just want one part I can put between the 3.3V Hammer pins and the outside to connect to 5V stuff. |
| 01:58.11 | Sienna | that needs to be on the Bot board |
| 01:58.22 | Sienna | is there a planned release date for that? |
| 01:58.49 | RobotGuy | Sienna: March/April timeframe. It has not been released for beta yet |
| 01:59.22 | SpeedEvil | Sienna: from that page |
| 01:59.28 | SpeedEvil | ' |
| 01:59.28 | SpeedEvil | SKU#: SEN-00758 |
| 01:59.28 | SpeedEvil | Price: $43.95 |
| 01:59.28 | SpeedEvil | 10-99: $39.56 each (10% off) |
| 01:59.28 | SpeedEvil | 100 or more: $35.16 each (20% off) |
| 01:59.29 | RobotGuy | I am most anxious to get my hands on one. |
| 01:59.30 | SpeedEvil | Description: Breakout board for the 3 axis LIS3LV02DQ accelerometer from ST Micro. Please view the datasheet for this impressive sensor! With a digital interface (SPI or I2C), the user only has to initiate commands to get 1mg resolution! Independent channels can be shut off, the sensor can be powered down, even an interrupt threshold can be set and used to awaken the sensor from power down! Selectable ranges are available for +/-2g and 6g. This senso |
| 01:59.36 | SpeedEvil | We were thoroughly impressed with the flexibility and precision of this sensor. This is a great sensor to play with if you need triple axis accelerations in digital form and you don't want to deal with analog conversions or filtering. |
| 01:59.40 | SpeedEvil | Dimensions: 21x23mm (0.8x0.9") |
| 01:59.42 | SpeedEvil | Documents: |
| 01:59.44 | SpeedEvil | <PROTECTED> |
| 01:59.46 | SpeedEvil | <PROTECTED> |
| 01:59.48 | SpeedEvil | <PROTECTED> |
| 01:59.50 | SpeedEvil | <PROTECTED> |
| 01:59.52 | SpeedEvil | Interface: The LIS3LV02DQ is a 3.3V part and outputs SPI or I2C digital. ' |
| 01:59.53 | RobotGuy | I have the ADXL330 3-axis accelerometer breakout board |
| 01:59.54 | SpeedEvil | sie: |
| 01:59.56 | SpeedEvil | oh damn |
| 02:00.02 | SpeedEvil | how the hell wasn't that a 1 line paste |
| 02:00.05 | Sienna | thats good! mid april i get my next paycheck for robot parts ;P |
| 02:00.06 | RobotGuy | ~fishslap SpeedEvil |
| 02:00.07 | purl | ACTION slaps SpeedEvil up side the head with a wet fish. |
| 02:00.11 | RobotGuy | Hee hee |
| 02:01.05 | SpeedEvil | And why diddn't I get flood kicked. |
| 02:01.08 | SpeedEvil | sighs. |
| 02:01.16 | RobotGuy | I want a Hammer powered BRAT. :D |
| 02:01.28 | GPSFan | ~lart SpeedEvil for flooding |
| 02:01.28 | purl | steals SpeedEvil's mojo for flooding |
| 02:02.33 | Sienna | SpeedEvil: Yeah, looking at the datasheet, at 3.3V all of the tolerances seem to more then double over 2.5V |
| 02:02.36 | RobotGuy | Sienna: Have I told you lately that you ROCK? ;) |
| 02:03.22 | SpeedEvil | Sienna: see page 39 |
| 02:03.25 | SpeedEvil | of the datasheet |
| 02:03.35 | SpeedEvil | That's basically initial tolerances |
| 02:03.44 | SpeedEvil | you almost don't care about those |
| 02:03.53 | SpeedEvil | as you can calibrate it so easily |
| 02:04.41 | Sienna | i was looking at their tolerances for long term accuracy |
| 02:05.00 | Sienna | but yeah, those initial numbers to me also look scary at 3.3! |
| 02:05.24 | SpeedEvil | What are you trying to do? |
| 02:05.42 | SpeedEvil | You're not thinking you can do inertial navigation? Because you can't. |
| 02:05.51 | SpeedEvil | Well - for more than a few seconds. |
| 02:06.42 | Sienna | not entirely inertial nav, but balancing. No, if I wanted inertial nav, I would get a much more expensive IMU |
| 02:06.43 | SpeedEvil | Also - the graphs on p39 of the datasheet are from a whole bunch of devices |
| 02:06.55 | RobotGuy | A 5DOF IMU + another (I think) gyro or maybe it's an accelerometer |
| 02:07.13 | SpeedEvil | each one is one line within that spread |
| 02:07.35 | SpeedEvil | each part |
| 02:07.45 | Sienna | i understand |
| 02:07.45 | RobotGuy | Sienna: There is a 6DOF IMU that uses two Spark Fun components, but it is designed for Bioloid. |
| 02:07.51 | RobotGuy | $250.00 |
| 02:08.00 | SpeedEvil | But for balancing, 20mg offset is nothing |
| 02:08.12 | Sienna | And actually RobotGuy, i was hoping that the board I wanted desgined had all the .1" headers for a complete 6DOF |
| 02:08.16 | SpeedEvil | that's 1 degree or so out |
| 02:08.46 | Sienna | I have never done a balancer before, so i was being cautios |
| 02:08.58 | SpeedEvil | it's also irrelevant |
| 02:09.04 | Sienna | if 1 degree is normally acceptable, i guess thats fine |
| 02:09.14 | SpeedEvil | because you're not going to mount it (typically) within 1 degree of known position |
| 02:09.27 | SpeedEvil | you'll need to calibrate exactly where your balancers down is anyway |
| 02:09.39 | Sienna | actually, i am going to mount it exactly in the vertical center of the bot |
| 02:09.55 | SpeedEvil | Also - you calibrate the accellerometer by turning it over and over several times and looking at the max/min |
| 02:11.00 | Sienna | I guess my application is just a bit odder then most. I Have to balance on two point contacts, that don't move with respect to the ground |
| 02:11.59 | SpeedEvil | basically - you put the robot straight up, tell your software that it's straight up, and then calibrate that direction as down |
| 02:12.08 | Sienna | *nod* |
| 02:12.10 | SpeedEvil | so any offsets are irrelevant |
| 02:12.29 | SpeedEvil | the offsets are consistent - if you keep the same supply voltage, the offset stays the same |
| 02:13.45 | Sienna | then the only thing I am missing is a 3.3V to 5V conterter for the I2C bus |
| 02:14.05 | Sienna | and SFE doesn't make those :( |
| 02:14.10 | SpeedEvil | what's this for again? |
| 02:14.22 | Sienna | interfacing SRF-08 sonars to the Hammer |
| 02:15.28 | SpeedEvil | Oh - right |
| 02:15.30 | SpeedEvil | http://www.tau.ac.il/~stoledo/lego/i2c-8591-lis3l02as4/ |
| 02:15.35 | SpeedEvil | gives a hint. |
| 02:15.42 | SpeedEvil | You don't pullup the I2C lines to 5V |
| 02:15.46 | SpeedEvil | you pull them up to 3.3V |
| 02:15.49 | SpeedEvil | and it all just works |
| 02:16.12 | SpeedEvil | :) |
| 02:17.38 | RobotGuy | Sienna: Check out |
| 02:17.41 | RobotGuy | http://www.mono-project.com/Main_Page |
| 02:17.49 | RobotGuy | No, wait, not that |
| 02:18.17 | Sienna | SpeedEvil, but what happens when my 5V sensor sends a 5V reply back to the Hammer? (that link confused me) |
| 02:18.34 | RobotGuy | Sienna: These - http://www.sparkfun.com/commerce/product_info.php?products_id=741 and http://www.sparkfun.com/commerce/product_info.php?products_id=395 |
| 02:19.27 | RobotGuy | Sienna: That's what the intermediary chip does - translates the levels if I remember correctly. |
| 02:20.16 | RobotGuy | Is that right, SpeedEvil? |
| 02:20.31 | SpeedEvil | RobotGuy: sort-of-but-irrelevant-in-this-case. |
| 02:20.38 | SpeedEvil | I2C is open collector. |
| 02:20.48 | SpeedEvil | What this means is that the chips never drive the lines high. |
| 02:20.51 | Sienna | ok, thats 3.3V, but thats all analog measurement RobotGuy |
| 02:20.53 | SpeedEvil | They only drive it low |
| 02:20.58 | RobotGuy | That's why we need 4.7 K pull ups on the SCL and SDA lines. |
| 02:21.09 | SpeedEvil | The lines are pulled high by external resistors. |
| 02:21.33 | RobotGuy | I'm going to be using I2C a lot with Hammer and various things. |
| 02:21.42 | Sienna | so, instead of 4.7k, by using a higher ohm (82k in picture) it pulls it instead to 3.3 |
| 02:21.43 | Sienna | ? |
| 02:21.44 | SpeedEvil | If you connect these external resistors to 3.3V, not 5V, then all that happens is that when either chip 'lets go' of the line - doesn't pull it low - it springs back to 3.3V |
| 02:21.56 | RobotGuy | sneaks one of Sienna's I2C sensors away. :) |
| 02:22.05 | SpeedEvil | no - it's that it's connected to a 3.3V power supply. |
| 02:22.12 | SpeedEvil | The pullup that is. |
| 02:22.21 | Sienna | Ok, so power it from 5V, and pullup from 3.3V? |
| 02:22.33 | Sienna | and then they only drain it, so i am not going to kill the hammer? |
| 02:22.38 | SpeedEvil | The absolute value of the pullup isn't relevant to that case. |
| 02:22.51 | SpeedEvil | Yes, because nothing pulls the data lines higher than 3.3V |
| 02:22.55 | SpeedEvil | the I2C lines |
| 02:22.58 | Sienna | (and RobotGuy, if you wanna borrow one, i will let you, as long as you make me a nice C api that I can call :P) |
| 02:23.03 | SpeedEvil | as that's what the resistor is connected to. |
| 02:23.11 | Sienna | wonderful! *hugs SpeedEvil* |
| 02:23.14 | RobotGuy | Sienna: Can I do it for Linux? |
| 02:23.15 | SpeedEvil | :) |
| 02:23.20 | Sienna | this is getting easier by the second |
| 02:23.21 | SpeedEvil | However. |
| 02:23.31 | Sienna | RobotGuy: C! Of course linux |
| 02:23.40 | SpeedEvil | I'd put a shottkey diode on each input - just on general principles. |
| 02:23.41 | RobotGuy | Sienna: OK, you have a deal! |
| 02:23.55 | RobotGuy | One the I2C lines? |
| 02:23.56 | Sienna | go pm me your addy over on the LM forum |
| 02:23.57 | SpeedEvil | to stop the line going beyond +3.3V - in the case of wierdness |
| 02:24.02 | RobotGuy | s/One/On/ |
| 02:24.07 | SpeedEvil | yes |
| 02:24.23 | RobotGuy | Diodes are one way streets though. I2C is bi-directional. |
| 02:24.44 | SpeedEvil | no - the diode only stops it if the voltage on the I2C bus rises over 3.3V |
| 02:24.52 | SpeedEvil | (which it should never do) |
| 02:25.17 | RobotGuy | SpeedEvil: That still doesn't solve the bi-directional situation though. |
| 02:25.24 | SpeedEvil | It does |
| 02:25.34 | SpeedEvil | however, the diodes are irrelevant to that. |
| 02:25.39 | RobotGuy | Explain please. |
| 02:25.49 | SpeedEvil | I2C is an open collector bus. |
| 02:25.52 | *** join/#edev cajaka (n=none@cpe-24-58-193-200.twcny.res.rr.com) |
| 02:25.52 | RobotGuy | Right |
| 02:26.03 | SpeedEvil | No IC pulls the bus up. |
| 02:26.05 | SpeedEvil | Only down. |
| 02:26.17 | SpeedEvil | It's pulled up by external resistors hooked to an external voltage. |
| 02:26.23 | SpeedEvil | this voltage sets the bus voltage |
| 02:26.29 | RobotGuy | Right, the 4.7K pull ups. |
| 02:26.45 | SpeedEvil | If it's 3.3V, then the signal will swing from about 0V-3.3V |
| 02:26.58 | RobotGuy | The external voltage should be the same as the Hammer's VL. |
| 02:27.16 | RobotGuy | How does that interface with 5V parts though?? |
| 02:27.19 | SpeedEvil | this works for a 5V device, as the 3.3V is considerably above the 2V or so threshold that the 5V chip will be using to pick 0s from 1s. |
| 02:27.26 | RobotGuy | Most of the sensors we use are 5V. |
| 02:27.42 | Sienna | RobotGuy, i understand him! comeon :P |
| 02:27.55 | SpeedEvil | So the signal never gets pulled higher than 3.3V - the hammer is happy and no smoke comes out. |
| 02:28.00 | RobotGuy | Sienna: OK, but I have to see it work. |
| 02:28.03 | Sienna | power the SRF08 from 5V, but tie the SDA/SDC lines to 3.3V |
| 02:28.16 | Sienna | hey, it would be a good experiment! |
| 02:28.18 | SpeedEvil | And the signal of 3.3V is quite adequate for the 5V part to understand it as a '1'. |
| 02:28.23 | RobotGuy | I just bought five 3.3V and five 5V regulators. |
| 02:28.28 | Sienna | now, how to tie a 5V Sharp ranger to a hammer? |
| 02:28.54 | RobotGuy | This would require the level shifting chip, I think. |
| 02:28.57 | SpeedEvil | However. I'd connect 100R resistors in series with the hammers SDA/SDC lines anyway - just in case. As it can't hurt. |
| 02:29.00 | RobotGuy | As would all 5V sensors. |
| 02:29.03 | SpeedEvil | Datasheet si? |
| 02:29.23 | RobotGuy | 100R? |
| 02:29.27 | SpeedEvil | 100 ohm |
| 02:29.29 | RobotGuy | OK |
| 02:29.39 | Sienna | http://www.junun.org/MarkIII/datasheets/GP2D12_15.pdf |
| 02:29.52 | Sienna | the 100ohm are for what? |
| 02:30.20 | RobotGuy | Sorry for being a bit dense tonight. I have not been feeling well - head pains. |
| 02:30.56 | Sienna | *hugs* |
| 02:31.09 | RobotGuy | Awww, thanks, Sienna. :) <hugs back> |
| 02:32.02 | SpeedEvil | Sienna: you just want an analog distance? |
| 02:32.23 | RobotGuy | I've been using http://www.lynxmotion.com/Product.aspx?productID=260&CategoryID=8 but am not real happy with them. |
| 02:32.29 | SpeedEvil | Sienna: they are oops resistances. They stop large curents flowing in the events of a fault. |
| 02:33.27 | Sienna | Ok SpeedEvil |
| 02:33.34 | SpeedEvil | Sienna: I'd simply - for the analog input - connect in series with a 1K resistor and hook up to an ADC - if you just want two |
| 02:33.37 | Sienna | RobotGuy, thats the same sensor i linked |
| 02:33.41 | RobotGuy | So what is the chip we need to make Hammer I/Os 5V tolerant? |
| 02:33.45 | Sienna | and SpeedEvil, yes, just analog. |
| 02:34.16 | Sienna | unless someone builds a pic to take a whole ring of 5V devices and turn it into a 3.3V SPI bus ;P |
| 02:34.28 | RobotGuy | Hammer only has 2 ADCs, but I am working to program some PICs to give more ADCs and other stuff. |
| 02:36.02 | Sienna | thats why i was looking at the SPI accel |
| 02:36.06 | Sienna | doesnt use an ADC |
| 02:36.09 | Sienna | or require a mux |
| 02:36.22 | RobotGuy | I will have firmware for dsPIC4012's to give 5 more ADCs, 1 quadrature encoder module, and maybe even 6Ch motor control PWM. These could interface using I2C or SPI, but I have never worked with SPI. |
| 02:36.49 | Sienna | i looked at the spec on SPI... up to 70MHz!!! |
| 02:37.05 | RobotGuy | Sienna: Have you worked with SPI? |
| 02:37.18 | RobotGuy | Hammer has two SPI modules. |
| 02:37.32 | Sienna | nope, i haven't even worked i2c |
| 02:37.32 | RobotGuy | One goes to the SD card slot. |
| 02:37.41 | Sienna | i can do mechanical, and software |
| 02:37.41 | RobotGuy | I have worked quite a bit with I2C. |
| 02:37.49 | Sienna | that stuff in the middle..... not my forte |
| 02:38.14 | RobotGuy | That must make me the "Interfacing Guy" then. ;) |
| 02:38.37 | RobotGuy | Sienna: Do you work with PICs/dsPICs? |
| 02:39.07 | RobotGuy | Sienna: You can do my mechanical and I will do your interfacing. :) |
| 02:39.55 | RobotGuy | I had a heck of a time assembling my BRAT with the 3mm hardware, but I got it done and think I am OK with it now. |
| 02:40.52 | Sienna | thats a fair trade for me! |
| 02:41.28 | RobotGuy | :D |
| 02:41.49 | RobotGuy | Sienna: Check out http://www.trossenrobotics.com/ - their forums are very active and lots of excellent info. |
| 02:42.33 | Sienna | ok, i will join their forums too |
| 02:42.59 | RobotGuy | You will find me there also. :D |
| 02:44.24 | Sienna | ok, registered |
| 02:44.31 | RobotGuy | YAY! |
| 03:01.41 | *** join/#edev prpplague (n=dave123_@ppp-70-244-160-94.dsl.rcsntx.swbell.net) |
| 03:02.49 | prpplague | ~seen landley |
| 03:02.54 | purl | landley <n=landley@dsl093-216-022.aus1.dsl.speakeasy.net> was last seen on IRC in channel #edev, 2d 4h 30m 15s ago, saying: 'Could be.'. |
| 03:04.25 | SpeedEvil | wwwasic.kip.uni-heidelberg.de/lhcb/Publications/external/AN97055.pdf Sienna RobotGuy is worth reading prolly |
| 03:05.04 | Sienna | yeah, i downloaded that for reading later |
| 03:05.31 | RobotGuy | Ahh, thanks for that, SpeedEvil. |
| 03:05.46 | RobotGuy | OK, I gotta get going for a bit. I want to to a visiting. :D |
| 03:07.06 | Sienna | night! |
| 03:55.27 | *** join/#edev Redhatter (n=vk4fsjl@2001:388:f000:0:0:0:0:279) |
| 04:19.02 | *** join/#edev Onyxyte (i=Onyxyte@r74-192-234-77.lfkncmta01.lfkntx.tl.dh.suddenlink.net) |
| 04:46.25 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 05:02.11 | *** join/#edev DrZimmerman (n=theo@tbaumgartner.broker.freenet6.net) |
| 05:16.41 | *** join/#edev redblue (i=star@ppp033.108-253-207.mtl.mt.videotron.ca) |
| 05:45.58 | *** join/#edev harpal (n=harpal@124.125.76.195) |
| 06:27.33 | *** join/#edev steliosk (n=Stelios@athedsl-105864.home.otenet.gr) |
| 06:28.28 | *** join/#edev harpal_ (n=harpal@124.125.76.195) |
| 06:41.20 | Mirell | Wee, I'm 23 now. |
| 08:03.47 | *** join/#edev howtoo (n=howtoo@195.159.62.66) |
| 08:24.23 | *** join/#edev roxfan (n=dunno@164.14-200-80.adsl-dyn.isp.belgacom.be) |
| 09:43.31 | *** join/#edev lyakh (n=lyakh@p57BD281E.dip0.t-ipconnect.de) |
| 10:12.41 | *** join/#edev Onyxyte (i=Onyxyte@r74-192-234-77.lfkncmta01.lfkntx.tl.dh.suddenlink.net) |
| 11:02.34 | *** join/#edev redblue (i=star@ppp049.108-253-207.mtl.mt.videotron.ca) |
| 12:20.56 | agaffney | Mirell: I hit that particular milestone ~13 months ago :P |
| 12:35.54 | chouimat | wow me it was ~11 years ago :( |
| 12:56.10 | *** join/#edev MothaWork_ (n=monmotha@mail.forwardengineering.com) |
| 13:11.49 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 13:18.45 | *** join/#edev GPSFan (n=kenm@12.10.255.246) |
| 13:18.45 | *** mode/#edev [+o GPSFan] by ChanServ |
| 13:19.11 | prpplague | ~lart GPSFan |
| 13:19.11 | purl | steals GPSFan's mojo |
| 13:19.36 | GPSFan | prpplague: g'day |
| 13:19.43 | prpplague | GPSFan: whats cookin? |
| 13:20.01 | GPSFan | grabs another quart of mojo from the mojo locker. |
| 13:20.11 | prpplague | hehe |
| 13:20.18 | GPSFan | spring is here... |
| 13:20.42 | prpplague | GPSFan: never met Spring... is she good looking? |
| 13:20.47 | GPSFan | prpplague: did you say something about a GSPI libertas driver in 2,6.25? |
| 13:21.14 | prpplague | GPSFan: yea, the core portion is there, a small section for uploading the firmware needs to be added |
| 13:21.33 | prpplague | GPSFan: iirc there are several people working on a patch for 2.6.26 |
| 13:21.44 | GPSFan | ah, I looked and all I found was a reference to the gspi helper.bin |
| 13:22.01 | prpplague | GPSFan: yea, its not 100% complete |
| 13:22.16 | GPSFan | but progress. |
| 13:22.26 | prpplague | GPSFan: sorry, in my line of work, i rarely expect that 100% of what i need is in the main stream kernel |
| 13:22.56 | GPSFan | understood. |
| 13:22.58 | prpplague | GPSFan: it is almost a given that i expect on 50% of what i need to be in the main tree |
| 13:23.16 | prpplague | sometimes i forget that most people aren't used to patching things together |
| 13:23.45 | prpplague | GPSFan: the firmware you grabbed from the zipitv2 is the gspi firmware for the 8686 |
| 13:24.07 | prpplague | GPSFan: i've still not found any 8385 gspi firmware that is openly available |
| 13:24.32 | GPSFan | prpplague: yes, but only available as source under NDA |
| 13:24.48 | prpplague | GPSFan: but you don't need the source to the firmware |
| 13:24.55 | prpplague | GPSFan: just need the binaries |
| 13:25.13 | GPSFan | corrrect, I meant the source to the gspi driver |
| 13:26.00 | prpplague | GPSFan: ahh, yea, but the document for implementing the gspi interface was accidentally posted on the OLPC mailing list, so implementing the interface is easy |
| 13:26.17 | prpplague | GPSFan: the rest of the code is exactly the same as the CF, usb , and sdio |
| 13:26.28 | GPSFan | easy, but not complete as of yet. |
| 13:26.35 | prpplague | GPSFan: correct |
| 13:26.56 | prpplague | GPSFan: i expect this time next month it will be functional |
| 13:27.07 | prpplague | GPSFan: i actually have some xg-182 modules on order |
| 13:27.34 | GPSFan | are they shipping yet? or jusrt the xg-180? |
| 13:27.48 | prpplague | GPSFan: i wish i could have talked management into doing a module for the wi2wi instead of doing it onboard |
| 13:28.00 | prpplague | GPSFan: yea, they've been shipping since november 2007 |
| 13:29.12 | prpplague | http://www.zcomax.com/embedded/XG-182M/dsxg182mv4.pdf |
| 13:29.21 | GPSFan | how bad is that connector to work with? it's .5mm IIRC. |
| 13:29.24 | *** join/#edev linac (n=lin@121.41.48.184) |
| 13:29.34 | prpplague | GPSFan: bad |
| 13:29.55 | prpplague | GPSFan: a real pain in the ass |
| 13:29.58 | GPSFan | it's not like they need all those pins. |
| 13:30.42 | prpplague | GPSFan: yea, the 8686 supports CF, sdio and gspi, but it also has a number of control lines for suspend modes, and a series of gpio status pins |
| 13:31.12 | prpplague | GPSFan: plus they want to make sure its backward compatible to the 8385 based modules |
| 13:31.50 | prpplague | GPSFan: zcomax doesn't appear to be the actual manufacturer of the modules, since the same model is available from a number of vendors |
| 13:32.08 | GPSFan | ah |
| 13:32.18 | prpplague | GPSFan: i suspect that marvell has some authorized manufacturer that then resales the modules to larger distributors |
| 13:32.27 | GPSFan | do you have the fccid? |
| 13:33.03 | prpplague | yea |
| 13:33.44 | GPSFan | the fcc docs might have a clue to who makes it. |
| 13:34.40 | prpplague | GPSFan: yea, just never dug into it |
| 13:34.57 | prpplague | GPSFan: we have a long term contract with zcomax so we just purchase from them |
| 13:35.03 | prpplague | GPSFan: we have terms and they stock for us |
| 13:35.15 | GPSFan | no matter, as long as you can get them for the life of your project. |
| 13:35.26 | prpplague | GPSFan: yea |
| 13:35.37 | prpplague | GPSFan: we have an 8 year guarantee |
| 13:36.20 | prpplague | GPSFan: basically they have to make it available in the qty and price or else they have substite at their expense |
| 13:37.09 | GPSFan | and you get to write new drivers. but you have NDA source so it's only a matter of doing it (if neccessary) |
| 13:37.45 | *** join/#edev cajaka (n=none@cpe-24-58-193-200.twcny.res.rr.com) |
| 13:43.01 | prpplague | GPSFan: yea, i made sure that i was free and clear on OSS dev |
| 13:43.53 | prpplague | GPSFan: but that kinda the thing, doing GSPI section is just a small addition to the libertas driver, the upper level code for libertas is exactly same across the interface types |
| 13:44.31 | GPSFan | that's a "good thing" (tm) |
| 13:45.18 | prpplague | GPSFan: indeed |
| 13:46.47 | *** join/#edev chouimat|work (n=dieu@209.217.106.98) |
| 13:52.26 | *** join/#edev Onyxyte (i=Onyxyte@r74-192-234-77.lfkncmta01.lfkntx.tl.dh.suddenlink.net) |
| 13:54.41 | *** join/#edev ce_geek (n=ce_geek@roger-blanchette-4.umelst.maine.edu) |
| 14:01.35 | *** join/#edev CosmicPenguin (n=nobody@163.181.251.103) |
| 14:01.35 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 14:12.52 | *** join/#edev sjhill (n=sjhill@real.realitydiluted.com) |
| 14:12.52 | *** mode/#edev [+o sjhill] by ChanServ |
| 14:33.56 | *** join/#edev steliosk (n=Stelios@athedsl-25485.home.otenet.gr) |
| 14:53.21 | *** join/#edev redblue (i=star@ppp166.108-253-207.mtl.mt.videotron.ca) |
| 14:56.00 | *** join/#edev linac (n=lin@121.41.45.171) |
| 15:01.29 | prpplague | sjhill: you find your flash problem? |
| 15:01.34 | prpplague | CosmicPenguin: greetings |
| 15:01.39 | CosmicPenguin | yo |
| 15:01.49 | prpplague | CosmicPenguin: whats cookin? |
| 15:01.55 | CosmicPenguin | same ol |
| 15:01.59 | chouimat|work | morning |
| 15:02.21 | prpplague | chouimat|work: hey frenchy, whats up in your neck of the woods? |
| 15:02.27 | CosmicPenguin | They are taking away my Unix mail server, which is causing me some sorrow right now |
| 15:02.52 | prpplague | doh |
| 15:03.12 | CosmicPenguin | Mainly because it has a static password, and the one on the exchange server changes every 90 days, so I'll need to go in and change half a dozen passwords |
| 15:03.15 | chouimat|work | prpplague: nothing much ... got gentoo, fedora working on my avr32 boards ... (was bored) and now reading about statfs,statvfs and ustat |
| 15:03.26 | CosmicPenguin | and of course, I'll forget - and everything will stop working every three months like clockwork |
| 15:03.37 | chouimat|work | ocuh |
| 15:03.57 | CosmicPenguin | Looking forward to that |
| 15:04.49 | prpplague | hehe |
| 15:05.20 | prpplague | hmm |
| 15:05.33 | prpplague | i wonder who's perlbot no_body belongs to |
| 15:06.09 | CosmicPenguin | kick ban it |
| 15:06.15 | prpplague | no_body: help |
| 15:06.19 | CosmicPenguin | unapproved bots suck |
| 15:06.38 | *** mode/#edev [+o prpplague] by ChanServ |
| 15:07.30 | prpplague | no_body: listplugins |
| 15:08.15 | prpplague | no_body: help |
| 15:09.46 | prpplague | hmm, doesn't identify itself |
| 15:09.52 | prpplague | it has an ipv6 address |
| 15:09.57 | sjhill | prpplague: yeah |
| 15:10.04 | prpplague | sjhill: what was it? |
| 15:10.05 | sjhill | prpplague: i'll never trust someone's code again |
| 15:10.09 | prpplague | sjhill: hehe |
| 15:10.19 | prpplague | sjhill: end point calculation error? |
| 15:10.24 | sjhill | prpplague: a WP bit that was set and should not have been |
| 15:10.32 | prpplague | sjhill: ahh |
| 15:10.47 | *** kick/#edev [no_body!n=dave@mail.americanmicrosystems.com] by prpplague (prpplague) |
| 15:11.47 | sjhill | prpplague: i was assured by another engineer that it was not honored, gee, guess what, it is |
| 15:12.05 | prpplague | hehe |
| 15:12.14 | prpplague | sjhill: famous last words |
| 15:12.17 | sjhill | so i blew 7 hours yesterday |
| 15:12.36 | sjhill | sure, i'm glad i found the bug, but it shouldn't have been there and i should not have trusted him |
| 15:13.08 | sjhill | and because of that, i have to re-release my kernel image and build system...again |
| 15:13.22 | prpplague | doh, hate it when that happens |
| 15:13.36 | prpplague | sjhill: i'm struggling to bring up a new set of boards for a handheld |
| 15:13.43 | prpplague | sjhill: the EE did a terrible job |
| 15:14.04 | prpplague | sjhill: its gonna take 10 times as long to bring it up because of all the crap he did |
| 15:14.09 | prpplague | very disappointing |
| 15:14.17 | sjhill | that is crappy |
| 15:14.36 | prpplague | sjhill: very hard to be motivated about working on hardware you know is a complete and utter disaster |
| 15:15.40 | prpplague | if someone owns the perlbot no_body, it can rejoin #edev if you add a help feature to identify the owner of the bot |
| 15:15.56 | prpplague | for example the way purl does |
| 15:15.57 | prpplague | ~owner |
| 15:15.58 | purl | TimRiker is my owner |
| 15:16.18 | chouimat|work | prpplague: ouch |
| 15:16.20 | CosmicPenguin | ban it too |
| 15:16.34 | prpplague | CosmicPenguin: why ban purl? |
| 15:16.40 | CosmicPenguin | no - no_body |
| 15:16.46 | CosmicPenguin | you just kicked it... |
| 15:17.13 | prpplague | oh yea, i figured if i kick'd it i'd get the attention of the owner |
| 15:18.07 | prpplague | sjhill: real disappointed with the EE's work |
| 15:18.14 | prpplague | sjhill: problem is management loves him |
| 15:18.53 | sjhill | ah, tenure for the incompotent |
| 15:18.56 | sjhill | love it |
| 15:19.11 | prpplague | sjhill: i can't seem to get management to understand that the EE is basically screwing the rest of the engineering department because of his incompotency |
| 15:19.26 | prpplague | sjhill: thats just it, he's only been with the company 9 months |
| 15:20.56 | prpplague | anyone ever read that post on the dailywtf about the programmer that spends months working on something and all they had was something like 5 lines of code that printed "i am a freakin genius!" |
| 15:21.17 | prpplague | well thats this EE |
| 15:22.47 | chouimat|work | prpplague: cool :D |
| 15:24.29 | prpplague | hehe not so cool when he hands you a board and says "here i'm done, you make it work!" |
| 15:25.03 | prpplague | chouimat|work: his first rev of the board had no jtag and no serial port |
| 15:25.17 | prpplague | i'm like, sooooo, how do you expect to program this thing? |
| 15:25.21 | chouimat|work | prpplague: wow I'm impressed |
| 15:25.33 | prpplague | and he's like: i don't know thats your job to do programming |
| 15:25.57 | adyer | prpplague: don't you guys have design reviews? |
| 15:26.29 | prpplague | adyer: we used to, but some reason right after we hired this guy, the reviews stopped |
| 15:26.38 | adyer | prpplague: lol |
| 15:26.53 | chouimat|work | prpplague: at least he's not a fan of flashing a 2GB flash using xmodem over 9600 :) |
| 15:27.11 | prpplague | adyer: we have regular engineering meetings, but management seems to be completely deaf to the complaints from the other engineers |
| 15:27.19 | prpplague | chouimat|work: hehe |
| 15:27.40 | *** part/#edev Onyxyte (i=Onyxyte@r74-192-234-77.lfkncmta01.lfkntx.tl.dh.suddenlink.net) |
| 15:28.17 | chouimat|work | prpplague: or implementing a parser in the device firmware because the guy want to use the CANopen ascii mapping |
| 15:28.31 | *** join/#edev T0mW (n=Tom@24.115.218.135.res-cmts.sth.ptd.net) |
| 15:29.16 | prpplague | adyer: my favorite was when i was explaining to management about the missing serial port for debugging they had this epiphany and it was "wow i have this great idea! we should always put a serial port on the board somewhere for us to do debugging!" |
| 15:29.50 | chouimat|work | bangs his head on the desk |
| 15:30.30 | prpplague | it was like that fedex/kinkos commercial where the guy makes the suggest to use their services, and the boss dismisses it, then the boss says the same thing be moves his hands around when he says it |
| 15:30.46 | prpplague | s/be/but |
| 15:30.56 | adyer | prpplague: it's not like a 3-pin header with tx/rx/gnd is gonna kill you :-) |
| 15:31.05 | prpplague | adyer: indead |
| 15:31.08 | prpplague | indeed |
| 15:31.12 | prpplague | can't spell today |
| 15:31.20 | adyer | prpplague: assuming you have a spare port :-) |
| 15:31.29 | prpplague | adyer: three unused |
| 15:32.58 | adyer | prpplague: well, after you finish programming the thing by osmosis, you can alsways spit out debug stuff over the usb <ducks> |
| 15:33.22 | prpplague | adyer: yea, i have no problems setting up the USB |
| 15:33.58 | prpplague | adyer: its the fact that we told management and the EE about the jtag and serial port, it they totally ignored us |
| 15:34.17 | prpplague | adyer: $12k of prototypes in the trash |
| 15:34.23 | prpplague | adyer: i don't think thats good business |
| 15:35.17 | adyer | prpplague: nope, maybe you need a consultant to tell management what you already know (we are availabel for a small fee :-) |
| 15:35.37 | prpplague | adyer: hehe, yea well i don't think that will happen |
| 15:35.50 | chouimat|work | adyer: some have a really nice definition of "small" ;) |
| 15:35.59 | prpplague | adyer: my fear is that managment is going to eventually get unhappy with the dev cycle |
| 15:36.15 | prpplague | adyer: they are either going to realize that the EE is the problem and fire him or... |
| 15:36.29 | chouimat|work | prpplague: make hime a manager :) |
| 15:36.36 | prpplague | adyer: they aren't going to bother to figure out what the problem is and just fire the entire engineering department |
| 15:37.43 | prpplague | chouimat|work: yea, thats a political mindfield |
| 15:37.53 | adyer | prpplague: write down your concerns, and keep the emails to point to at a convenient later date, frame it as "we need to improve this, here's how" |
| 15:38.46 | prpplague | chouimat|work: if he hires outside the company, thats gonna piss alot people off, and if he promotes from within the company, its going to be a problem because there is a conflict between the senor most engineer and the person who is really qualified to do the job |
| 15:40.06 | chouimat|work | prpplague: I like those minefield </sarcasm> |
| 15:41.41 | prpplague | hehe |
| 15:42.18 | prpplague | adyer: i'm beyond that, i've already tried making things better, now days i just keep to myself and do what i'm told |
| 15:43.07 | chouimat|work | prpplague: and keep your resume up to date ? |
| 15:43.16 | prpplague | indeed |
| 15:46.14 | T0mW | prpplague: yeah, well, I seem to be hitting the same problem as you. Customer sends me a document which loosely specifies what his new unit will be. Then, sends me email to "mention" things that he forgot. |
| 15:46.46 | T0mW | "Oh, btw, we are going to use USB to connect to the thermal printer internally"... |
| 15:46.50 | prpplague | T0mW: i'm used to that kind of thing, its dealing with a clueless EE thats creating my nightmare |
| 15:47.17 | T0mW | prpplague: and, I have a clueless software type that is causing me grief. |
| 15:48.35 | T0mW | He figures that we can store several hundred megabytes of video image in an SD card, then pump them across ethernet to the front panel unit whenever he needs them. |
| 15:48.48 | chouimat|work | T0mW: prpplague hopefully this is over for me |
| 15:48.48 | T0mW | too bad the SD card runs at 12.5MHz |
| 15:49.12 | prpplague | T0mW: done much with the hammer? |
| 15:49.16 | T0mW | prpplague: feature creep |
| 15:49.32 | T0mW | prpplague: nothing, been tied up with other crap |
| 15:49.50 | T0mW | prpplague: plus shopping for a house |
| 15:49.55 | prpplague | ahh |
| 15:51.13 | T0mW | prpplague: if you want to diddle with the design, I can ship it down to you? |
| 15:52.21 | prpplague | naw, between what i have to do for AML and tinkering for tincantools, i'm completely burned out, plus right now i'm totally preocuppied by the office politics craps |
| 15:54.53 | ce_geek | prpplague: Gosh! Joy and fun all over, eh? |
| 15:55.03 | prpplague | ce_geek: indeed |
| 15:55.29 | prpplague | ce_geek: just a note, the stuff going on at AML has no bearing on tincantools, they are sister companies, but independent |
| 15:55.35 | ce_geek | prpplague: What is your background, by the way? Just curious. I thought you were an EE / CE. |
| 15:55.50 | ce_geek | prpplague: Haha, good to know. |
| 15:57.33 | prpplague | ce_geek: i generally refer to myself as an "intellectual prostitute" |
| 15:57.41 | prpplague | ce_geek: i solve problems |
| 15:57.49 | ce_geek | prpplague; HAHA. |
| 15:58.02 | ce_geek | prpplague; Excellent. |
| 15:58.38 | prpplague | ce_geek: to be honest i never finished college, i ran out of money and joined the army |
| 15:58.57 | ce_geek | prpplague; I've learned here at least in academia that often times the piece of paper doesn't mean shit. Some of the best educators I've had have only had a Master's, but the college refuses to hire anyone with less than a PhD nowadays, and everyone they hire is outrageously useless. |
| 15:58.59 | prpplague | ce_geek: i've been doing computer design since back in the alter days |
| 15:59.16 | prpplague | ce_geek: indeed |
| 15:59.18 | ce_geek | prpplague; Good! Well, whatever you did, you definitely know your shit. |
| 15:59.32 | prpplague | ce_geek: hehe, naw, i just am good at fooling people |
| 15:59.44 | ce_geek | Well, that's an art too ;-) |
| 15:59.51 | prpplague | indeed |
| 16:00.00 | ce_geek | spreads it thick on his thesis. |
| 16:00.30 | adyer | heh the art is in knowing what you what you don't know |
| 16:00.45 | prpplague | ce_geek: i served in the army signal corp and that got my foot in the door for alot of places. lots of on the job experience |
| 16:00.51 | prpplague | adyer: exactly |
| 16:00.56 | ce_geek | prpplague: Slick |
| 16:01.32 | prpplague | ce_geek: my speciality is problem solving. i'm used to being handed a stack of datasheets and being told "figure out how to make this work" |
| 16:02.31 | prpplague | ce_geek: we are often paid to clean up other contractors messes |
| 16:02.50 | *** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net) |
| 16:02.52 | ce_geek | prpplague; There's an alarming amount of that kind of stuff going on all over the place, and it's damn expensive. |
| 16:02.59 | prpplague | ce_geek: i.e. the contractor is given 6 months to do a job, and after the 6 months it still isn't working, so they bring it to us |
| 16:03.08 | ce_geek | prpplague: Oh excellent. |
| 16:03.30 | prpplague | ce_geek: but when they bring it to us, we will only get 45 days to make it work, hence we charge a whole lot to clean up the mess |
| 16:03.50 | ce_geek | prpplague: As you should :) |
| 16:04.28 | prpplague | problem is that our EE retired back in march of last year, and we hired a new one that just isn't cut out for this type of work |
| 16:13.12 | *** join/#edev ico2 (n=ico2@modem-2641.cougar.dialup.pol.co.uk) |
| 16:15.10 | T0mW | prpplague: that's it. I just sent email to the customer and told them that I was not going to do their design unless they gave me a definative spec. They keep changing sh!t all the time! Either we are going with an S3C2440 @ 400MHz + 64Meg SDRAM or an Intel XScale IOP348 @ 1200MHz + 256Meg DDR2... |
| 16:20.38 | chouimat|work | T0mW: ouch |
| 16:20.52 | T0mW | I'm fed up with them |
| 16:50.49 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 16:56.23 | *** join/#edev lyakh (n=lyakh@p57BD281E.dip0.t-ipconnect.de) |
| 16:58.17 | RobotGuy | I think hard coded paths are kicking my backside with buildroot. |
| 17:03.32 | *** join/#edev ico2 (n=ico2@modem-2641.cougar.dialup.pol.co.uk) |
| 17:07.21 | RobotGuy | Has anyone told the Debian folks their site sucks with respect to finding the package a person needs? |
| 17:08.10 | CosmicPenguin | really? I find packages.debian.org (and packages.ubuntu.com) to be very well organized |
| 17:08.30 | RobotGuy | Hmmm, maybe I am using the wrong thing then. |
| 17:09.27 | RobotGuy | It would be so much easier if they did not split the devel stuff from the libraries. |
| 17:09.43 | RobotGuy | I usually resort to just building from sources |
| 17:12.00 | RobotGuy | It would be nicer if they had a search engine to find packages. |
| 17:14.39 | prpplague | returns from lunch |
| 17:14.42 | prpplague | RobotGuy: greetings |
| 17:14.46 | RobotGuy | I found what I needed. |
| 17:14.56 | RobotGuy | prpplague: Greetings |
| 17:15.07 | prpplague | oh oh cute |
| 17:15.14 | RobotGuy | What? |
| 17:15.18 | prpplague | i'm getting spam from leonidas@sparta.com |
| 17:15.47 | prpplague | how funny is that |
| 17:15.59 | RobotGuy | Facinating.. |
| 17:16.37 | RobotGuy | I broke buildroot again. I seem to be very good at thaat. |
| 17:18.25 | prpplague | RobotGuy: indeed |
| 17:18.30 | RobotGuy | If I were going to write a Makefile, I would want to learn to do it right, not hard code movs and copies and such. |
| 17:19.24 | RobotGuy | I hate hard coded paths. |
| 17:20.02 | RobotGuy | I think I know what I did to break it this time though. |
| 17:22.42 | prpplague | MothaWork_: ping |
| 17:26.47 | RobotGuy | joins the "I hate Buildroot Makefiles club" |
| 17:27.35 | RobotGuy | I do not understand why people are doing things the way they are in these Makefiles. |
| 17:28.30 | CosmicPenguin | like for example? |
| 17:30.00 | RobotGuy | Like doing forced copiees of stuff from /lib to /usr/lib and such |
| 17:30.48 | RobotGuy | And forced moves and removes |
| 17:35.04 | RobotGuy | There are configure options that take care of making sure things are put where they are supposed to be. |
| 17:35.58 | T0mW | CosmicPenguin: FYI: Intel Q6600 Quad core does OE very nicely |
| 17:36.10 | T0mW | CosmicPenguin: with 8Gig of DDR2 |
| 17:36.16 | RobotGuy | Well, maybe it has something to do with being a cross-compile. I don't know. |
| 17:38.24 | CosmicPenguin | T0mW: yeah - my OE machine is a quad processor box |
| 17:38.25 | CosmicPenguin | very sexy |
| 17:38.56 | chouimat|work | soon I will have a dual quad core :) |
| 17:38.59 | CosmicPenguin | RobotGuy: whats wrong with renames? if packages are putting things in the wrong place, then buildroot should be the one to sort that out |
| 17:39.12 | T0mW | CosmicPenguin: heh, mucho-betterer than a dual core |
| 17:39.36 | T0mW | chouimat|work: yeah, I was considering that |
| 17:39.43 | RobotGuy | CosmicPenguin: It's the moves, copies, removes. It makes it extremely hard to have things put in different places. |
| 17:39.54 | chouimat|work | T0mW: this one will be for home not work :) |
| 17:40.04 | RobotGuy | Maybe it is just because it is a cross-compile. I don't know. |
| 17:42.10 | RobotGuy | This stuff just makes my head hurt, and that is NOT a good thing. |
| 17:45.27 | CosmicPenguin | its tough stuff - everybody just assumes that build systems are easy |
| 17:45.30 | CosmicPenguin | until they start hacking them |
| 17:46.56 | T0mW | chouimat|work: tyan motherboard? |
| 17:51.07 | RobotGuy | CosmicPenguin: I never assume that. I know they are complicated and often difficult to deal with. I just think buildroot is unncessarily so sometimes. |
| 17:52.11 | RobotGuy | But then, I have never gotten this deep into any other build systems either. |
| 17:52.47 | RobotGuy | I actually found and fixed a few buts in the current svn buildroot. :) |
| 17:53.35 | RobotGuy | s/buts/bugs/ |
| 17:56.33 | RobotGuy | Right now I am working on getting PyRo installed and figuring out what I will need actually ON Hammer to make PyRo work for controlling W.A.L.T.E.R. |
| 17:56.57 | RobotGuy | And, of course, I am working on getting a good software build for Hammer. |
| 17:57.23 | RobotGuy | It can be enough to make a crazy person sane! :) |
| 18:04.27 | ds2 | wow busy morning |
| 18:15.03 | RobotGuy | It's my continuing saga of tinkering with buildroot. :) |
| 18:15.56 | chouimat|work | T0mW: don't know yet |
| 18:27.41 | RobotGuy | I'm going back to a default packages dir. |
| 18:45.25 | *** part/#edev adyer (n=adyer@adsl-69-223-13-78.dsl.chcgil.ameritech.net) |
| 18:46.13 | RobotGuy | Don't everyone talk at once now.. :) |
| 18:55.04 | RobotGuy | I have patched binutils 2.18* into the 'official' buildroot. |
| 18:59.49 | *** join/#edev kajaca (n=rakhus@32.139.3.97) |
| 19:10.17 | *** join/#edev ico2 (n=ico2@modem-2641.cougar.dialup.pol.co.uk) |
| 19:20.21 | *** join/#edev Redhatter (n=vk4fsjl@2001:388:f000:0:0:0:0:279) |
| 19:22.00 | prpplague | Redhatter: ping |
| 19:22.19 | Redhatter | prpplague: Yep? |
| 19:22.28 | prpplague | Redhatter: no_body your bot? |
| 19:22.37 | Redhatter | It is |
| 19:22.56 | prpplague | Redhatter: ahh ok, i kicked it early because i didn't know who's it was |
| 19:23.18 | prpplague | Redhatter: if you can please add an option to response as the owner |
| 19:23.24 | prpplague | Redhatter: i.e. the way purl does |
| 19:23.27 | prpplague | ~owner |
| 19:23.27 | purl | TimRiker is my owner |
| 19:23.42 | Redhatter | Ahh okay.. I'll look into it |
| 19:23.56 | *** join/#edev rodserling (n=rodserli@unaffiliated/rodserling) |
| 19:24.04 | rodserling | Hello! |
| 19:24.07 | prpplague | Redhatter: don't mind it for logging and such, just want to make sure who owns it |
| 19:24.19 | prpplague | rodserling: greetings |
| 19:24.22 | Redhatter | It doesn't do a hell of a lot... purl probably does more useful things for this channel. |
| 19:24.38 | prpplague | yea purl is pretty robust |
| 19:24.58 | Redhatter | no_body is based on a heavily hacked up version of Perlbot |
| 19:26.26 | Redhatter | Used to have google there...but that's broke now and I haven't gotten around to fixing it yet. |
| 19:26.37 | *** join/#edev magnetron (n=magnetro@unaffiliated/magnetron) |
| 19:27.09 | RobotGuy | I had a bot that had pretty good AI - using Alice. I wish I had all my configs for that bot now. |
| 19:28.28 | Redhatter | no_body's AI, with the right DB, can come up with some pretty funny stuff. |
| 19:29.06 | Redhatter | http://qdb.longlandclan.yi.org <-- some of the early ones show a bot 'jeeves' (same code as no_body) coming up with some good one-liners. |
| 19:29.11 | rodserling | Anything I can do with some old dvd players? |
| 19:30.25 | Redhatter | Depends what they are... I'm thinking likely, not a lot, as I doubt a DVD player would have that fast a CPU or that much RAM. |
| 19:30.45 | RobotGuy | rodserling: Canibalize the working parts and use them in a robot? |
| 19:30.47 | SpeedEvil | typically not |
| 19:30.59 | SpeedEvil | the DVD decoder is a seperate chip typically |
| 19:31.01 | Redhatter | They generally rely on some hardware MPEG2 decoder to do most of the magic. |
| 19:31.07 | rodserling | :D I have very little clue about electronics :P |
| 19:31.48 | Redhatter | What did you have in mind, rodserling? |
| 19:32.10 | rodserling | Anything, I just want to learn about circuitry |
| 19:32.27 | Redhatter | I'd suggest that's probably not the best place to start ;-) |
| 19:32.32 | rodserling | :P |
| 19:32.49 | prpplague | http://blog.makezine.com/archive/2005/07/how_to_dvd_play.html |
| 19:33.24 | Redhatter | Some of these N-in-one (where N is some figure over 20) electronics kits will teach you a lot more. |
| 19:33.36 | Redhatter | As will a breadboard and some components, with a suitable book. |
| 19:35.02 | rodserling | I think I remember some program that could do what a breadboard could |
| 19:35.30 | Redhatter | There are plenty about.. mostly for Win32... some like Qucs are cross-platform. |
| 19:36.08 | Redhatter | Nothing beats actually fiddling with real hardware though. |
| 19:37.08 | RobotGuy | I just bought a really cool robotics kit the other day - has three robots you can assemble without soldering. |
| 19:37.50 | rodserling | Awesome. |
| 19:37.53 | RobotGuy | No, it's not for me, but to be able to demonstrate robots to people in care facilities. |
| 19:39.38 | Redhatter | rodserling: Approximately what level of understanding are you at with electronics? |
| 19:40.08 | rodserling | beginner, I'd guess |
| 19:41.01 | Redhatter | i.e. if I gave you a light bulb, a switch, some wires and a battery, could you hook them up and make it work? |
| 19:41.31 | Redhatter | do things like "Ohm's Law" ring a bell? |
| 19:42.14 | rodserling | I could |
| 19:43.29 | RobotGuy | YAY! My build of wxPython worked this time, after updating my sources. :D |
| 19:44.15 | Redhatter | rodserling: I'd probably go down to the local electronics shop, and see what kits are available. Also look into some book that covers some circuit basics. |
| 19:44.33 | RobotGuy | rodserling: Where do you live? |
| 19:44.40 | rodserling | Alrighty. |
| 19:44.44 | Redhatter | rodserling: A local amateur radio club, may also be able to assist. |
| 19:44.44 | *** join/#edev adyer (n=adyer@adsl-69-223-13-78.dsl.chcgil.ameritech.net) |
| 19:50.24 | RobotGuy | Redhatter: Oh yeah, that would be a great resource. |
| 19:52.00 | dijenerate | hi all |
| 19:52.49 | RobotGuy | dijenerate: Greets |
| 19:53.23 | dijenerate | Question to anyone... how would you probe for a ps/2 trackstick controller that doesn't seem to be PnP? |
| 19:53.30 | RobotGuy | rodserling: Check for a local robotics club also. We robotics folks are usually very helpful. |
| 19:53.46 | dijenerate | dmesg does not show it as being seen at boot |
| 19:54.23 | prpplague | dijenerate: hehe, my choice would be to disassemble the device and follow the traces |
| 19:54.30 | prpplague | dijenerate: but i doubt you are going to do that |
| 19:54.40 | prpplague | dijenerate: you get a good chiplist on the dinoboard? |
| 19:55.01 | prpplague | runs an errand |
| 19:55.06 | dijenerate | prpplague: with my jackhammer hands, disassembly is always a last resort |
| 19:56.04 | dijenerate | prpplague: not yet on the Dinoboard but apparently someone at the manufacturer's is interested in helping |
| 19:56.05 | RobotGuy | Oh, a name for a new TCT board, like one with an sc3244x - Jack Hammer. :D |
| 19:56.41 | dijenerate | lol |
| 19:56.48 | RobotGuy | prpplague: You mentioned the other day a board with an s3c2440.. ;) Something with lots of RAM and Flash?? |
| 19:56.53 | dijenerate | I'm touched *sniff* |
| 19:58.26 | dijenerate | I know it's on the PS/2 bus with the Synaptic touchpad... it shows in windows, has a special driver and all, but can't seem to get it to show under linux |
| 19:59.08 | adyer | dijenerate: I've never heard of anything ps/2 being PnP :-) |
| 19:59.30 | dijenerate | well Mice have been made to be.... lol |
| 20:00.20 | dijenerate | my prob is I have little experience with this... (I/O devices in general) |
| 20:00.56 | dijenerate | so I don't know how to probe for or detect a PS/2 device like this |
| 20:01.40 | dijenerate | I've been spoiled by new kernels' tendencies to auto detect them |
| 20:02.46 | dijenerate | adyer: how would you detect such a device... I'm trying to see if existing drivers will work with it through trial and error first |
| 20:03.14 | dijenerate | before attempting anything more difficult |
| 20:03.41 | adyer | dijenerate: I think ps/2 mouse is like the keyboard - you talk to a microcontroller, whihc passes messages back and forth to the actual device. |
| 20:04.33 | dijenerate | so in other words I need the microcontroller's command set if this doesn't conform to standards... |
| 20:04.49 | adyer | dijenerate: yep. |
| 20:05.20 | dijenerate | and chances are, if it conformed to standards it would already be working... |
| 20:05.55 | adyer | dijenerate: is it a mouse replacement? |
| 20:06.22 | dijenerate | yeah, along with the synaptics trackpad on the same laptop |
| 20:06.36 | dijenerate | which shares the bus with it |
| 20:06.48 | adyer | dijenerate: extra icky when there are two things on a ps/2 bus |
| 20:07.13 | adyer | dijenerate: maybe the windows didn't probe it, but shows it because a driver was loaded for it |
| 20:07.41 | dijenerate | well there is a driver and it has it's own controller |
| 20:08.29 | dijenerate | but I figured they might not have spent that much time and money to make their own device when tracksticks are pretty standard |
| 20:09.09 | adyer | dijenerate: I think windows is lying to you :-) |
| 20:09.23 | dijenerate | you think the fact they share the bus maybe why the trackstick doesn't show? |
| 20:09.47 | RobotGuy | prpplague: PING |
| 20:10.07 | dijenerate | adyer: what's new there? windows lies to us all |
| 20:10.17 | RobotGuy | wants to eventually have a Hammer controlled Biped. :) |
| 20:11.06 | chouimat|work | RobotGuy: I had one ... but he rebelled and wacked me with the hammer |
| 20:11.14 | adyer | dijenerate: ps/2 is really meant for only one device. there are a lot of adapters (like bar code readers and such) that fake it out in different ways |
| 20:11.34 | RobotGuy | Can somebody help me find a 0.1" space connector I can use for stuff like power connections? A right angle locking connector would be ideal, but others are OK. |
| 20:11.48 | SpeedEvil | power? |
| 20:11.53 | RobotGuy | chouimat|work: Ha ha! :P |
| 20:12.04 | SpeedEvil | ordinary IDC connectors work fine at about a half amp a pin |
| 20:12.05 | RobotGuy | SpeedEvil: Yes, like battery. |
| 20:12.07 | adyer | RobotGuy: how much power? |
| 20:12.09 | chouimat|work | RobotGuy: Iknow it was too hard to resist :) |
| 20:12.11 | SpeedEvil | look in the RC world |
| 20:12.12 | dijenerate | adyer: any suggestions where to start? |
| 20:12.27 | RobotGuy | adyer: Has to be able to handle 10 Amps minimum. |
| 20:12.39 | adyer | RobotGuy: ROFL |
| 20:12.42 | SpeedEvil | RobotGuy: google miniature battery connector R/C |
| 20:13.12 | RobotGuy | SpeedEvil: This is where I get frustrated, because I have been having problem finding what I need. That's why I am asking for help. |
| 20:13.30 | adyer | RobotGuy: try something like a minifit or minifit jr - not 0.100" but much better for power type stuff |
| 20:13.30 | RobotGuy | A hexapod can draw 8 amps when walking. |
| 20:13.30 | SpeedEvil | And I gave you some google terms that should generate good hits |
| 20:14.15 | RobotGuy | The search engines on places like digikey always frustrate me. |
| 20:14.32 | SpeedEvil | That's why I suggested google. |
| 20:15.40 | adyer | dijenerate: not sure how to go other than to google the mfg, and maybe hook up a logic analyzer, capture the ps/2 port under windows and disassemble. |
| 20:18.10 | dijenerate | hmmm... *sigh* disassembly... |
| 20:19.47 | SpeedEvil | http://www.electrifly.com/miscproducts/adapters.html forex |
| 20:21.29 | RobotGuy | signs |
| 20:21.34 | RobotGuy | sighs too |
| 20:23.11 | prpplague | RobotGuy: pong |
| 20:23.49 | RobotGuy | prpplague: Ah, there you are. :D What were you saying about 128/128 the other day, and something that has an S3C2440 on it?? :D |
| 20:23.49 | adyer | RobotGuy: http://www.molex.com/pdm_docs/ps/PS-5556-001.pdf specs for minifit-jr |
| 20:24.09 | prpplague | RobotGuy: no clue |
| 20:24.34 | adyer | RobotGuy: don't forget that most people consider it wise not to come close to the rated power per pin :-) |
| 20:25.04 | RobotGuy | adyer: I know. Specs are fine, but I need connectors I can USE. |
| 20:27.33 | SpeedEvil | http://www3.towerhobbies.com/cgi-bin/wti0001p?&I=LXDKA2 |
| 20:27.44 | SpeedEvil | looks easy and suitable |
| 20:31.05 | *** join/#edev emil|k (i=emil@e179234007.adsl.alicedsl.de) |
| 20:32.43 | adyer | RobotGuy: same kind of connectors in a PC power supply - what's no usable about that? |
| 20:33.09 | adyer | RobotGuy: http://dkc3.digikey.com/PDF/T081/0195-0196.pdf |
| 20:33.28 | RobotGuy | adyer: Has to be for pcb. |
| 20:34.02 | adyer | RobotGuy: you pcb-pcb or wire-pcb? |
| 20:34.34 | RobotGuy | adyer: Has to be PCB to connector for cable. |
| 20:38.44 | prpplague | RobotGuy: i think you were refering to the capabilities of the s3c24xx series for maximum sdram configuration |
| 20:39.06 | prpplague | RobotGuy: in order max the sdram our for the s3c24xx you would need 4 sdram devices |
| 20:39.30 | prpplague | RobotGuy: each being a 512mb x 16 |
| 20:39.33 | RobotGuy | prpplague: Possibly. I would like to put a full LAMP configuration on Hammer but don't know if that would be workable in 32 MB RAM. Especially the M part. |
| 20:39.44 | prpplague | RobotGuy: that woudld give you a total of 256mb sdram |
| 20:39.53 | adyer | RobotGuy: how about a pcb mount terminal block? http://microcontrollershop.com/product_info.php?currency=USD&products_id=1956 |
| 20:40.00 | RobotGuy | prpplague: Wow. I'd be happy with 128 Meg. |
| 20:40.00 | prpplague | RobotGuy: why would you want to do that? |
| 20:40.22 | prpplague | LAMP on the hammer doesn't make much sense |
| 20:40.44 | RobotGuy | adyer: Yes, something like that but has to be sturdier than what Lynxmotion uses on their boards. |
| 20:40.55 | RobotGuy | prpplague: Why not? |
| 20:41.28 | prpplague | RobotGuy: LAMP is a configuration for a server based device, the hammer is an embedded device |
| 20:41.37 | prpplague | RobotGuy: doesn't really fit the need/resources of the device |
| 20:41.48 | RobotGuy | prpplague: Yes, I know the difference. |
| 20:42.14 | prpplague | RobotGuy: what would do with it if it was running a LAMP config? |
| 20:42.48 | RobotGuy | prpplague: Yeah, people always tell me this can't be done or isn't workable or isn't possible. :( Isn't that what strangles progress and creativeness? |
| 20:42.54 | RobotGuy | sighs |
| 20:43.12 | RobotGuy | prpplague: Web interface to W.A.L.T.E.R. for one thing. |
| 20:43.27 | RobotGuy | prpplague: Using PHP and/or Python. |
| 20:43.39 | chouimat|work | RobotGuy: look at lighttpd |
| 20:43.50 | prpplague | RobotGuy: there is a big difference between being possible and being practical |
| 20:44.05 | prpplague | RobotGuy: you can do a very robust web interface without a LAMP config |
| 20:44.17 | RobotGuy | prpplague: Why would LAMP not be practical on an embedded processor having the appropriate resources? |
| 20:44.18 | prpplague | RobotGuy: we have complete web interface for our handhelds |
| 20:44.30 | Redhatter | I was surprised to discover how much I could do with busybox ash and httpd. |
| 20:44.42 | RobotGuy | chouimat|work: I have been looking into boa |
| 20:44.57 | prpplague | RobotGuy: 1) LAMP is designed for server sized devices 2) the power and size of the software has to match the intended usage |
| 20:45.02 | RobotGuy | I will look at lighthttpd also |
| 20:45.18 | Redhatter | Had a small but simple web-based interface, running quite responsively, on a 33MHz 386 |
| 20:45.37 | RobotGuy | prpplague: Does it -really- have to match the intended usage? |
| 20:45.50 | RobotGuy | Sometimes I like to do things just because I want to. :D |
| 20:45.57 | prpplague | RobotGuy: uh, thats kinda the whole statement of embedded |
| 20:46.18 | RobotGuy | Or to answer questions like: Hmmm, is it feasible to put full LAMP on an embedded system? |
| 20:46.42 | Redhatter | I wouldn't think so |
| 20:46.56 | RobotGuy | Sometimes I think there is too much negativity here. |
| 20:47.00 | Redhatter | If you need a DB, I'd be looking at something like sqlite, no need for a daemon |
| 20:47.01 | prpplague | RobotGuy: to answer your question yes, it is also possible to use a VW bug as a dump drug, but really, why? |
| 20:47.25 | RobotGuy | prpplague: I don't think the analogy is appropriate. |
| 20:47.26 | Redhatter | Perl or Python for a lightweight language to run the scripts. |
| 20:47.41 | RobotGuy | Redhatter: Yes, I am planning to use Python. |
| 20:47.43 | prpplague | RobotGuy: not negativetly, its just that you are still thinking in desktop/server terms rather than in embedded terms |
| 20:48.03 | RobotGuy | prpplague: Maybe I just want to see if it can be done? What is wrong with that? |
| 20:48.12 | adyer | RobotGuy: for production systems where you get dinged on the cost, the question is usually what's the minimum necessary to do the job. Can't think too much like a hobby here. |
| 20:48.14 | prpplague | RobotGuy: nothing at all |
| 20:48.17 | Redhatter | recently tried coding a webapp in pure C to see what it was like. |
| 20:48.19 | prpplague | RobotGuy: you are welcome to do it |
| 20:48.40 | RobotGuy | I also want an IRC interface for W.A.L.T.E.R. |
| 20:48.41 | prpplague | RobotGuy: but re-inventing the wheel doesn't really increase you capability or knowledge |
| 20:48.42 | Redhatter | http://gallery.longlandclan.yi.org <-- code at http://git.longlandclan.yi.org/ ... wasn't that painful really. |
| 20:49.14 | RobotGuy | Maybe it is just me, but sometimes being here leaves me feeling down. |
| 20:49.42 | prpplague | RobotGuy: hehe, again, i think your problem is you are still think in terms of desktop |
| 20:49.49 | Redhatter | You're welcome to try... but I'm thinking it'd be a bit slow. |
| 20:49.51 | RobotGuy | prpplague: Maybe. |
| 20:50.20 | Redhatter | PHP is awful on my Qube2 (QED/MIPS RM5231 at 250MHz) |
| 20:50.22 | chouimat|work | sends RobotGuy a few lcp2148 boards |
| 20:50.31 | RobotGuy | Aall I think I would really need is more RAM. I already run LAMP on a Pentium 3/733 MHz with 192 Meg of RAM. |
| 20:50.44 | prpplague | RobotGuy: you have to realize that linux developers have been doing embedded designs for along time, they've tried doing pretty much everything you've talked about doing and they've developed solutions specific for the embedded design |
| 20:50.59 | RobotGuy | I have two LPC2148 boards - a header board and a proto board from Spark Fun/Olimex. :D |
| 20:51.34 | prpplague | RobotGuy: there is nothing you can't do for any embedded web interfae with LAMP, that you can't do with boa+sqlite+shell script |
| 20:51.56 | RobotGuy | prpplague: I am looking at sqlite and boa. |
| 20:52.22 | RobotGuy | I'm pretty good at writing shell scripts in various languages. :) |
| 20:52.38 | RobotGuy | I would probably glue it all together with Python. |
| 20:53.18 | *** join/#edev dijenerate (n=dijenera@72.22.132.215) |
| 20:53.35 | Redhatter | What about Apache httpd with mod_python? Or does lighttpd or boa have an equivalent? |
| 20:54.03 | prpplague | dijenerate: my wife is dying for a roti! can you run by chefette and fedex her a chick and potato roti? |
| 20:54.03 | RobotGuy | I still have to figure out a software configuration for W.A.L.T.E.R. that works. |
| 20:54.04 | Redhatter | Having the python interpreter resident might just help performance a bit |
| 20:54.48 | prpplague | i do all of our web interface with standard shell script using busybox |
| 20:55.04 | RobotGuy | Redhatter: I would love to have Apache and mod_python on Hammer, with mod_php also. |
| 20:55.35 | RobotGuy | I think I could probably still use sqlite for the database and I have a 2 GB Thumb Drive. |
| 20:55.42 | Redhatter | prpplague: How do you handle GET/POST variables... I could sorta do the former, but couldn't figure out the latter. I wound up writing a C app that would decode that stuff. |
| 20:55.59 | dijenerate | I'll do my best, but fedex closed 20 minutes ago |
| 20:56.06 | RobotGuy | There are Python modules available that handle all that. :D |
| 20:56.11 | prpplague | one sec phone |
| 20:56.44 | Redhatter | RobotGuy: Indeed... as with Perl... and they're built into PHP. |
| 20:59.29 | prpplague | dijenerate: hehe |
| 20:59.50 | prpplague | RobotGuy: boa and the other light httpd's support the usage of shell scripts as cgi |
| 21:00.01 | prpplague | RobotGuy: very easy to use |
| 21:01.00 | prpplague | RobotGuy: we do complete control items such as check boxes to turn features on/off , upload of custom config files, upload of update kernel and root filesystems, boot screen image configuration, and cloning functionality |
| 21:01.31 | RobotGuy | prpplague: Di they support using PHP or Python? Both of those can be used for shell scripting also. |
| 21:01.41 | prpplague | RobotGuy: we even have an option where you can get web page which shows the exact same information as it is posted on the framebuffer of the handheld |
| 21:01.54 | RobotGuy | prpplague: Where is all of this? |
| 21:02.09 | prpplague | RobotGuy: i'm sure they do, but if you have shell why do you need the php or python? |
| 21:02.17 | prpplague | RobotGuy: on our handheld devices |
| 21:03.16 | RobotGuy | prpplague: For me, Python is easier and already has modules available for stuff I want to do. Both PHP and Python have many many modules available for doing just about anything one might want to do. |
| 21:03.34 | RobotGuy | prpplague: I gather there is no available toolkit others might use then? |
| 21:03.49 | prpplague | RobotGuy: we have a sdk, but you have to purchase a handheld |
| 21:04.06 | RobotGuy | prpplague: Bingo, this is exactly why I have to keep reinventing the wheel.. |
| 21:04.31 | prpplague | RobotGuy: there are tons of other devices that use the same operations |
| 21:04.53 | RobotGuy | All devices I probably could never afford. |
| 21:04.57 | prpplague | RobotGuy: what kind of operations did you have in mind? |
| 21:05.14 | prpplague | RobotGuy: there are plenty of them with the source already posted |
| 21:05.18 | RobotGuy | Web interfaces like to my robot. |
| 21:05.42 | prpplague | RobotGuy: give me an example of an ideal screen? |
| 21:05.53 | RobotGuy | Making a GUI for people to use to interface to my robot. |
| 21:06.37 | RobotGuy | I'm starting to look at OpenMoko, but can't seem to find the software. |
| 21:06.55 | prpplague | RobotGuy: ok so , what an eight way directional arrow along with some sensor reports? |
| 21:06.57 | *** join/#edev ico2 (n=ico2@modem-2641.cougar.dialup.pol.co.uk) |
| 21:07.07 | RobotGuy | It would be cool to have Hammer supported by Open Embedded too. |
| 21:07.14 | prpplague | RobotGuy: openmoko is all OE based |
| 21:07.19 | RobotGuy | prpplague: That might be simplistic, but yes, a start. |
| 21:07.29 | prpplague | RobotGuy: thats like a 5 minute deal |
| 21:07.59 | RobotGuy | prpplague: I know you are way more experienced at this stuff than me. I want to learn. |
| 21:08.17 | RobotGuy | Examples are good. |
| 21:08.21 | SpeedEvil | RobotGuy: look on the wiki. search mokomakefile |
| 21:08.35 | prpplague | RobotGuy: my suggestion would be to compile boa for your desktop and just tinker with it's capabilities |
| 21:08.53 | prpplague | RobotGuy: see what you can do _without_ adding anything else |
| 21:08.54 | RobotGuy | prpplague: OK. |
| 21:09.07 | RobotGuy | I will do that. |
| 21:09.08 | prpplague | RobotGuy: then add _just_ the things you lack with just boa |
| 21:09.38 | *** join/#edev bbradley (n=bbradley@82-32-35-42.cable.ubr05.azte.blueyonder.co.uk) |
| 21:09.47 | prpplague | it is much easier to add in what is missing than to know what to take out that isn't nessary |
| 21:09.48 | RobotGuy | I can set boa up on this PC, instead of my server. |
| 21:10.11 | prpplague | yep, then you can use your browser to just connect to localhost |
| 21:10.26 | RobotGuy | prpplague: True, but I don't know what I would want to take out of other stuff. I tend to error on the side of having more capabilities than less than I may need at some point. |
| 21:10.45 | prpplague | RobotGuy: which is exactly my point |
| 21:10.48 | prpplague | it is much easier to add in what is missing than to know what to take out that isn't nessary |
| 21:11.20 | RobotGuy | I understand what you are saying. I'll give boa a try on my PC. |
| 21:12.09 | RobotGuy | I'm also still trying to get a software config that works for what I want to do with Hammer. |
| 21:12.23 | RobotGuy | Maybe I need to rethink my initrd setup. |
| 21:12.44 | RobotGuy | Uncompressed it is 7.9 MB |
| 21:13.30 | RobotGuy | Hee hee, I got binutils 2.18* patched into my 'official' buildroot. :D |
| 21:14.31 | RobotGuy | I'd still really like something with more memory, more uarts accesible, sound, and lcd stuff. More flash really isn't needed with my thumb drive. |
| 21:15.30 | RobotGuy | I talk about Hammer every chance I get. :D |
| 21:18.18 | prpplague | RobotGuy: thats what i was trying to tell you last week, i'd either use a very small initrd(or none at all) along with a jffs2 rootfs |
| 21:18.27 | prpplague | RobotGuy: that will free up alot of memory for you |
| 21:18.49 | RobotGuy | prpplague: Oh, OK. I did not fully understand then, sorry. |
| 21:18.51 | prpplague | RobotGuy: with just the kernel along with a jffs2 root, you can get around 24mb sdram |
| 21:19.10 | RobotGuy | Using JFFS2 for the ramdisk, right? |
| 21:19.21 | prpplague | RobotGuy: no, jffs2 for rootfs |
| 21:19.32 | RobotGuy | initrd? |
| 21:19.45 | prpplague | RobotGuy: hehe no, |
| 21:20.01 | prpplague | RobotGuy: you know the jffs2 partition you have on the hammer now? |
| 21:20.23 | RobotGuy | prpplague: Yes, for the flash, which I seem to have lost access to. |
| 21:20.31 | *** join/#edev MonMotha (i=monmotha@c-69-245-153-193.hsd1.in.comcast.net) |
| 21:20.55 | prpplague | RobotGuy: ok, now, buildroot can generate an root file system image in the form of a jffs2 image, you would copy this jffs2 image to that partition |
| 21:21.12 | RobotGuy | prpplague: I am not using the flash anyway. |
| 21:21.39 | prpplague | RobotGuy: now, when the system boots, it uses a small initrd also generated by buildroot, once the system is up, it performs an operation called a chroot (change root) |
| 21:21.41 | RobotGuy | But I am using my thumb drive to put software and such on. The Flash is not large enough. |
| 21:21.57 | prpplague | RobotGuy: the root fs will switch from being the initrd to the jffs2 in flash |
| 21:22.11 | prpplague | RobotGuy: which means the initrd is no longer needed and can be removed from memory |
| 21:22.20 | RobotGuy | The Flash partition is not large enough. |
| 21:22.22 | *** join/#edev CosmicPenguin (n=nobody@163.181.251.103) |
| 21:22.22 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 21:22.49 | prpplague | RobotGuy: hmm, 14mb seems plenty large |
| 21:23.10 | RobotGuy | prpplague: Not for a native toolchain. |
| 21:23.13 | prpplague | RobotGuy: what size is your initrd uncompressed? |
| 21:23.25 | prpplague | RobotGuy: ug what do you need a native toolchain for? |
| 21:23.35 | prpplague | RobotGuy: and besides moving to jffs2 was to free sdram |
| 21:23.44 | RobotGuy | 7.9 MB uncompressed |
| 21:24.13 | prpplague | RobotGuy: the jffs2 partition is 14mb , so that fits with no problems |
| 21:24.54 | RobotGuy | Why do I need to justify it? There are some things I just want to build natively, or maybe make a change on the fly. |
| 21:25.18 | RobotGuy | I've lost access to the flash for some reason. :( |
| 21:25.33 | prpplague | RobotGuy: ok, like i said some of things you want to do don't make alot of sense |
| 21:25.47 | prpplague | RobotGuy: but anyway, you can do a jffs2 partition and free up alot of sdram |
| 21:25.52 | RobotGuy | prpplague: I understand that they don't make sense to you. |
| 21:26.24 | prpplague | RobotGuy: thats why i ask you to explain is so that i can understand |
| 21:26.32 | RobotGuy | I need to also have Python and whatever modules I need to add to Python, PyRo, etc. |
| 21:26.47 | prpplague | RobotGuy: thats fine, you can do all that on your sd card or thumbdrive |
| 21:26.47 | RobotGuy | prpplague: OK, I will try to explain from now on. |
| 21:27.09 | prpplague | RobotGuy: since you are tight on sdram space, its probably best you look at doing a jffs2 root |
| 21:27.14 | RobotGuy | prpplague: So I would use the flash for what would normally be in the initrd? |
| 21:27.28 | prpplague | RobotGuy: correct |
| 21:27.44 | RobotGuy | prpplague: OK, I follow so far, or at least think I do. |
| 21:27.49 | prpplague | RobotGuy: initrd uses ram, jffs2 uses flash |
| 21:28.14 | prpplague | if you are low on ram and have plenty of flash, use the flash for root instead of ram as root |
| 21:28.28 | RobotGuy | prpplague: Why did I lose access to my flash suddenly? What did that have to do with changing my kernel config? I didn't change any of the MTD stuff. |
| 21:28.46 | RobotGuy | prpplague: OK, and the Flash will hold a larger root. |
| 21:28.50 | prpplague | RobotGuy: there are a number of reasons you could be having problem |
| 21:29.14 | prpplague | RobotGuy: mismatched partition spaces, the space isn't formated, overlapping partitions |
| 21:29.17 | RobotGuy | prpplague: I keep having those length errors come and go and have no idea what I am doing to cause them. |
| 21:29.43 | prpplague | RobotGuy: your length errors are based on sizes that don't match |
| 21:30.24 | RobotGuy | prpplague: That is what I do not understand. |
| 21:30.51 | prpplague | RobotGuy: pastebin you mach-tct_hammer.c file |
| 21:30.59 | RobotGuy | OK |
| 21:34.01 | RobotGuy | prpplague: http://www.pastebin.ca/957275 |
| 21:35.36 | prpplague | RobotGuy: how big is your compressed initrd image? |
| 21:36.17 | RobotGuy | 2.3 MB compressed |
| 21:36.27 | RobotGuy | 7.9 MB uncompressed |
| 21:37.01 | prpplague | RobotGuy: ok so currently you have your partition set for 256k for bootloader, 1024k for kernel and 2048k for initrd |
| 21:37.30 | prpplague | 2.4mb > 2048k |
| 21:37.34 | prpplague | 2.3mb > 2048k |
| 21:37.37 | RobotGuy | prpplague: Not exactly. They are larger than what I need in some cases. |
| 21:37.56 | prpplague | RobotGuy: you just said that the compressed kernel is 2.3mb |
| 21:38.07 | RobotGuy | Compressed initrd |
| 21:38.26 | prpplague | ok well 2.3mb > 2048k |
| 21:38.42 | RobotGuy | I know. I have it as 4096 in APEX. |
| 21:39.06 | prpplague | RobotGuy: that doesn't mean anything if the kernel thinks the size of the initrd is no bigger than 2048k |
| 21:39.26 | RobotGuy | prpplague: I think I have it set to the same for the kernel. |
| 21:39.41 | prpplague | RobotGuy: you currently have it set for 0x340000 |
| 21:40.01 | prpplague | RobotGuy: which is 256k for bootloader, 1024k for kernel, and that leaves 2048k |
| 21:40.12 | RobotGuy | prpplague: Ok, I need that larger then. |
| 21:41.21 | prpplague | RobotGuy: yes, you should always calculate that size value by adding up the size of the bootloader 256k, plus the size of the kernel 1024k, plus the size of the compressed initrd +256k |
| 21:41.44 | prpplague | RobotGuy: ok now pastebin your apex .config file |
| 21:41.50 | RobotGuy | OK |
| 21:42.47 | RobotGuy | prpplague: http://www.pastebin.ca/957290 |
| 21:45.06 | prpplague | RobotGuy: ok that looks correct, now pastebin your kernel .config |
| 21:47.26 | *** join/#edev Crofton (n=balister@206.196.151.254) |
| 21:48.13 | prpplague | Crofton: hey hey |
| 21:50.28 | RobotGuy | prpplague: http://www.pastebin.ca/957304 I know you are gonna give me a hard time about my config.. |
| 21:50.45 | RobotGuy | And I need to add some more stuffs. |
| 21:50.49 | prpplague | just making sure the right options are on |
| 21:50.53 | RobotGuy | OK |
| 21:51.41 | Crofton | hi |
| 21:51.47 | Crofton | I'm on vacation :) |
| 21:52.19 | prpplague | Crofton: ahh |
| 21:52.35 | prpplague | RobotGuy: first items cmdline partitions and afs need to be turned off |
| 21:52.52 | RobotGuy | prpplague: I thought I had those turned off. |
| 21:53.07 | prpplague | RobotGuy: config says they are on |
| 21:53.09 | RobotGuy | My mistake |
| 21:53.34 | RobotGuy | is this in the MTD section? |
| 21:54.43 | RobotGuy | I found the cmdline in MTD |
| 21:55.14 | prpplague | RobotGuy: yea in the MTD section |
| 21:55.22 | RobotGuy | OK, I figured out your abbreviation - not good to do when teaching somebody. :P |
| 21:55.29 | RobotGuy | Got them turned off now. |
| 21:55.50 | prpplague | RobotGuy: and generally speaking you can;t access a jffs2 partition if your kernel doesn't have jffs2 support |
| 21:55.54 | prpplague | RobotGuy: need to turn jffs2 on |
| 21:56.06 | RobotGuy | prpplague: Oh, yeah, silly me. |
| 21:56.44 | prpplague | RobotGuy: that should be enought to get the jffs2 mounted |
| 21:57.11 | RobotGuy | I see JFFS in File Systems, but I think that is not right? |
| 21:57.41 | prpplague | RobotGuy: its under misc filesystems |
| 21:57.48 | RobotGuy | OK |
| 21:58.18 | Redhatter | Silly question, why does JFFS2, etc, *depend* on MTD? |
| 21:58.33 | Redhatter | Why can't one use it, on for instance, a USB flash drive? |
| 21:59.00 | RobotGuy | I don't know (or remember) how that got turned off. |
| 21:59.07 | prpplague | Redhatter: the more important question is why would you want to use it on a USB flash drive? |
| 21:59.23 | Redhatter | prpplague: Wear levelling? |
| 21:59.38 | prpplague | Redhatter: thats just it, thumb drive controller chips do the wear leveling |
| 21:59.47 | prpplague | Redhatter: its part of their core programming |
| 21:59.58 | Redhatter | What about CF/SD cards though? |
| 22:00.02 | prpplague | Redhatter: why do double wear leveling? ir just makes it slow |
| 22:00.05 | prpplague | Redhatter: same |
| 22:00.21 | Redhatter | Ahh okay. |
| 22:00.34 | RobotGuy | I need to turn on Phidgets support too. |
| 22:00.55 | Redhatter | Still... suppose you had a generic block device, that happened to be flash-based, and didn't implement wear leveling... |
| 22:00.56 | prpplague | Redhatter: JFFS2 was designed to be used with flash devices that are in raw access, which means there is no controller chip doing the wear leveling or crc checks |
| 22:01.45 | prpplague | Redhatter: if you have flash with no wear leveling you need MTD in order to access it in the raw mode nessary to implement wear leveling |
| 22:02.03 | RobotGuy | rebuilds kernel |
| 22:02.04 | Redhatter | Fair enough... well that explains it somewhat :-) |
| 22:05.03 | Redhatter | So if I manage to burn a hole in a USB flash drive, or a CF/SD card, it's a problem with the controller not doing the leveling properly? |
| 22:05.12 | Redhatter | Rather than the choice of filesystem? |
| 22:07.26 | prpplague | Redhatter: correct |
| 22:07.58 | prpplague | Redhatter: that is why choosing the right device makes a difference, sandisk wear leveling is much better than say the cheapo Adata devices |
| 22:09.46 | prpplague | Redhatter: the other choice is to use a device that has not controller such as an xD card |
| 22:09.58 | prpplague | Redhatter: if you use something like xD you can run jffs2 on it |
| 22:10.09 | prpplague | Redhatter: giving you control over the wear leveling |
| 22:10.12 | ds2 | wear leveling HE HE HE HE HE |
| 22:12.01 | *** join/#edev bbradley (n=bbradley@82-32-35-42.cable.ubr05.azte.blueyonder.co.uk) |
| 22:13.07 | prpplague | RobotGuy: btw, here is a great reference on usbnetworking - http://wiki.openmoko.org/wiki/USB_Networking |
| 22:24.34 | RobotGuy | Durn, he gone. |
| 22:32.24 | *** join/#edev redblue (i=star@ppp249.108-253-207.mtl.mt.videotron.ca) |
| 22:51.59 | *** join/#edev adyer (n=adyer@adsl-69-223-13-78.dsl.chcgil.ameritech.net) |
| 23:05.43 | *** join/#edev ico2_ (n=ico2@modem-145.chimpanzee.dialup.pol.co.uk) |
| 23:08.31 | *** join/#edev redblue (i=star@ppp104.108-253-207.mtl.mt.videotron.ca) |
| 23:22.06 | *** part/#edev adyer (n=adyer@adsl-69-223-13-78.dsl.chcgil.ameritech.net) |
| 23:53.53 | *** join/#edev prpplague (n=dave123_@ppp-70-244-160-94.dsl.rcsntx.swbell.net) |
| 23:54.03 | prpplague | boo |
| 23:56.04 | prpplague | are we having fun yet? |
| 23:58.08 | prpplague | checks out the help wanted ads |