00:11.34 | hunger | The new page looks great, but it has so little information... |
00:16.48 | hunger | Looks like I managed to build picogui. Now, how do I run it with a nice theme? |
00:18.03 | hunger | I keep getting "connection broken" whenever I try to access the information on the old wiki. |
00:19.54 | *** join/#picogui kidder (~s@h24-65-171-188.gv.shawcable.net) |
00:40.24 | *** join/#picogui file (jwired@mctn1-3043.nb.aliant.net) |
00:40.42 | *** join/#picogui gonkulator (~brandon@h-69-3-79-3.DNVTCO56.covad.net) |
00:42.27 | file | hi gonkulator |
00:43.41 | gonkulator | hi file |
00:44.05 | file | I missed PnP :( |
00:44.41 | gonkulator | where did Xentac and I go wrong????? |
00:44.48 | Xentac | gonkulator: I don't know... |
00:44.50 | Xentac | ;) |
00:44.52 | file | lol |
00:45.16 | file | now... for the battle of the 29 Windows 95 install floppies |
00:46.26 | file | gonkulator: how are you? |
00:52.06 | file | carpman: Detonation circuits disabled. |
00:52.10 | file | carpman: Please read the manual. |
00:53.20 | gonkulator | file: I am doing well |
00:53.38 | file | gonkulator: that's great |
00:53.50 | file | is this D&D? |
00:53.57 | gonkulator | I dunno |
00:54.03 | file | Magic the Gathering? |
00:54.04 | carpman | file: evangelion and D&D aparently |
00:54.10 | file | oh ic |
00:54.42 | file | gonkulator: Have a pleasant dinner. |
00:54.42 | carpman | Ugh. My head hurts, but the Ego forums are almost done. |
00:55.54 | file | I've got tests tomorrow... then another test Friday |
01:29.12 | *** join/#picogui decaff (~khedspet@clt74-76-015.carolina.rr.com) |
01:51.33 | kidder | hey xen you there??? |
02:16.20 | *** join/#picogui darth_iBook (~silme@aden2-18-dhcp.resnet.Colorado.EDU) |
02:45.10 | carpman | MUAHAHAH!!!! AoF has forums!!!!! |
02:47.43 | Xentac | hehehe |
03:13.40 | gonkulator | PicoBot: seen jojo |
03:13.40 | PicoBot | jojo was last seen on #picogui 107 days, 1 hours, 32 minutes and 43 seconds ago, saying: No. [Mon Nov 11 12:49:54 2002] |
03:13.47 | gonkulator | dang |
03:13.53 | gonkulator | thats a long long time ago... |
03:14.57 | gonkulator | lurgyman: hey dan! |
03:22.06 | lurgyman | gonkulator: bedtime |
03:22.13 | carpman | indeed |
03:22.20 | carpman | night all |
03:23.12 | gonkulator | :) |
03:25.05 | gonkulator | captain_proton: hey David! |
03:25.53 | gonkulator | actually, it is homework time |
04:29.45 | *** join/#picogui invader_zim (~see_it_ch@aden2-135-dhcp.resnet.Colorado.EDU) |
04:39.11 | *** join/#picogui nixman (~Xman@HSE-Ottawa-ppp236174.sympatico.ca) |
04:39.56 | nixman | anyone know pages covering (small) lcd srcrren output from linux ? |
04:45.20 | g-i-r | yay me is back |
05:04.27 | invader_zim | g-i-r: I MUST STUDY.....leave me alone......HAHAAHHAHAHAHAMAHAHHAHAA |
05:04.42 | invader_zim | i will conquer this pitiful planet |
05:04.58 | invader_zim | Ang-l|5d-til22: what does that mean? |
05:05.03 | invader_zim | oh!!! |
05:10.36 | *** join/#picogui lalo[out] (~lalo@200.203.12.232) |
05:16.19 | captain_proton | scanline is here instead |
05:18.02 | g-i-r | captain_proton: what was the url for the cirque music? |
05:19.39 | captain_proton | http://bzflag.net:3000/public.e17a/music/Cirque_du_Soleil |
05:20.06 | scanline | silly auto-randomizing URLs... |
05:20.42 | captain_proton | i think mine isn't quite auto-randomizing though |
05:20.58 | scanline | hmm |
05:23.30 | captain_proton | oh my god |
05:23.37 | captain_proton | we have all the blakes7 episodes |
05:23.38 | scanline | captain_proton: I still think it's worth trying that oscilloscope test using multiple monitor refresh rates |
05:23.44 | scanline | oh no! |
05:25.35 | captain_proton | i'm scared too |
05:27.28 | g-i-r | lol |
05:36.14 | captain_proton | scanline: what is the full part # of the sharp chip? |
05:37.59 | scanline | IR3T07A |
05:41.00 | captain_proton | wow...so much documentation on this chip |
05:41.48 | scanline | well, it's probably at least 20 years old... |
05:49.07 | g-i-r | w00t! fizix done |
05:50.32 | g-i-r | scanline: how do you get the eject key on the keyboard to work? |
05:50.46 | scanline | I think you don't... there might be a daemon for it, but I haven't seen one |
05:50.59 | g-i-r | ok then, really dumb question, how do you eject the cdrom? |
05:51.07 | g-i-r | nm |
05:51.09 | scanline | eject command |
05:51.53 | g-i-r | right, /me just figured that out |
05:52.12 | g-i-r | that was too obvious |
06:05.03 | g-i-r | I love this movie so much |
06:22.50 | captain_proton | your favorite show? |
06:24.07 | g-i-r | what a great filme |
06:24.11 | g-i-r | s/filme/film/ |
06:24.39 | scanline | 'tis committed |
06:36.23 | captain_proton | scanline: argh! need docs on the 3T07 |
06:36.34 | scanline | mmyep |
06:38.53 | *** join/#picogui lurgyman (~lurgyman@aden2-123-dhcp.resnet.Colorado.EDU) |
06:39.35 | *** join/#picogui lurgyman (~lurgyman@aden2-123-dhcp.resnet.Colorado.EDU) |
06:46.26 | scanline | hopefully this will recognize between two different hand positions |
06:47.00 | njs | neural nets are fun. |
06:47.57 | scanline | Well, I now have a sensor that gives 8 readings.. so I could use inverse electrostatics to determine the hand position from that, but I thought a neural net approach would be worth a try |
06:48.42 | scanline | if this is practical I'll do one in C++ |
06:48.51 | njs | I haven't had a reason to write one in python, but it's hardly difficult :-) |
06:49.25 | njs | good luck; getting neural nets to work often requires some thought. |
06:49.42 | scanline | thanks |
06:49.47 | scanline | this is the first time I've played with them |
06:49.55 | njs | (it sucks; you should just be able to throw arbitrary data at learning algorithms and have it work. but it doesn't. grumph.) |
06:50.55 | scanline | It'd be really nice if the easy approach actually works... train the neural net by giving it actual data for a bunch of coordinates, then use it to convert readings to coorcinates |
06:51.26 | captain_proton | scanline: your output layer is going to be huge |
06:51.38 | scanline | output layer? |
06:51.41 | *** join/#picogui lurgyman (~lurgyman@aden2-123-dhcp.resnet.Colorado.EDU) |
06:51.55 | *** join/#picogui lurgyman (~lurgyman@aden2-123-dhcp.resnet.Colorado.EDU) |
06:52.03 | njs | the main trick is to preprocess the input into features that are actually useful for solving the problem at hand; the best backgammon players in the world are neural nets, but you have to precompute a few tactical things to feed your net... |
06:52.22 | captain_proton | scanline: in the network |
06:52.38 | captain_proton | scanline: each node only has two states - on or off |
06:53.01 | scanline | yeah.. I need to keep thinking about whether it's best to use absolute readings or relative to a baseline, whether to scale them relative to each other, how to filter the input.. |
06:53.07 | scanline | captain_proton: no, this one is all floating point |
06:53.19 | captain_proton | floating point neural net? huh? |
06:53.20 | njs | (I had to write a backgrammon player in lisp for my AI class, and they expected us to do some really lame and boring strategic evaluater, which was, well, lame and boring, so I wrote what should be a ~world class player instead.) |
06:53.41 | njs | captain_proton: neural nets are usually floating point (unlike neurons, but eh.) |
06:54.14 | njs | (I had to get it done a bit early, since the training took all my cpu for ~4 days.) |
06:54.17 | scanline | This is just a python module I found that implements a 3-layer back propagation neural net.. looks pretty simple |
06:54.48 | scanline | though I should really either accelerate it with Numeric, or write the neural net as a C++ module |
06:57.41 | scanline | neat, it worked |
06:57.59 | scanline | sort of |
06:58.04 | njs | "worked" == "actually gives correct output"? |
06:58.09 | scanline | it's not linear, but that's to be expected giving it only two points to train with |
06:58.33 | scanline | "worked" == the number it gives increases when I move my hand toward the point I trained as "1" and decreases when I move to the point I trained as "0" |
06:58.45 | njs | cool |
06:59.07 | scanline | it's very nonlinear though |
06:59.47 | njs | well, if you train it to death with only 2 examples, it'll give all sorts of junk for other values :-) |
07:09.40 | scanline | it's odd that the error goes up during the first part of training, then gradually curves toward going down.. |
07:10.52 | captain_proton | scanline: not really |
07:11.29 | captain_proton | initially the modifications its making are big, so changing one thing changes the entire system - when its been trained a lot, the modifications it makes are small and localized |
07:11.51 | scanline | An input device with AI... wacky :) |
07:12.14 | njs | captain_proton: depends on your training method |
07:12.55 | captain_proton | njs: true, but with most standard back-propagation nets its how it works |
07:13.24 | captain_proton | scanline: is that 10uF cap on pin 9 of the 3T07 just a decoupling cap? |
07:13.41 | scanline | yeah, filters power supply noise |
07:14.01 | *** join/#picogui Tal-DCed (~talez@talez.arach.net.au) |
07:14.07 | scanline | the larger capacitors on the side of the chip near the photodiode might be for a filter, or maybe they just set the amp gain |
07:14.14 | scanline | ooh, mst3k |
07:14.23 | lurgyman | "The Day the Earth Froze" |
07:14.26 | njs | captain_proton: eh, not the ones I've seen, though it's true that adjusting your training parameter down as you train can be a nice optimization. |
07:14.26 | scanline | something to watch while training neural nets and procrastinating homework :P |
07:14.30 | lurgyman | "The Invasion of the Neptune Men" |
07:14.41 | lurgyman | "The Amazing Transparent Man" |
07:15.40 | scanline | whoa! |
07:15.55 | scanline | with psyco this neural net goes way super faster |
07:16.09 | scanline | JIT for python... spiffy |
07:16.32 | captain_proton | scanline: hm, that 1uf cap and 22 ohm resistor look like a low-pass on the photodiode signal? |
07:16.43 | scanline | could be, but those are screwy values to use for a filter |
07:16.47 | njs | *tsk tsk* it's not JIT, exactly. |
07:17.01 | scanline | well, yeah |
07:17.11 | scanline | it compiles when you run the program rather than when the function is run |
07:17.58 | njs | a conventional JIT is just a normal compiler embedded in your runtime environment, which is sorta silly when you think about it. |
07:23.05 | captain_proton | distinctly un-silly |
07:24.00 | captain_proton | scanline: how easy would it be to add a capacitor in there to just widen up the pulse? ;) |
07:24.40 | scanline | not sure where to put it |
07:25.30 | scanline | we could try just putting a capacitor in parallel with the photodiode maybe... |
07:25.43 | scanline | it's really the only place we can be sure it's before the filtering takes place |
07:26.18 | captain_proton | indeed |
07:26.22 | scanline | but the value would have to be just right, or we'd mess it up completely by smoothing out flashes enough that they wouldn't pass the filter |
07:26.46 | captain_proton | well, we know that your flasher works |
07:26.55 | scanline | uh huh |
07:27.03 | captain_proton | and your monitor is at what, 85Hz? |
07:27.33 | scanline | well, I don't think the refresh rate is the issue |
07:28.02 | captain_proton | ah true, we'd need to measure the phosphor blinkiness |
07:28.04 | scanline | on the scope we could see small pulses at about where the monitor refresh should be, but the shape of the wave would indicate that the sensor is a good bit faster than the monitor's phosphors |
07:29.52 | scanline | hehe, I was just thinking about TEMPEST and such the other day :) |
07:30.43 | invader_zim | hahahahha |
07:35.02 | captain_proton | =) |
07:35.08 | lurgyman | yay! |
07:35.36 | lurgyman | wow |
07:35.38 | scanline | bah, these files aren't world-read! |
07:35.39 | lurgyman | lurgyman@llama /mnt/navi/media/series> du -sh 12:32AM |
07:35.39 | lurgyman | 119G . |
07:36.25 | scanline | lurgyman: big numbers :) |
07:36.32 | lurgyman | there should be a way of having a permissions mask on a directory... which supercedes that of its subdirectory |
07:36.52 | lurgyman | llamaman, llamaman, he can do anything a llama can |
07:36.59 | g-i-r | rofl |
07:37.04 | captain_proton | scanline: hm, what would the possibility be of building a new light sensor circuit? |
07:37.18 | scanline | hehe, I forgot how funny the mst3k theme song was |
07:37.44 | scanline | captain_proton: I'd rather make nintendo's preamp work for what we need than build a completely new one :) |
07:38.58 | captain_proton | scanline: me too :P |
07:42.40 | scanline | "put that in your llama and smell it" |
07:42.51 | scanline | this mst3k has almost enough llama content |
07:43.30 | invader_zim | good night all |
07:43.37 | *** part/#picogui invader_zim (~see_it_ch@aden2-135-dhcp.resnet.Colorado.EDU) |
08:03.02 | captain_proton | showertime |
08:11.02 | hardwire | scanline: good free PIC C compiler? |
08:11.36 | scanline | for small values of "good", there's sdcc |
08:11.51 | hardwire | whats the most idiot proof one :) |
08:11.53 | scanline | it mostly works, though it was designed for 8051 and the PIC port is a still a bit buggy |
08:12.17 | scanline | sdcc is the only open source one.. there are a few that run under linux but aren't open.. I think there's a shareware one called c2c |
08:13.07 | hardwire | danke |
08:13.42 | scanline | there's also a "lite" version of the hitech C compiler.. it's sort of an industry standard, but it's windows-only and the 'normal' version of it is pretty expensive |
08:14.50 | hardwire | A friend was asking advice on lunix ones |
08:15.20 | scanline | I've used C for a couple projects.. but honestly I don't see much advantage over assemblu |
08:15.24 | scanline | err, assembly |
08:15.35 | hardwire | I do |
08:15.36 | hardwire | 1 |
08:15.38 | scanline | the PIC is really easy to program in assembly, and C doesn't really map well to it's architecture |
08:15.41 | hardwire | I can't grok assembly |
08:15.45 | hardwire | hehe |
08:16.01 | scanline | I mean, it's RISC- only 33 instructions I think' |
08:16.07 | hardwire | oh really? |
08:16.13 | scanline | yeah |
08:16.30 | hardwire | hmm |
08:16.44 | hardwire | I am reallly newb to asm |
08:16.52 | hardwire | I would be.. HEY.. GIMME A FOR LOOP DAMNIT! |
08:17.00 | scanline | well, pic asm is one of the easiest I've seen to learn, even if it's a bit funky |
08:17.07 | scanline | well.. here's a simple loop: |
08:17.10 | scanline | movlw 10 |
08:17.14 | scanline | movwf foo |
08:17.20 | scanline | myLoop: |
08:17.22 | scanline | ... |
08:17.27 | scanline | decfsz foo |
08:17.31 | scanline | goto myLoop |
08:17.45 | hardwire | rawr? |
08:17.57 | scanline | the movlw and movwf instructions do "foo = 10" |
08:18.13 | hardwire | whats lw stand for? |
08:18.15 | scanline | decfsz == "decrement file register and skip the next instruction if it's zero" |
08:18.28 | scanline | movlw == "move literal to the 'w' register" |
08:18.38 | hardwire | ookay |
08:18.58 | scanline | the pic has a 'w' register that's used like an accumulator, and the rest of the registers are in the same address space as the general-purpose RAM |
08:19.03 | hardwire | the asm compiler groks foo? |
08:19.03 | *** join/#picogui darth_mall (~silme@12-253-83-21.client.attbi.com) |
08:19.13 | scanline | well, I'm assuming 'foo' is a variable |
08:19.18 | hardwire | neet |
08:19.23 | scanline | you'd assign that to a memory address somehow beforehand |
08:19.32 | hardwire | yeh |
08:19.43 | hardwire | hand malloc basically? |
08:20.06 | scanline | you almost never need dynamic memory allocation |
08:20.13 | hardwire | err.. I understand |
08:20.17 | hardwire | but you map it all out by hand |
08:20.20 | hardwire | ? |
08:20.27 | scanline | well, there are assembler directives that make it easier |
08:20.32 | hardwire | ok |
08:20.42 | scanline | like 'cblock', which lets you specify a starting address then lays out all the variables you give it after that address |
08:20.46 | hardwire | push a variable of x length into the memory stack |
08:21.05 | scanline | it's not really a stack.. just a block of memory |
08:21.10 | scanline | the PIC has very little RAM |
08:21.20 | scanline | the larger ones have around 300 bytes of RAM |
08:21.23 | hardwire | can we use poo.. just for this conversation |
08:21.23 | hardwire | memory poo |
08:21.25 | hardwire | and.. can the io be a Taco Beam? |
08:21.33 | scanline | :) |
08:21.45 | scanline | so all the I/O is mapped into a register |
08:21.53 | hardwire | yeh.. I remember that part |
08:22.22 | scanline | If you had a taco beam on pin RB5 for example, you could use "bsf PORTB, 5" to turn it on |
08:22.22 | hardwire | just.. in C |
08:23.08 | hardwire | PORTB is the return data? |
08:23.12 | hardwire | err |
08:23.12 | hardwire | crap |
08:23.14 | hardwire | sorrt |
08:23.16 | hardwire | turn on is out |
08:23.28 | scanline | bsf == "bit set in file register" |
08:23.28 | hardwire | PORTB is in this case "1"? |
08:23.39 | scanline | PORTB is a byte with 8 I/O pins |
08:24.09 | hardwire | bsf == 1? |
08:24.27 | scanline | yep, and there's a bcf instruction to clear the bit to 0 |
08:24.32 | hardwire | PORTB is the memory location for the 8 bits of I/O. and the 5 declared which bit to change? |
08:24.36 | hardwire | err |
08:24.37 | scanline | yep |
08:24.39 | hardwire | I/O location |
08:25.00 | hardwire | so converting PORTB to an INT is done how? |
08:25.13 | scanline | eh? |
08:25.54 | hardwire | returning the status of all the pins on PORTB as an INT.. to say the serial port |
08:26.02 | hardwire | well |
08:26.03 | hardwire | Char |
08:26.20 | scanline | PORTB is a byte.. you could separate each bit out into a separate byte if you wanted |
08:26.33 | hardwire | err.. sorry :) |
08:26.41 | scanline | a few PICs have built in UARTs, but on most of them you have to bit-bang the serial data yourself |
08:26.46 | hardwire | basically.. how do you return the byte that is in PORTB |
08:26.47 | hardwire | heheh |
08:27.07 | hardwire | memory-mapped TACO BEAMS! |
08:27.12 | scanline | :) |
08:27.20 | captain_proton | sorry |
08:27.42 | scanline | dang this neural net is taking a long time to train... |
08:27.55 | hardwire | scanline for what? |
08:28.08 | scanline | trying to use a neural net to analyze sensor readings from my field sensor box |
08:28.11 | captain_proton | hardwire: input devices with ARTIFICIAL INTELLIGENCE! BWAHAHAHAHAHA |
08:28.29 | hardwire | scanline: ok |
08:28.39 | scanline | if this works, it should be able to report horizontal position, assuming the other axes stay constant |
08:29.39 | hardwire | will you be able to plot the similar paths the neural net is using.. and form a somewhat faster system for reading it? |
08:30.18 | scanline | this net is pretty small, so it shouldn't be a speed problem at runtime.. training is slow though |
08:30.37 | hardwire | gI don't have experience with neural net programming.. or even theory |
08:30.40 | scanline | 25 neurons :) |
08:30.48 | hardwire | I did have fun playing with RF triangulation |
08:30.49 | scanline | this is the first I've played with neural nets |
08:30.51 | hardwire | that was fun |
08:30.55 | hardwire | found a cheap way to do it |
08:31.13 | hardwire | took samples of a beacon from three points around town |
08:31.40 | hardwire | lined up the samples with the iteration of the beacon and onverted the sounds to 16bit FFT bitmaps |
08:32.05 | hardwire | normalized the FFT's after doing noise reduction |
08:32.20 | hardwire | then basically did a histogram of that.. and got what I wanted.. |
08:32.24 | hardwire | ame within 150 feet |
08:32.55 | hardwire | I doubt the second test will follow the blind luck that one gave me |
08:33.34 | scanline | neat |
08:33.52 | hardwire | soon my cop tracker system will be working :) |
08:34.05 | scanline | haha |
08:34.13 | scanline | tracking them by their CB radios? |
08:34.21 | hardwire | the VDT's in the cars |
08:34.34 | scanline | vehicle data transmitter? |
08:34.58 | hardwire | they are using an encrpyted VDT system now.. but it posts the callsign before each packet set. like regular packet radio.. unencrypted |
08:35.06 | hardwire | Video Display Terminals? |
08:35.10 | scanline | ah |
08:35.17 | hardwire | its packet radio |
08:35.24 | hardwire | just.. now the packets are RF encypted |
08:35.29 | hardwire | but not the callsign |
08:35.45 | hardwire | so I can decode the callsign.. and measure the quality of each packet from 3 sources |
08:36.25 | hardwire | then just upload the data to state and nationwide APRS Servers with a narow relay. and anybody watching internet APRS servers will see "TheMan-CB8970" pop up on their screen |
08:36.43 | hardwire | and then repeat the APRS packets to the APRS RF gateways here in town |
08:37.01 | hardwire | then anybody listening to APRS with their packet radios in their cars.. will know if there is a cop ahead |
08:37.04 | hardwire | WAY WAY ahead |
08:37.28 | hardwire | its my civil duty to do this |
08:37.35 | scanline | it's amazing the length people go to, to avoid speeding tickets :) |
08:37.45 | hardwire | bah.. I don't even have a car |
08:38.09 | hardwire | the good thing about it is.. I can then pick up the bus's as well :) |
08:38.14 | hardwire | and UPS |
08:38.15 | hardwire | FEDEX |
08:38.19 | hardwire | and everybody else that uses a radio |
08:38.38 | scanline | where are you putting the receivers? |
08:38.44 | hardwire | just start training something that will learn when to switch channels |
08:38.50 | hardwire | scanline: anywhere people wanna be a part of it |
08:39.45 | hardwire | it just sends the waveforms to a master node on the network.. then that masternode will consolidate and analyze the data and send it off to a main server node.. |
08:40.30 | hardwire | so.. each radio listener node just crops and tags packets to the atomic clock |
08:40.38 | hardwire | sends em off to the master nodes |
08:40.58 | hardwire | tis a dream I have |
08:41.18 | hardwire | when you think about it, you are gonna learn alot about te police in your area |
08:41.24 | hardwire | like where those awesome donut stops are |
08:41.38 | scanline | hmm |
08:41.40 | hardwire | how many times one house can get visited by cops in one day |
08:41.41 | scanline | how much hardware does it take to do this? |
08:41.48 | hardwire | bout 60 bucks for a scanner |
08:42.05 | hardwire | the soundmodem needed to decode the CW from the packets is based on the sound card |
08:42.19 | hardwire | so.. a 200 somewhad mhz system |
08:42.24 | hardwire | at each listener node |
08:42.38 | scanline | I've had a little interest recently in building a really super cheap LF radio transmitter |
08:42.49 | hardwire | then possibly making custom RS232 hardware and a quick interface to the keypad on the scanners |
08:42.55 | scanline | in the 160khz band you can do modulation in software on a cheap microcontroller |
08:43.05 | hardwire | really wide modulations |
08:43.08 | hardwire | PSK31 |
08:43.23 | hardwire | apt-get install soundmodem :) |
08:43.53 | hardwire | using PSK31 modulation at that low of frequency.. you should expect 200-300bps |
08:45.10 | scanline | I've heard of people transmitting morse code over a 160khz link, using dot times of around 30 seconds :) |
08:45.21 | hardwire | yeesh |
08:45.31 | hardwire | well crap.. whats the range for shortwave? |
08:45.54 | scanline | shortwave is a lot better of course, but you can't modulate it in software with supercheap hardware |
08:46.13 | hardwire | modulate packet data.. or the frequency sweep? |
08:46.18 | scanline | everything |
08:46.21 | hardwire | ah |
08:46.28 | scanline | generate the RF signal entirely in software |
08:46.43 | scanline | it's possible on a PIC pretty easily for such low frequencies :) |
08:46.51 | hardwire | wellif you can make a shortwave xmitter.. you can definatly finr a way to take soundcard out to it.. and use PSK31 |
08:47.05 | hardwire | PSK31 is used over shortwave for sending packet data around the world |
08:47.10 | scanline | well yeah, but it's not as fun as doing the transmitter itself in software :) |
08:47.15 | hardwire | its also used on the shuttle missions |
08:47.23 | hardwire | and its on the International Space Station |
08:47.27 | hardwire | sheesh.. so is SSTV |
08:47.31 | hardwire | played with that for a while |
08:47.32 | hardwire | scary stuff |
08:48.04 | hardwire | scanline: you wantLF radio transmitters for what again? |
08:48.17 | scanline | no reason :) |
08:48.20 | scanline | just pondering |
08:48.22 | hardwire | haha |
08:48.56 | hardwire | how did you dig into working with a neural net? |
08:49.38 | scanline | thought to myself "Hmm, the inverse electrostatics problem for this will be a bitch.. how about I get the computer to solve it for me!" and I start playing with a python module for a backpropagation neural net |
08:49.59 | hardwire | I was thinking about getting a group together up here to start training a neural net for CCD Based object detection |
08:51.30 | hardwire | there are some neat CCD's being developed in a few uni's that send off an excitement map |
08:51.49 | hardwire | oooh.. this clusters of elements just got VERY excited.. WEE! |
08:52.17 | hardwire | it would be more than cool to start playing with one of those for research on some stuff I am doing for a company |
08:52.18 | scanline | how is that different than taking the difference between subsequent frames and doing a little filtering? |
08:52.42 | hardwire | scanline: it is the basic principle |
08:52.52 | hardwire | however lots of things are not all that exciting |
08:56.16 | hardwire | the network of elements is seperated into micro clusters of more elements.. that report to their neighbors.. one an element is excited it passes its whoop along to its neighbors.. who are also probably very excited about something.. the excitement gets weaker as it filters thruout the CCD array. the elements can also be completely unsatisfied with their current excitement level and make sharp edges in the excitement |
08:56.24 | hardwire | it can be implemented in software |
08:56.36 | hardwire | but the CCD's basically do most of the work for you |
08:56.48 | hardwire | and it seems to be refined more than just standard difference mapping |
08:56.55 | scanline | oh, that's neat |
08:56.57 | hardwire | to the point where it can easily draw a line around moving objects |
08:57.20 | hardwire | it can also distinguish the difference inbetween an object and a pan |
08:57.30 | hardwire | unless the object is the size of the CCD's focal path |
08:57.31 | hardwire | and it moves |
08:58.07 | hardwire | the element matrix is circular as well |
08:58.21 | hardwire | basically a really excited ripple in a pond |
08:59.10 | hardwire | from what the person that mentioned this too me said, the human eye acts on excitetory functions, more than color for object recognition |
08:59.26 | hardwire | and this is R and D to develop excitetory CCD's |
09:01.07 | hardwire | my human CCD's are getting kinda dry and need rest |
09:01.17 | scanline | hehe |
09:01.44 | hardwire | I love teaching people about linux and how to use it |
09:01.49 | hardwire | he is getting the grasp of it |
09:01.52 | hardwire | but the man needs xhost - |
09:02.50 | hardwire | my personal favorite is I have a rndom sine wave based r/g/b gamma value script |
09:03.07 | hardwire | 0.4 to 3.0 on each channels gamma |
09:03.08 | scanline | scary |
09:03.09 | hardwire | hehe |
09:03.11 | hardwire | its just evil |
09:03.18 | captain_proton | yay |
09:03.22 | captain_proton | my evil C# program works |
09:03.25 | scanline | that might even be more scary that lurgyman's program to mess with the console palette |
09:03.30 | hardwire | sleep well trusty taco beam |
09:03.36 | scanline | :) |
09:16.23 | captain_proton | scanline: you still have that firmware that made the 4 LEDs blink? |
09:17.04 | scanline | yeah |
09:17.19 | scanline | here, lemme just DCC you a tarball of what I have |
09:17.48 | captain_proton | just wondering - i'm not going to mess with it until tomorrow, but that seems like a good program to make sure that everything works |
09:17.58 | scanline | oh, ok |
09:18.06 | captain_proton | i can hook up a couple LEDs or maybe just use the logic probe |
09:19.08 | scanline | the 'fieldbox' directory has all the stuff I've done so far for this box.. 'blinkytest' under that has what you're looking for. I also have a modified version of the "xwisp" downloader utility |
09:20.45 | scanline | I gave this one 16 samples on a plane.. hopefully it will give me 2D position of my finger if it's touching the surface of the box |
09:21.04 | scanline | but it looks like this will take a really long time to train... |
09:22.27 | captain_proton | welp |
09:22.37 | captain_proton | my C# (uff) program is working swell, so... |
09:22.38 | captain_proton | bedtime! |
09:22.42 | scanline | good night |
11:29.36 | *** join/#picogui vlado (~vlado@193.68.201.82) |
11:43.44 | *** join/#picogui file (jwired@mctn1-1665.nb.aliant.net) |
12:58.04 | *** join/#picogui hunger (foobar@pD9ED6CA6.dip.t-dialin.net) |
12:58.20 | hunger | Is there one in CVS? |
12:58.32 | scanline | The source and the compiler is in the 'themetools' directory in CVS |
12:58.55 | hunger | scanline: Yes, got that. I'm looking for something to put ofteh pgserver -t |
12:59.09 | hunger | s/often/after/ |
12:59.16 | scanline | you run themec on a .ths file and it produces the .th file you use with pgserver |
12:59.53 | scanline | the 'photonic' theme is nice, though not quite complete |
13:00.18 | scanline | you can also find compiled themes at http://old.picogui.org/themes/ |
13:01.03 | vlado | how to use ttf fonts in picogui ? |
13:01.14 | hunger | scanline: Thanks! |
13:01.43 | hunger | scanline: I'm gonna do a presentation on "cool GUI systems" and I want to show picoGUI. |
13:02.24 | hunger | scanline: Of yourse it will be very fresco-centered:-) |
13:02.57 | scanline | neat |
13:03.10 | scanline | vlado: turn on a font engine that supports them, probably the 'freetype' engine |
13:03.22 | hunger | scanline: But the motivation is to raise the awareness that X is not the technology we want for the next 20 years;-) |
13:03.34 | scanline | vlado: you'll need to give it a path for the fonts, and specify that font engine if another one is the default- that info is in README.configfile |
13:03.43 | scanline | hunger: cool! good luck |
13:03.52 | vlado | i tried .pgserverrc with "[font-freetype] ...." but the pc gets freezed when run pgserver |
13:04.22 | scanline | vlado: what do you put after [font-freetype]? that's the important part :) |
13:04.58 | scanline | vlado: and just to make sure, you're compiling pgserver with the freetype engine enabled in the menuconfig? |
13:05.04 | hunger | scanline: I probably won't to picogui justice though. I know too little about it. |
13:05.10 | vlado | path = /usr/share/fonts/picogui |
13:05.19 | scanline | hunger: well, awareness is better than nothing |
13:05.41 | vlado | scan_list = /usr/share/fonts/picogui/font.index |
13:05.51 | scanline | vlado: there could be a problem with your freetype installation? |
13:06.20 | hunger | scanline: I want to mention that it shares ideas with Fresco, but tends to favour speed over concepts (embedded systems!) and the "microkernel"-approach for the next version. |
13:07.12 | vlado | scanline: hm... dunno |
13:07.16 | scanline | hunger: the current implementation does favor speed over concepts, but it's not what I like to emphasize ;) |
13:07.46 | vlado | scanline: well, can i use non-ttf unicode fonts then? |
13:08.00 | hunger | scanline: Well, there needs to be something fresco is better at for my presentation;-) |
13:08.38 | scanline | vlado: sure. The old engine uses BDF files, but it's a bit inconvenient since you have to have it compile the font into your pgserver binary. It would be much nicer if you could get freetype working |
13:08.47 | scanline | hunger: :) |
13:10.20 | vlado | scanline: yes... but it doesn't work for me :( |
13:10.35 | scanline | can you try using strace to see where it's hanging? |
13:12.32 | hunger | scanline: Your buildsystem depends heavily on /bin/sh being a link to /bin/bash by the way. |
13:13.05 | scanline | hunger: Hmm, I thought that had been fixed... oh well, I really hate the current build system |
13:13.07 | hunger | scanline: It might be better to replace the SHELL-vars in your Makefiles with /bin/bash instead of /bin/sh |
13:13.24 | scanline | I think I fixed it at some point then someone else had a reason to put it back |
13:13.35 | scanline | the new build system will have no dependencies other than pythong :) |
13:13.43 | scanline | err, python |
13:14.06 | hunger | scanline: I thought I hab /bin/sh point to /bin/bash... so I tried a while to build picogui yesterday night till I found out that debian chenged sh to be dash. |
13:14.16 | scanline | dash? |
13:14.40 | hunger | Dunno what that is... maybe debian specific? |
13:14.46 | scanline | guess so |
13:14.52 | hunger | It's supposed to be more POSIXy then bash... |
13:32.35 | vlado | haaaaaaaa got it :))) |
13:32.56 | vlado | where is set the default font ? |
13:32.57 | scanline | vlado: neat, what was the problem? |
13:33.16 | scanline | it's one of the config options in the font-freetype section, look in README.configfile |
13:34.28 | hunger | How do I make windows overlap? |
13:34.35 | hunger | I read that's possible now? |
13:34.55 | scanline | picogui doesn't support it yet, but you can use rootless mode to put each picogui root widget in a separate X window |
13:35.33 | vlado | scanline: dunno :) i didn't understand what i did to get it working |
13:35.36 | hunger | scanline: Great! |
13:37.16 | vlado | scanline: where is set the default font ? in the theme ? |
13:37.52 | scanline | vlado: the actual default font is set by the font engine- in this case it's an option in the font-freetype config section. However, the theme may override it. |
14:01.50 | vlado | how to change keyboard mapping ? |
14:02.00 | vlado | is there a way ? |
14:02.09 | scanline | keyboard mapping for what? |
14:04.32 | hunger | scanline: Your new site looks really nice. But the old one contained way more information. |
14:05.49 | vlado | i want cyrillic keyboard |
14:06.01 | scanline | hunger: the new site did have pretty much all the information the old one did, but we had problems with the code for the links and screenshot boxes using too much CPU, so they're temporarily disabled.. it's all in the newest news item |
14:06.28 | scanline | vlado: that depends on the input driver. If you're running under X, it's X's responsibility to do that. If you're using the rawttykb driver on the console, it should copy the kernel's keymapping |
14:07.42 | *** join/#picogui Dire (~Admin@100.8.33.65.cfl.rr.com) |
14:07.53 | hunger | scanline: Ah! That explains it then. |
14:08.15 | scanline | hunger: I don't know enough about Plone to fix them.. lalo's going to do it, but he's been busy |
14:08.22 | hunger | scanline: Now that I found the old site I'm more then happy again anyway:-) |
14:09.50 | scanline | :) |
14:10.04 | scanline | I haven't worked on picogui the last several weeks though.. been distracted by many things |
14:11.05 | scanline | I just opened a fortune cookie that reads "You love Chinese food" |
14:11.42 | hunger | scanline: lol |
14:12.13 | hunger | scanline: You will like chinese food after reading a couple of hundred of those messages;-) |
14:41.59 | *** join/#picogui carpman{mayuka} (~djackson@dsl081-122-066.dfw1.dsl.speakeasy.net) |
14:42.20 | *** join/#picogui PuckCh (~bornet@adsl-62-167-184-157.adslplus.ch) |
14:54.58 | *** join/#picogui balor (~balor@pgh17.cs.may.ie) |
15:02.47 | *** join/#picogui PuckCh (~bornet@adsl-62-167-193-142.adslplus.ch) |
15:03.04 | *** join/#picogui invader_zim (~see_it_ch@aden2-135-dhcp.resnet.Colorado.EDU) |
15:08.15 | *** join/#picogui fjkldas (~Admin@100.8.33.65.cfl.rr.com) |
15:41.05 | *** join/#picogui decaff (~khedspet@216-187-196-82.dsl.btitelecom.net) |
15:43.02 | *** part/#picogui hunger (foobar@pD9ED6CA6.dip.t-dialin.net) |
15:43.44 | *** part/#picogui vlado (~vlado@193.68.201.82) |
15:51.58 | *** join/#picogui bigthor (~bigthor@80.Red-80-32-133.pooles.rima-tde.net) |
15:53.15 | invader_aim | invader_zim |
15:54.09 | invader_zim | g-i-r: you should wake up!!!! it's already 9 in the morning!!!! |
16:16.42 | carpman{mayuka} | invader_zim: must have stayed up to late watching static. |
16:16.47 | carpman{mayuka} | s/to/too |
16:17.06 | *** join/#picogui bigthor (~bigthor@80.Red-80-32-133.pooles.rima-tde.net) |
16:17.46 | *** join/#picogui lalo[out] (~lalo@200.203.12.232) |
16:18.02 | *** join/#picogui prpplague (~JoeBob1@12.148.134.9) |
16:59.45 | invader_zim | carpman{mayuka}: nope, been studying all night....but i did watch ZIM!!!!! HHAHAHAHAHAH |
16:59.52 | invader_zim | Dire|class: hello |
17:10.23 | carpman{mayuka} | Hi floor! Make me a sammich |
17:10.38 | floor | carpman{mayuka}: poof, you're a sammich |
17:10.52 | *** join/#picogui bigthor (~bigthor@80.Red-80-32-133.pooles.rima-tde.net) |
17:27.08 | hardwire | wheres dougy! |
17:27.11 | hardwire | boohoo! |
17:35.36 | *** join/#picogui decaff (~khedspet@216-187-196-82.dsl.btitelecom.net) |
17:36.04 | *** join/#picogui bigthor (~bigthor@80.Red-80-32-133.pooles.rima-tde.net) |
18:07.32 | *** join/#picogui darth_iBook (~silme@wireless-163.cs.colorado.edu) |
18:13.52 | darth_iBook | hello Dire |
18:15.17 | lurgyman | hello darth_iBook |
18:15.25 | Dire | howdy |
18:16.26 | darth_iBook | wanna eat at noon ish? |
18:16.39 | captain_proton | hardwire: dougy? wrong channel |
18:16.51 | hardwire | augh |
18:16.53 | hardwire | you are so right |
18:17.02 | captain_proton | #picogui's doug is file |
18:17.31 | Dire | mr rogers died :( |
18:17.52 | darth_iBook | that's sad |
18:31.36 | darth_iBook | lurgyman captain_proton scanline & invader_zim: lunch in about 1/2 hour? |
18:31.45 | captain_proton | sure, why not |
18:32.19 | darth_iBook | ^_^ |
18:32.31 | *** join/#picogui fragglet (~fraggle@tuneless.plus.com) |
18:34.06 | captain_proton | darth_iBook: sooner perhaps? |
18:37.28 | captain_proton | so leave |
18:37.40 | darth_iBook | oki, one sec |
18:37.47 | captain_proton | :-) |
18:39.41 | Dire|out | goin' to see Old School, hope it's funny :p |
18:41.09 | darth_iBook | oki, /me is on his way to aden |
18:41.13 | darth_iBook | will call from lobby |
19:14.57 | *** join/#picogui prae (~prae@sherpadown.net) |
19:18.33 | *** join/#picogui lime (~nope222@public1-stre1-6-cust37.bagu.broadband.ntl.com) |
19:22.07 | darth_llama | mwahahahaha |
19:22.24 | lime | hi, looking for lonetech, anyone seen him? |
19:22.35 | darth_llama | PicoBot: seen lonetech |
19:22.36 | PicoBot | lonetech was last seen on #picogui 50 days, 12 hours, 1 minutes and 40 seconds ago, saying: while full modules, such as SIMMs, DIMMs, and RIMMs, are measured in kibibytes or mebibytes. [Tue Jan 7 18:29:59 2003] |
19:29.52 | lime | anyone had any experience with case mods on old Indy's ? |
19:31.49 | carpman{mayuka} | case mods on an Indy? Why, they look cool already. |
19:55.18 | *** join/#picogui file (jwired@mctn1-2576.nb.aliant.net) |
20:11.15 | *** join/#picogui darth_iBook (~silme@aden2-18-dhcp.resnet.Colorado.EDU) |
20:11.34 | gonk{work} | hi darth_ |
20:11.36 | gonk{work} | hi darth_iBook |
20:11.40 | darth_iBook | hi gonk{work} |
20:12.07 | gonk{work} | whats up? |
20:13.49 | darth_iBook | just chillin in david and dan's room doin some coding |
20:14.20 | darth_iBook | watchin david squish cans |
20:19.52 | carpman{mayuka} | hello gonk{work} |
20:20.36 | gonk{work} | hey carpman{mayuka} |
20:20.49 | gonk{work} | carpman{mayuka}: we have two dead/dying machines here :) |
20:21.38 | carpman{mayuka} | what kind of machines? |
20:22.17 | *** join/#picogui Dire|out (~Admin@100.8.33.65.cfl.rr.com) |
20:22.32 | gonk{work} | carpman{mayuka}: PCs |
20:22.38 | carpman{mayuka} | BAH |
20:22.43 | carpman{mayuka} | I got one of these... |
20:23.03 | carpman{mayuka} | http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=3403156098&category=11214 |
20:24.43 | *** join/#picogui captain_proton (~jupiter@aden2-23-dhcp.resnet.Colorado.EDU) |
20:27.42 | *** join/#picogui file (jwired@mctn1-2576.nb.aliant.net) |
20:28.26 | carpman{mayuka} | ...? |
20:28.34 | carpman{mayuka} | file: eh? |
20:28.48 | file | carpman{mayuka}: king of an ancient greece society |
20:29.43 | carpman{mayuka} | file: eh. |
20:29.55 | file | oh well |
20:29.56 | carpman{mayuka} | Eh. |
20:30.00 | file | carpman{mayuka}: eh eh |
20:30.10 | carpman{mayuka} | toque eh. |
20:30.36 | gonk{work} | PicoBot: emulate lur |
20:30.37 | PicoBot | gonk{work}: what? |
20:30.39 | gonk{work} | PicoBot: emulate lurgyman |
20:30.52 | carpman{mayuka} | gonk{work}: emulate carpman |
20:31.00 | carpman{mayuka} | err... |
20:31.06 | carpman{mayuka} | PicoBot: emulate carpman |
20:31.09 | carpman{mayuka} | :) |
20:31.14 | gonk{work} | PicoBot: emulate lurgyman is also flips a pencil until his music theory homework is done |
20:31.14 | PicoBot | okay, gonk{work}. |
20:31.30 | file | the USS Enterforaprize! |
20:32.00 | *** join/#picogui Xent (jchu@amga45cyy30x9.bc.hsia.telus.net) |
20:32.07 | file | hi Xentac |
20:32.46 | Xent | hi file |
20:32.51 | space-commander- | ack! |
20:33.06 | space-cmndr-trav | arg |
20:33.15 | file | Xent: how are you? |
20:33.36 | carpman{mayuka} | Xent: what have you done with ac!?!?!?! |
20:33.53 | Xent | carpman{mayuka}: he's still on my desktop... |
20:33.57 | carpman{mayuka} | oh, ok. |
20:35.16 | darth_iBook | there |
20:35.18 | darth_iBook | much better |
20:37.13 | *** join/#picogui darth_llama (~lurgyman@aden2-123-dhcp.resnet.Colorado.EDU) |
20:37.18 | gonk{work} | oh |
20:37.26 | Xent | oh? |
20:37.29 | gonk{work} | darth_iBook: did I tell you what they replaced on my iBook |
20:38.20 | darth_iBook | gonk{work}: cool |
20:38.28 | darth_iBook | so it works now? |
20:39.17 | gonk{work} | darth_iBook: indeed :) |
20:43.06 | carpman{mayuka} | Xent: esperanto... so you'll be able to talk to William Shatner in his native language... |
20:48.22 | captain_proton | gonk{work}: whenwhere? |
20:50.05 | gonk{work} | captain_proton: when my friend is working, and at the sushi bar right by the market street station called Sonodas |
20:50.29 | captain_proton | hmmmm |
20:50.30 | captain_proton | =) |
20:51.27 | Xent | carpman{mayuka}: that's right... |
20:52.28 | gonk{work} | captain_proton: take the prices, and divide by 10 :) |
20:59.07 | darth_iBook | lala cheese |
21:15.44 | darth_iBook | right |
21:42.02 | *** join/#picogui file (~joshnet@mctn1-0574.nb.aliant.net) |
22:03.45 | *** join/#picogui file (~joshnet@mctn1-0574.nb.aliant.net) |
22:05.43 | invader_zim | it's the end of the world as we know it |
22:05.50 | invader_zim | it's the end of the world as we know it |
22:05.58 | invader_zim | it's the end of the world as we know it |
22:06.02 | invader_zim | and i feel fine... |
22:06.26 | invader_zim | THEY ARE FIXING MY DOOR!!!! |
22:06.34 | invader_zim | captain_proton: why would you scare me off? |
23:13.33 | darth_llama | brrrr |
23:13.59 | darth_mall | darth_llama: you still haven't changed your nick back? ^_^ |
23:14.12 | darth_llama | i am invader_zim!!! ZIm is is my name |
23:14.17 | darth_mall | ah |
23:14.23 | invader__zim | heheh |
23:15.40 | invader__zim | darth_mall: you need a mullet |
23:16.14 | invader__zim | you should borrow micah |
23:16.15 | invader__zim | 's |
23:16.43 | invader__zim | gir: I am bored |
23:16.45 | invader__zim | amuse me |
23:17.35 | darth_mall | gir is not here |
23:17.56 | lurgyman | muffins |
23:18.02 | darth_mall | ooh |
23:18.28 | lurgyman | suggest amusement |
23:18.50 | darth_mall | stand on your head |
23:18.51 | lurgyman | nm...we have amusement |
23:19.01 | lurgyman | we are going to go build a card house |
23:23.30 | *** join/#picogui Soopaman (~soopaman@h24-66-55-163.wp.shawcable.net) |
23:29.23 | Soopaman | ping |
23:29.23 | Soopaman | pong |
23:38.29 | gonk{work} | w00t |
23:40.45 | Soopaman | whoa |
23:40.45 | Soopaman | where did you get that from/? |
23:41.14 | gonk{work} | Soopaman: notice the {work} in my gonk{work} nick |
23:41.25 | gonk{work} | hmm... |
23:42.10 | gonk{work} | wb Dire |
23:42.33 | Dire | 'lo |
23:42.57 | darth_mall | Ang-l|4d-til22: why 4 you grumbleded |
23:43.18 | darth_mall | hi gonk{work} |
23:43.33 | darth_mall | <GIR>look at me! I'm dancin like a monkey!!</GIR> |
23:45.36 | Dire | Old School was pretty funny :p |
23:46.10 | darth_mall | Dire: worth the $7 to see it |
23:46.12 | darth_mall | ? |
23:46.23 | Dire | If you like silly comedy's, for sure. |
23:46.31 | darth_mall | coo |
23:49.13 | gonk{work} | later everyone |
23:49.15 | gonk{work} | Ang-l|4d-til22: take care |
23:49.18 | gonk{work} | darth_mall: be well |
23:49.21 | gonk{work} | Dire: have fun |
23:49.22 | darth_mall | lates gonk{work} |
23:49.28 | darth_mall | ttyl |
23:49.29 | gonk{work} | Soopaman: for the love of God, behave |
23:49.30 | darth_mall | drive safe |
23:49.36 | darth_mall | roads am badish here at least |
23:49.41 | darth_mall | dunno bout at mineral and all |
23:49.44 | gonk{work} | ok |
23:49.54 | Dire | tata |
23:49.55 | darth_mall | doesn't take far |
23:57.35 | Soopaman | gonk{work}, god doesn;t like me |
23:57.35 | Soopaman | after our convo of merlin262 |
23:57.35 | Soopaman | :/ |