#qi-hardware IRC log for Saturday, 2012-03-10

wolfspraulxiangfu: we should also try to find out whether anyone is still using ks7010 or plans to use it00:31
wolfspraulI think it's time to retire this officially, it's totally dead-end.00:32
wolfspraulkyak had some, but I think all I heard from him about them was trouble :-)00:32
xiangfuks7010 ok. 00:49
wolfspraulxiangfu: let's see whether anyone still uses them00:52
wolfspraulmy point is that we may be wasting time on something that is entirely worthless00:52
wolfsprauland then I much rather spend the same time on wpan which has a future00:53
wolfspraulthe successor of the crazy spectec wifi solution is the equally crazy solution of backpacking the ben onto a MR11U wifi router :-)00:53
wolfspraulbut more practically, the Ben remains a device with only USB-networking, and an experimental ben-wpan00:54
kristianpaulnice, MR11U looks like portable workaround solution 01:09
wolfspraulyes and no, not to me01:10
wolfspraulI believe in easy to use products, something cheap and simple and just super easy to use.01:10
wolfsprauland that's not going in the right direction01:10
wolfspraulcan you or anyone hack it up? of course01:10
wolfspraullike 1000 other things01:10
kristianpaulyes i got you point01:11
wolfspraulbut I believe we can make integrated and easy to use products01:11
kristianpauland of course not the right direction01:11
kristianpaulfor us01:11
wolfspraulit's a long and hard path to get there though, but we are on it01:11
kristianpaulintegrated, yes good point !01:11
kristianpaulha even works as a battery01:12
Aylahi08:58
Aylalooks like LLVM/MIPS is progressing: http://lists.cs.uiuc.edu/pipermail/llvmdev/2012-March/048131.html08:58
kyakwolfspraul: yeah, i gave up on ks701009:16
wolfspraulkyak: yes, that's what I remembered and it make sense09:25
wolfspraulyou were probably the last holdout ks7010 user :-)09:25
wolfspraulthat's why I think we can officially retire ks7010 support to the code history, from where it will most likely never get resurrected...09:26
wolfspraulkyak: do you have atben & atusb?09:26
viricwhitequark: two years ago, xburst-tools built fine with libusb 0.1 :)10:16
viricwhitequark: Now that I supplied libusb 1.0.0, it tells me it wants libusb 0.1! :D11:18
virichttp://hydra.nixos.org/build/227515511:18
viricxiangfu: ^ any idea? What does xbursttools want, libusb 0.1 or libusb 1.0?11:18
viricor both?11:19
viric(sorry, I supplied libusb 1.0.8, not 1.0.0)11:20
lindi-viric: maybe it needs both?11:21
viricI'm supposing that xiangfu did not care about some detail, and made it depend on both11:22
lindi-jzboot needs 1.011:23
lindi-maybe usbboot needs 0.111:24
viricmh ok11:24
lindi-it's a different API11:24
viricI'll supply both then11:25
larscmaybe it's just time to get rid of usbboot and just provide a thin wrapper which maps usbboot commands to jzboot commands11:26
viricwell, that's xbursttools from december. I don't know if much changed.11:27
Aylaor just trash usbboot and force people to use jzboot :)11:27
viricI only know to use usbboot :)11:27
xiangfuviric, it's needs both.11:30
xiangfuviric, jzboot : libusb 1.011:30
larscAyla: you can't break backwards compatibility if you don't want to annoy people11:30
xiangfuviric, usbboot: libusb 0.111:30
viricok11:30
viricI hope I get it to build now11:30
viricthank you!11:30
larscxiangfu: is there anything in usbboot which jzboot can't do?11:30
Aylabooting linux? :p11:31
xiangfularsc, I don't know the detail. I think kyak do some test on jzboot11:31
larscok, i'll give it a try and come up with something thats usbboot compatible but uses the jzboot codebase11:32
Aylaperhaps a script is enough?11:33
viricthis week we updated libc, gcc, ... and all that at nixpkgs. I've still to test the up to date nanonixos.11:33
larscAyla: probably not, if you want to support the interactive mode11:36
larscand usbboot uses a different config format11:36
Aylaok11:37
xiangfularsc, just quit check the command subset mostly same. usbboot: nprog,  jzboot: nprogram and usbboot have two more gpio and reset command11:38
larscxiangfu: ok, sounds promising11:38
larscbut first some grocery shopping11:39
xiangfuand jzboot and usbboot both usb '-c' for run semicolon-separated commands.11:40
viricjob nixpkgs:trunk:xbursttools has11:41
viricchanged from 'Failed with non-zero exit code' to 'Success'11:41
viricperfect!11:41
whitequarkviric: libusb-1.0 ships with a shim12:02
whitequarkthat is, a "libusb-0.1" which actually uses 1.0, but is api-compatible with 0.112:02
viricnot our libusb112:02
whitequarker12:02
whitequarkthere's maybe a libusb-0.1-compat package or something like this12:03
viricI'll check12:03
viricmaybe that would be better than using the old libusb12:04
viricwhitequark: our build log: http://hydra.nixos.org/build/2189755/nixlog/1/raw12:04
kristianpaulwolfspraul: http://oshwgnss.blogspot.com/ ;)12:08
kristianpauland here are the sources eagle afaik http://www.sensorcomm.co.jp/osqzss/fpga/max2769/12:08
wolfsprauloh wow12:09
wolfspraulhow did you get this?12:09
kristianpauli'll ask takuji ebinuma for licensenses12:09
kristianpaulArtyom asked for some sources to him, i'll followed the thread ;-)12:09
wolfspraulgrea12:09
wolfspraulgreat12:10
wolfspraul:-)12:10
kristianpaulgood, pcb design is simple, hope not simpler ;)12:14
kristianpauls/design/layout12:15
wpwrakhmm, is "hope" used as a verb or a noun here ? :)12:47
kristianpaulverb12:49
kristianpaulwhy?12:49
wpwraki was pondering the philosophical implication of interpreting it as a noun :)12:49
kristianpauli can imagine :-)12:50
kristianpaulbetter not ;)12:50
wpwrakthe result of twisting my mind around logic where A < B and B > A can be true at the same time :)12:54
larscdepends on the definition of '<' and '>' ;)12:55
kristianpaullol12:55
wpwrakyup ;-)12:55
wpwrakA and B can be sets. A < B means that there is at least one pair of elements a, b for which a < b.12:56
wpwrakthe tricky bit is using such a logic for searching. and for detecting redundant conditions.12:56
larscare you using that for the flickernoise compiler?13:02
wpwrakoh dear, no :)13:03
wpwrakthat's for boom13:03
larscboom?13:04
wpwrakthe BOM processor. the thing that takes component descriptions from the schematics and turns them into a shopping list13:04
larschm, interesting13:06
kristianpaulargh, the mix og being and end inside multiple if else statements begin to get messier :S13:08
kristianpauls/og/of13:09
wpwrakah, more philosophy. now we approach the continuity of existence, to be or to end, this is the question ;)13:10
kristianpaulindeed ! :)13:11
wpwrakor maybe the choices aren't even mutually exclusive ? if we have a _mix_ of being and end ... where's schroedinger when you need them ? :)13:12
kristianpauls/being/begin13:12
wpwrakbah, now you're trivializing it13:13
larscbeging and end, the eternal circle of life13:14
larscbegining13:15
whitequarkDocScrutinizer: re GSM modules22:24
whitequarkI just discovered an AT+CRLP command which allows you to set radio link settings22:24
whitequarkInternetworking window size, Mobile window size, Ack timer T1, Retransmission attempts N2, RLP version, T422:25
whitequarkcan anything interesting be done with this?22:25
DocScrutinizerprobably yes22:44
DocScrutinizerhighly intriguing anyway22:44
whitequarkDocScrutinizer: also, I have read the changelog22:51
whitequark11.Solved the problem of DNS crash while using AT+CIPSHUT in decrypting process22:52
whitequarkafaik by "crash" they mean a buffer overflow in the module itself22:52
whitequark... which should be in privileged code22:52
DocScrutinizerLOL22:53
DocScrutinizerwhich gives you some ideas?22:53
whitequarkexactly.22:54
whitequarkcombined with the fact that I have SDK where I have endpoints to write to UART, etc...22:54
whitequarkthis is _very_ interesting22:54
whitequark(well, I don't really have it atm, but it won't be hard to get. a few letters maybe.)22:55
whitequarkto be honest, I'm quite certain that their software has zillions of buffer overflows waiting to be exploited anyway22:57
whitequarkgiven how it's written and what often appears in changelog22:58
tickratehey guys. I haven't been here in a while. Gotta a Questions for ya. I've been playing with uboot and messed some things up. Got it all back except one thing.23:11
tickrateI lost the failsafe on 2.6.38+ kernel. How do you get it back? Not important, but would like to know.23:11
--- Sun Mar 11 201200:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!