Back
[03:16:20] <rue_shop3> sometimes I'm worried I'm shopping with sleep deprevation
[03:16:50] <rue_shop3> I got 1Mhz resonators, 20pf caps, and 2.048Mhz crystals
[03:17:12] <rue_shop3> casue I just dug thru the shop for 3 hours to find a low freq crystal and failed.
[03:37:47] <deshipu> ah, the time crystals
[08:45:42] <veverak> DOC generated
[08:45:46] <veverak> so, i got formated code
[08:45:50] <veverak> I've got comments
[08:45:53] <veverak> I've got tests
[08:45:55] <veverak> I've got doc
[08:45:59] <veverak> hmm
[08:46:02] <veverak> one part done!
[08:46:04] <veverak> :)
[08:53:11] <veverak> damn it
[08:53:14] <veverak> ignored on #Python
[08:53:20] <veverak> guys, how do you commit doc in repo?
[08:53:25] <veverak> I have doc/ with source files
[08:53:36] <veverak> and wondering if I should commit only source
[08:53:39] <veverak> or also generated files
[12:54:08] <deshipu> do you have setup.py and did you register it on pypi?
[12:54:25] <deshipu> veverak: only source
[12:54:37] <veverak> it's not exactly pypi compatible
[12:54:39] <veverak> :D
[12:54:42] <deshipu> veverak: add the rest to gitignore
[12:54:49] <veverak> deshipu: I see
[12:55:07] <deshipu> veverak: you can make it build on readthedocs and point people there
[12:56:09] <veverak> deshipu: won't work
[12:56:15] <veverak> doc needs ROS env to build
[12:56:32] <veverak> but, ros got this solved it's own way, so I won't worry about that hopefully
[12:57:11] <veverak> anyway, need to work on the doc
[12:57:13] <veverak> it's messy
[12:57:15] <veverak> ;)
[12:59:14] <deshipu> fun
[12:59:38] <deshipu> I just spent the day working on 1 and 16-bit framebuffer implementations
[12:59:54] <deshipu> if I see one more bit shift and mask, I'm going to bite someone
[13:00:29] <veverak> :D
[13:07:10] <deshipu> I was supposed to try implementing a software uart...
[13:07:21] <deshipu> hmm, let's try it
[13:08:59] <veverak> :D
[13:15:34] <rue_house> maybe its a bad to machine wash something MADE OUT OF GEL
[13:38:05] <rue_house> I think I lost the finger of a 12' robot in my bedroom
[14:24:57] <deshipu> I won't even ask what it was doing there
[14:36:13] <veverak> well
[14:36:20] <veverak> doc: RTFS
[14:36:22] <veverak> is also way
[14:36:24] <veverak> :)
[14:46:29] <z64555> veverak: another codebase I work with has its documentation in an ignored subdir. This is so that people pulling from the repo don't have to load a sizable chunk of metadata that can be generated by a tool
[14:47:48] <veverak> makes sense
[14:47:55] <veverak> anyway
[14:48:02] <veverak> I LOVE config files in .py
[14:48:09] <deshipu> I don't
[14:48:20] <veverak> one can pull a lot of info from one source easily
[14:48:38] <deshipu> but I have to maintain deployments that are several years old that have python config files
[14:48:57] <deshipu> veverak: at some point your config file has more logic than your app
[14:50:34] <veverak> yep
[14:50:36] <veverak> :)
[14:58:35] <deshipu> and then you have to upgrade...
[15:09:57] <veverak> :D
[15:09:59] <veverak> well
[15:10:01] <veverak> one last thing
[15:10:10] <veverak> I have this dynamic reconfigure module used
[15:10:26] <veverak> thing is, that it creates python object, that got parameters definition in json structure
[15:10:39] <veverak> hmm
[15:10:42] <veverak> nt object, class!
[15:10:47] <veverak> so
[15:11:02] <veverak> have to figure out how to create sensible output from that json into sphinx
[15:15:24] <deshipu> just write it manually
[15:15:29] <veverak> big nope
[15:15:37] * veverak is freak on this things
[15:15:40] <veverak> hate to write doc
[15:15:48] <veverak> so I double hate something that can be generated
[15:16:25] <deshipu> see, it's your user API
[15:16:29] <deshipu> it shouldn't be generated
[15:16:35] <deshipu> it shouldn't ever change
[15:16:41] <veverak> it's not api
[15:16:46] <veverak> it's configurable parameters
[15:16:46] <deshipu> it is
[15:16:55] <veverak> and this part of doc is actually reported as "generated part"
[15:16:59] <deshipu> any configuration is part of the API
[15:19:36] <veverak> it's configuration of dynamic reconfigure module
[15:19:42] <veverak> conf itself is generated by python module
[15:19:47] <veverak> and than viewed in nice gui
[15:20:31] <veverak> https://github.com/Schpin/schpin-ros/blob/master/schpin_dev/cfg/PWMJoint.cfg it's not exactly friendly to check what can be configured
[15:20:39] <veverak> and I thought that it would be nice in doc
[15:20:49] <veverak> especially if I want to use this frequently
[16:39:08] <veverak> deshipu:
http://squirrel.veverak.org/ros/html/pwm_servo_joint.html#dynamic-params
[16:39:10] <veverak> got it there
[16:39:26] <veverak> but I will move that source autodoc elsewhere
[17:13:58] <rue_shop3> "no, Pikachu!, use lightnight bolt, not smash with hammer!!!, ARG, PIKACHU!!!!!"
[17:15:20] <rue_shop3> I think your wasting your time using ros, you have successfully, severly overcomplixified something too simple to put a minutes thought into
[17:16:48] <veverak> lol :)
[17:48:57] <rue_shop3> SWEEEEEEEEEEET I have a 16 channel analog -> pwm working
[17:50:06] <rue_shop3> I'm just some math away from an 8 channel loop controller
[17:50:37] <deshipu> while (;;) { AnalogWrite(AnalogRead())) }
[17:51:12] <rue_shop3> heh 10 bit adc * 16 channels, 12bit pwm * 16 channels @ 600Hz
[17:51:51] <rue_shop3> atmega 8
[17:53:16] <rue_shop3> I suppose the new avrs with the pin change interrupt can be used as adc's
[17:53:43] <rue_shop3> you reset the timer, snapshot the value when the interrupt trips
[17:53:44] <rue_shop3> huh
[17:54:34] <rue_shop3> I recon you could get 16 bits out of that too
[17:55:13] <rue_shop3> @244Hz
[17:55:33] <rue_shop3> no 122, it has to charge
[17:55:47] <veverak> well
[17:55:50] <veverak> one makes tests
[17:56:01] <veverak> one makes doc and changes ssome things, one needs to mod tests
[17:56:16] <rue_shop3> so, my tlc5940 library gets the channels backwards
[17:58:18] <veverak> so
[17:58:27] <veverak> deshipu: have you thought about leg-moving mechanism?
[17:58:31] <veverak> damn it
[17:58:42] <veverak> deshipu: about mechanims to detect if leg touched the ground
[17:58:56] * veverak was again writing faster than thinking
[17:59:58] <veverak> i was thinking that attaching simple smd-size pcb button at the end of leg would do the trick
[18:00:19] <veverak> than I just need a neat way how to detect one button is pressed from raspberry pi
[18:01:40] <veverak> second version is to go for properly supensioned legs and hall sensor
[18:01:41] <deshipu> veverak: normal switches are too hard to press
[18:01:51] <veverak> deshipu: my tote is little byt hevier
[18:01:53] <veverak> *heavier
[18:01:57] <deshipu> veverak: you need one of those switches for detecting if an sd card is inserted
[18:02:10] <deshipu> "detector switch"
[18:02:23] <deshipu> there are smd versions that are small and easy to attach to the legs
[18:03:35] <veverak> http://www.aliexpress.com/item/10-Pcs-Two-Way-Operation-SMD-Momentary-Detector-Switch-Black/32681474422.html?spm=2114.01010208.3.82.RcIbCO&ws_ab_test=searchweb201556_10,searchweb201602_1_10057_10056_10065_10037_10055_10054_10059_10058_10032_10017_107_10060_10061_10052_414_10062_10053_10050_10051,searchweb201603_9&btsid=f7e67388-974f-43ce-b8d8-a5cc38b284c2
[18:03:37] <veverak> looks interesting
[18:03:42] <veverak> deshipu: btw, super lucky here
[18:04:09] <deshipu> veverak: see
https://hackaday.io/project/6050-tote/log/22895-proprioception-revisited-touch
[18:04:34] <veverak> https://scontent-vie1-1.xx.fbcdn.net/v/t1.0-0/q82/p280x280/13882146_1756723417932955_2189489648822015933_n.jpg?oh=bbe6b3c0a30e6ef33c86fb8adf47f2ec&oe=58467A83
[18:04:41] <veverak> hackerspace cnc got SOME results
[18:05:01] <deshipu> nice
[18:05:31] <deshipu> that's for esp-12?
[18:06:02] <veverak> yep
[18:07:30] <veverak> http://www.ebay.de/itm/NEU-4-Achse-CNC-Router-3020-Gravurmaschine-Graviergeraet-fraesmaschine-Fraeser-3D-/141748376814?&autorefresh=true
[18:07:58] <veverak> all tests clear again!
