#qi-hardware IRC log for Saturday, 2015-03-14

wpwrakphew. logitech M560, the mouse from hell.00:32
wpwrakthe middle button switch on the scroll wheel of my M215 was dying. so i looked for a new critter. the M560 looked nice. some people claimed that there would be issues with the middle button since the wheel worked in an unusual way, but then yet others said the little button next to it can act as middle button. so all should be well, right ?00:34
wpwrakturns out that logitech implemented some win8-specific mouse-keyboard bastard, with virtually everything hard-wired to hostile functions. of the four more-or-less-crappy solutions/workaround i found on the intertubes, two didn't work, one requires installation of some closed-source tool that does a gazillion of things, maybe including "fixing" the mouse, and the fourth finally worked. needless to say, trying to puzzle it out by lookin00:38
wpwrakg at xev, xbindkeys -k or -mk, et al. was hopeless, too.00:38
DocScrutinizer05hmmm01:12
DocScrutinizer05tried a od -x /dev/* ?01:13
DocScrutinizer05err01:13
DocScrutinizer05od -x /dev/input/<*>01:13
DocScrutinizer05I seen that M560 critter in a blister pack lately and wondered if that "little button next to scrollwheel" wasn't again just a mechanical lever to clutch the pitch snap of the scrollwheel, for "freewheeling" vs "stepping" mode01:16
DocScrutinizer05btw weird scrollwheel middle button function and a little pushbutton switch next to it acting like the genuine middle mouse button is exactly the _default_ config of the MX-revolution aka MX550001:19
wpwraknaw, the freewheeling is what you get if you press and release the wheel. press and release again and it is stepped again.01:20
DocScrutinizer05also the mentioned "mouse+kbd hybrid" is exactly what MX-Revolution does, I found out by inspecting all devices in /dev/input/ with od. Anyway meanwhile the mapping of scancodes to key events is absolutely fine for the MX01:21
wpwrakthe button sends some windows curse, mainly as "keyboard"01:21
DocScrutinizer05yes, *exactly* what MX-Revo does in _default_ mode. You know, you can config/program the mouse itself01:22
wpwrakfor the m560 there are contradictory configurations. and the one you get with the "help" of the program that will then interpret the config is wrong. i'm not sure how they accomplish that :)01:22
wpwrakyes, with a window-only tool (setpoint) or that closed-source monstrosity (hidpoint). apparently, there's no public programming info01:23
DocScrutinizer05maybe you wanna check my revoLUTIONconTROL stuff?01:23
DocScrutinizer05http://maemo.cloud-7.de/mx-revolution/01:24
wpwrakrevoco.c looks interesting. thanks !01:27
wpwrakhow the heck does one change free-spinning vs. click by software ?01:28
Action: whitequark is looking at manufacturing ceramics01:28
wpwrakpottery ! :)01:28
whitequarkso i could make an alumina green body using direct coagulation casting01:28
whitequarktake alumina, ammonia polyacrylate, magnesium oxide, mix, then form01:29
DocScrutinizer05wpwrak: revoco -mode ? sth like that01:29
whitequarkyou don't need a press this way01:29
whitequarkit's a sol-gel process. you can even machine the resulting "gel", it's cohesive enough01:29
whitequarkthen you sinter it as usual01:29
whitequarkafter that you need a ceramic-metal seal. in the industry they sinter a molybdenum manganese oxide powder over the ceramic in the atmosphere of wet hydrogen01:30
DocScrutinizer05http://privatepaste.com/89e62636f401:30
DocScrutinizer05wpwrak: ^^^01:30
whitequarkand i want to get involved with literally no part of that process01:30
wpwrakDocScrutinizer05: no, i mean: that's a mechanical characteristic. how can software alter that ?01:30
whitequarkthere are nicer things like active brazing alloys01:30
DocScrutinizer05it has a little servo built in01:31
whitequarkyou have an AgCuTi alloy, where the Ti content apparently makes the solder eat the ceramic (!) thus wetting it01:31
whitequarkwhich actually sounds realistic for a change01:31
DocScrutinizer05the Revolution and MX-5500 had. No other mouse since had01:31
wpwrakhmm yes, this one feels all mechanical01:31
whitequarkthe annoying part is that they don't sell these alloys so i'll have to manufacture them myself01:31
wpwrakwhitequark: be thankful that the underlying physics are in place already :)01:32
whitequark... you say so, but i keep googling various steps of this and finding papers from like 201301:32
wpwrakwhitequark: that's merely their discovery :)01:33
DocScrutinizer05wtf is a "atmosphere of wet hydrogen" ?01:33
whitequarkhydrogen and water mix01:34
whitequarkwater vapor01:34
DocScrutinizer05ooh01:34
whitequarkit's happening at about 1100 degrees, too01:34
DocScrutinizer05:nod:01:34
whitequarkdeeply unpleasant process from every angle01:34
wpwrakDocScrutinizer05: alas, i don't see anything about programming buttons functions in revoco.c01:35
wpwrakeverybody just seems to do this with xbindkeys01:36
DocScrutinizer05errm nope, the mouse only delivers keycodes01:36
DocScrutinizer05err scancodes01:36
DocScrutinizer05you however can profram the action of scrollwheel button01:36
wpwrakthere's also an obscure udev-based approach but i didn't get that to work01:36
DocScrutinizer05progam01:36
DocScrutinizer05lemme check what I got for my MX01:37
DocScrutinizer05apart from revoco manual=301:38
DocScrutinizer05err nope01:38
DocScrutinizer05NOT revoco manual, revoco auto!!01:39
DocScrutinizer05I'm using auto=901:39
DocScrutinizer05jr@saturn:~> revoco mode01:40
DocScrutinizer05note: MX-5500 support is experimental01:40
DocScrutinizer05click-by-click01:40
DocScrutinizer05switching to freewheeling when speed of scrolling exceeds "9" (whatever units that is)01:40
DocScrutinizer05so _all_ buttons deliver some scancode for me, most in dev/input/mouse, some in /dev/input/event1(?)01:42
DocScrutinizer05something is mapping those scancodes to key/mouse events01:43
wpwrakthat may be udev. there, i saw rules for mapping scan codes. alas, the stuff that should supposedly work for the M560 doesn't01:44
wpwrakalso, the mouse sends a whole string of such nastiness, so that may trip things, too01:45
DocScrutinizer05I just checked my udev stuff and it seems systemd fucked it up01:46
wpwrakfrom poettering, with love :)01:46
DocScrutinizer05>:-(01:46
DocScrutinizer05recently found: http://compsoc.man.ac.uk/~shep/pedals.html (please read more than headline, it had useful info for me) and http://homepages.cwi.nl/~aeb/linux/kbd/scancodes-8.html#kccb201:48
wpwrakyes, that's (one of) the system(s) i tried to use. didn't have any effect.01:53
wpwrakanyway, i now have the critter halfway satisfactorily mapped01:53
DocScrutinizer05 /usr/lib/udev/hwdb.d/20-usb-vendor-model.hwdb  ::  01:54
DocScrutinizer05usb:v046DpC51A*01:54
DocScrutinizer05 ID_MODEL_FROM_DATABASE=MX Revolution/G7 Cordless Mouse01:54
wpwrakit goes keystrokes -> xte -> click, so i won't have "middle-click and drag" or such, but maybe that's enough01:54
DocScrutinizer05and 01:55
DocScrutinizer05usb:v046DpC525*01:55
DocScrutinizer05 ID_MODEL_FROM_DATABASE=MX Revolution Cordless Mouse01:55
DocScrutinizer05 /usr/lib/udev/rules.d/97-hid2hci.rules02:01
DocScrutinizer05has (excerpt): http://privatepaste.com/20a369d94302:02
DocScrutinizer05don't ask me wtf is hid2hci02:03
DocScrutinizer05need to paste full file, it's too funny   http://privatepaste.com/b1544beb0902:05
DocScrutinizer05aaah, /usr/lib/udev/hid2hci: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.32,BuildID[sha1]=11d11fc14fe19a91e80a9141c767d55bf5b319fd, stripped02:09
DocScrutinizer05cool this systemd-udevd02:09
DocScrutinizer05http://privatepaste.com/eba65a3fa7  makes sense, this is a BT wireless mouse02:11
DocScrutinizer05nfc who does mapping, and where02:12
DocScrutinizer05btw I prolly couldn't live without MMB-dragging, used to zoom in pdf viewers et al02:14
DocScrutinizer05used to pan in eagle02:14
DocScrutinizer05http://privatepaste.com/7b618fc71c xev MMB press + release, if that helps02:20
DocScrutinizer05and od /dev/input/mouse0 yields: 000c08 for MMB-press and 000000 for MMB-release02:28
DocScrutinizer05btw that small button next to scrollwheel is mapped to XF86_SEARCH(?) on my system/mouse02:36
DocScrutinizer05now maybe some savvy people in here can tell me what's the problem when:02:39
DocScrutinizer05neo900:~ # write dos1 pts/002:39
DocScrutinizer05write: you have write permission turned off02:39
DocScrutinizer05neo900:~ # mesg02:39
DocScrutinizer05is y02:39
DocScrutinizer05(dos1 as well has mesg y))02:39
DocScrutinizer05I've seen suggestions to check /var/log/audit but that's empty   02:41
DocScrutinizer05http://serverfault.com/questions/543019/cant-write-to-other-users-you-have-write-permission-turned-off02:42
eintopfback09:58
Mike25226Would this be a suitable place to discuss the creation of a powerful 2.4GHz/5GHz PCB Antenna for wifi?23:42
virico23:44
viricno23:44
Mike25226Any suggestions of where I might be able to find help with my topic? (If i understand correctly what Copyleft hardware is... my topic would be classified as such)23:57
whitequarkPCB antennas are fairly standard23:59
whitequarklook for appnotes from wifi chip vendors23:59
whitequarkor not even wifi, anything that runs on 2.423:59
--- Sun Mar 15 201500:00

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