#qi-hardware IRC log for Tuesday, 2014-03-18

whitequarkhmm, seems like my Z missing steps problem was solved by limiting velocity to 1025mm/s00:13
whitequarkfrom 1500mm/s00:13
whitequarkalso I'm inclined to believe the .05mm repeat positioning accuracy, because if it didn't actually provide that, my holes wouldn't hold nuts00:14
whitequarkand they do00:14
whitequarkAperture Science Object Attachment System, final version: http://i.imgur.com/W9zbvZ1.jpg00:53
whitequarkI broke five endmills doing that :/00:53
whitequarkseems like there's something weird going on with Z axis and/or the table is not parallel to XY plane00:54
whitequarkso I experimented a bit with copper plating bath and graphite03:57
whitequarkresult: copper is EVERYWHERE except where I want it to be03:57
whitequarkhttp://i.imgur.com/uNE8jBl.jpg04:01
whitequarkthat is the *reverse* side of one-sided PCB04:02
whitequarkI actually tried to scratch off the copper and it is not easy04:02
whitequarknot significantly easier than with factory-made copper04:02
whitequarkthe layer is about... 40 micron04:03
whitequarkbut that's likely wrong figure affected by my ability to measure it04:04
whitequarknah, comes right off if I try to tin it: http://i.imgur.com/wgBMNZy.jpg04:11
whitequarkwhich is... well, predictable04:12
whitequarkno reason for it to have good adhesion to surface, every reason not to04:12
DocScrutinizer05((table is not parallel to XY plane)) quite possible11:38
DocScrutinizer05what I elaborated on yesterday(?)11:38
DocScrutinizer05((15 mills)) ouch!11:38
DocScrutinizer05((imgur)) LOL11:40
kristianpaulwb14:04
wpwrakwelcome back, internet ! :)14:54
kristianpaullol15:27
zrafawpwrak: where is libubb living? if I try to build libswd I get a lot of 18:59
zrafaben.c:(.text+0x4f8): undefined reference to `ubb_mem'18:59
zrafaben.c:(.text+0x524): undefined reference to `ubb_mem'18:59
zrafaben.c:(.text+0x550): undefined reference to `ubb_mem'18:59
zrafabut I do not see ubb_mem defined in ubb.h or something like that18:59
wpwrakthe ben-blinkenlights project ;-)19:02
zrafawell, I am tired right now, maybe that is the error :)19:02
wpwrakundefined ubb_mem would suggest that you have the headers but not the library19:18
zrafawpwrak: the problem was that I modified Makefile and had a wrong CFLAG :P19:33
larscLDFLAG19:35
wpwrakLDLIBS ? :)19:37
wpwrakalthough a few years ago, you could still get away with putting libraries in LDFLAGS19:38
wpwrakmust have been gnu ld getting a bit stricter that ended that19:38
larscit now cares about the order of the libs or something19:48
wpwrakyup. they probably removed a loop that should never have been there19:49
ysionneauto get the same effect you can now use -Wl,--start-group and -Wl,--end-group or something like that19:50
ysionneau(and put your lib list between those two)19:51
larscgreat, google's autocomplete suggests '-wl --start-group' when you type 'wl--start-g'19:52
larscof course it doesn't find anything because - infront of a searchterm means it shouldn't appear in the result19:52
larsc'Using this option has a significant performance cost. It is best to use it only when there are unavoidable circular references between two or more archives.'19:53
ysionneauwell, for "small" projects with few libraries (which are not too big) you just don't care19:56
ysionneaubut when building Android ou Gecko ou whatever indeed you care :)19:56
ysionneauyou should not use that unless there is really a "circular dependency"19:56
ysionneauusing it by default is bad design19:56
wpwrakhaving a circular dependency would probably constitute bad design on its own ;-)19:58
ysionneauwell, that depends, but usually it's not the case19:59
ysionneau(I mean, usually you indeed don't have those)19:59
zrafalarsc: wpwrak : the problem was a CFLAG because I commented out some var (# CC2543 = ../../ybox/fw/cc/) and then CFLAGS was no so nice with CFLAGS = -Wall -I$(LIBUBB)/include -I$(CC2543)21:27
zrafawpwrak: so It ran gcc $(CFLAGS) -c source file21:28
zrafadont know why then the error was showed by ld anyway (after that gcc of that source file it then runs gcc to build the binary with ld)21:29
wpwrakhmm, what would you do with ybox/fw/cc ? that's firmware for a chip you don't have :)21:38
wpwrakor does some other makefile invoke that ?21:40
Action: whitequark finally got to preparing the copper hypophosphite solution23:19
zrafawpwrak: yes, libswd has usage of ybox/fw/cc and then I was removing cc stuff from Makefile23:31
zrafawpwrak: so some -I option of gcc was empty when running make23:32
wpwrakyes, there's a header dependency. but you shouldn't need to build things over there, just have the files23:33
wpwrakanelok/anelok and anelok/ybox currently have a highly incestuous relationship with includes and even vpaths going back and forth. so have your barf bag ready :)23:34
wpwrak(if you get to building firmware, then ben-wpan/atusb/fw/ also becomes part of the family. but you don't need that)23:35
wpwrak(at last not yet)23:35
whitequarkbleargh, ammonia stinks23:45
whitequarklike a thousand cats momentarily cried in terror and pissed all over the place23:45
whitequarkooo, very interesting23:56
whitequarkhttp://we.easyelectronics.ru/HomeTech/metallorezist-rebyata.html#comment8569223:56
whitequark"persulphates etch on boundaries of copper crystals"23:59
--- Wed Mar 19 201400:00

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