#qi-hardware IRC log for Saturday, 2012-09-01

kristianpauljesus there is also a brand field01:06
kristianpaulnot for me right now :)01:07
kristianpauland P/N01:07
kristianpaulwpwrak, i guess P/N is required for boom next features?01:08
GitHub130[m1-gps-expansion] kristianpaul pushed 1 new commit to master: https://github.com/kristianpaul/m1-gps-expansion/commit/df374862787c485752b59798dcd681ba55ba5e9802:07
GitHub130[m1-gps-expansion/master] - Naming values and units fixed to better Understandability. - Cristian Paul PeƱaranda Rojas02:07
kristianpaulnow going power symbols and no more eeschema edtiding for now..02:08
kristianpauland yes and add specifc boom files :-)02:08
kristianpaulwpwrak: is obvious if i want to use some components from the qi-he kicad liba i have to re-draw layout, right?19:55
wpwrakif you want to replace an already existing footprint with one from kicad-libs, then you have to re-layout that part, yes19:58
wpwrakif it's just one piece or a small number, you only need to redo the traces that connect to it. i.e., delete the 2nd to last segment, delete track on the side of the change (deleting the track will also get rid of any tiny "technical" segments kicad may have inserted), then redraw from the point where you cut it.20:00
wpwrakif you have lots of changes, maybe just keep a copy of the old layout for reference, then delete all traces and redo them. with the old layout as a reference, this usually doesn't take long.20:01
wpwrak(copy) preferably such that you can load it in pcbnew and examine every little detail, if needed20:02
kristianpaulI just want to improve some power labels for this desin https://github.com/kristianpaul/m1-gps-expansion.git20:22
kristianpaulwpwrak can you take a look at it when you have time?20:22
kristianpaulthe power supply "sheet" basically i want change that labels for somemething more "apropiate"20:23
kristianpaulperhaps no need20:23
kristianpaulsince the component cutoff was already made now i just need to reflect that on the layout and add other files boom may need20:24
wpwrakif you change power labels in the schematics, that shouldn't affect the layout at all20:34
kristianpaulgood !20:35
DocScrutinizer05hi folks20:35
kristianpaulhey Doc :)20:35
wpwrakare you sure you want to write 0u01 ? and not something like 10n ? :)20:36
kristianpaulha, i havent correlate that yet20:36
wpwraki know it can be tempting to make things not too obvious, but ... ;-)20:36
kristianpauli just took design from upstream and start working on it :)20:36
kristianpaulsure go ahead, all feedbak is welcome !20:37
kristianpaulstill some ICs to drop i just wanted to be conservative on the ref freq sheet as well20:38
DocScrutinizer05Hmm, I may plan for something *useful* now, like visiting dentist, some days on the countryside, or even starting some interesting work20:39
kristianpaulfor now i think i'll need some components from kicad-lib.. 20:39
kristianpaulso should i copy or .. link? (how?)20:39
wpwrakreference them in the .pro file. you can see in existing projects how it's done.20:41
kristianpaulah in the pro, i was looking at the .sch20:42
wpwraki.e., you can set the search path (board-m1) or just make an explicit path to the file (ben-wpan and so on)20:42
kristianpaulah LibDir=, quite easy :)20:42
kristianpaulDocScrutinizer05: how is that, are you providing consulting  to a dentist? or just going to the densist as any one do20:47
DocScrutinizer05just as anyone does20:48
DocScrutinizer05though most do faster than me I guess20:48
DocScrutinizer05my medical doctor always gets consulting for free :-D20:49
kristianpauland this interesting work, can you tell us a bit more, what is it about?20:49
DocScrutinizer05let's see what this new dentist will be like20:49
DocScrutinizer05well, for now it's more like a few visits to dentist plus offtime at the countryside. No new exciting tasks yet20:50
DocScrutinizer05I'm just a bit wiser now than I've been 9 months ago. I now know what a mess LTE is inside, and I know I don't really like clearcase20:53
DocScrutinizer05though maybe clearcase is just fine, when used without any windows around, and without weird wrappers to enforce weird rules for naming etc20:55
DocScrutinizer05and it seems I'm getting too old for mere coding monkey business20:56
DocScrutinizer05or code quality gets worse by the year, so I finnaly reached the state where I feel nausea trying to grok wtf they do 20:57
DocScrutinizer05anyway it's abit depressing to quit a job after you spent 6 month with 90% of the time only learning new weird stuff (like clearcase, chip docs, document management system and why you never find what you are serching for), then slowly go productive to maybe 50% of the time doing real work and only 50% learning weird useless stuff, and then you quit and would prefer to rm -rf /knowledge/busines/2012/* in your brain21:06
DocScrutinizer05I haven't touched any electronic meter or tool in the whole time :-/21:08
DocScrutinizer05except Lauterbach21:08
wpwrakat least something :)21:08
DocScrutinizer05yeah. Indeed Lauterbach is mighty mighty magic21:09
DocScrutinizer05not so much the hw, the sw is what does all the really amazing stuff21:09
wpwraknext, you'll join some electronics repair shop in a 3rd world country. back to basics ;-)21:09
mthpeople say the main attraction to ClearCase is the dynamic views, but to me it looks like a feature that optimizes to minimize the use of the cheapest resource21:10
mthas in, hard drive space is cheap nowadays, but server load and network bandwidth are not21:10
DocScrutinizer05I actually already forgot if we were supposed to use dynamic views or err the opposite (private?)21:11
mthgit optimizes the exact opposite way and that works really well21:11
DocScrutinizer05but yes, the views (basically bindmounted virtual volumes with directory strucure) were the most annoying thing, since they are slow, instable, and ... windows-infested21:13
DocScrutinizer05only thing more annoying than the views themselves is composing a view for your needs21:14
DocScrutinizer05since there's never been a simple revision/checkpoint/whatever to checkout, but you always needed to pick the right branch&version of each module to create a complete view for your own21:16
mthwe had about 100 people with the ClearCase server being a 60 MHz Sun server... I was so glad that my team could use CVS instead21:17
DocScrutinizer05and of course there been no HEAD or anything like that21:17
mth(this was in a time that our desktops were 500-1000 MHz)21:17
mthis the "no HEAD" a problem of ClearCase or of the company you were working at?21:20
mthdocument management system where you never find what you're looking for sounds familiar as well21:23
mthpeople figured out that the only search option that actually worked reliably was searching by author21:23
mthso you could find a document as long as you knew who wrote it21:23
DocScrutinizer05yes, kinda identical to my experience21:33
DocScrutinizer05only for us it's project number, like 168301521:34
DocScrutinizer05so if you know the project number of that chip, you might find the datasheet for it21:34
DocScrutinizer05the no-HEAD problem been an organisational one I guess21:35
DocScrutinizer05since basically on clearcase you could share views21:35
DocScrutinizer05or simply define a proper top mudule that includes all you n4eed21:36
DocScrutinizer05or simply define a proper top module that includes all you need21:36
DocScrutinizer05dang, typos21:36
DocScrutinizer05btw it turned out we got (at least) THREE concurrent DMS21:39
mthlarsc: in jz4740-pcm.c, there is this line: for (stream = 0; stream < SNDRV_PCM_STREAM_LAST; ++stream) {22:33
mthshouldn't that be "<=" instead?22:34
mthsince SNDRV_PCM_STREAM_LAST is equal to the last item in the enum (SNDRV_PCM_STREAM_CAPTURE)22:34
Aylait'd be a problem if SNDRV_PCM_STREAM_LAST is 022:34
mthit's 122:35
Aylahence the "would" :)22:35
mthit makes a diference for any _LAST value that is not negative22:36
mthI noticed because the Ingenic 4770 driver has very similar code to larsc's driver, but has a hardcoded "2" there22:36
mthso I checked whether I could replace 2 by SNDRV_PCM_STREAM_LAST and it turns out the values don't match22:37
mthlarsc's driver is the only user of that constant in the entire sound/ tree, by the way22:38
mthso we could just as easily declare the constant wrong and larsc's code right ;)22:38
mthI prefer exclusive upper bounds anyway22:39
mthit seems ALSA always uses _LAST for an inclusive upper bound, but most of the rest of the kernel uses it exclusive22:42
mththis is the kind of thing where an enforced coding style would really help22:42
mthactually, the rest of the kernel uses exclusive and inclusive about as often22:44
mthand sometimes it uses _AFTER_LAST to be explicit about being exclusive22:44
--- Sun Sep 2 201200:00

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