#qi-hardware IRC log for Wednesday, 2012-04-25

wolfspraulcladamw: good morning! :-)00:55
wolfspraulit looks like you and werner make good progress on the schematics00:55
cladamwhi good morning !00:56
wolfspraulplease let's continue full focus on that, bring the quality up...00:57
wolfsprauluntil even werner cannot find a problem anymore :-)00:57
wolfspraul(hopefully I won't eat those words one day :-))00:57
cladamwso I may keep cleaning up but not my priority one now. :-) phew~ Werner wrote too long to digest them now, but I put them into my todo works. :-)00:58
wolfsprauloh I think that's good to do fast00:58
wolfspraulwhat else do you have on your list?00:59
cladamwtrack house and book meeting to them. :(00:59
wolfspraulah ok00:59
cladamwthen back to quickly clean up. :)00:59
wolfspraulsure one by one01:00
wolfspraulbut definitely that is good feedback, exactly the quality stuff we were hoping to get out of the exercise01:00
wpwrakwolfspraul: i have a reputation of worst-case reviewer to uphold ;-)01:01
wolfspraulwell yeah. let's stay practical. perfection the enemy of good...01:01
wpwraki think it's worth the effort to spend a bit of time on this. each mistake we find now is much easier to fix than in later steps. also, each mistake we detect is part of the learning experience. the goal isn't only to finish a task but to become better at doing this sort of tasks.01:03
wolfspraulyes sure01:03
wolfspraulthat's why I encourage Adam to do it quuick01:03
wolfspraulset the gold standard, become more efficient later01:03
cladamwwpwrak, perfect >> good >> so so, so I face you as the one you are as reviewer to lead us to be as reputation you made. :-)01:03
wpwraki hope we can motivate more people to join the review. as individuals, we don't scale. as a group, we do.01:05
cladamwwolfspraul, wpwrak btw, i do some works then back to this great rules of exercises. :-)01:05
wpwrakalso, more inexperienced reviewers now means more experienced reviewers later.01:05
rohnew schems?01:07
rohis there a pdf?01:07
wpwrakroh: kicadized M1r4 schematics01:07
rohi see. still.. is there a pdf version?01:08
wpwrakwolfspraul: shall we try to bring schhist back up ? that would also provide PDF01:08
rohdont want to nag, but as long as one cannot just do apt-get install kicad and start using the files, i think a pdf version would make it easier to convince people to look into it and review ;)01:09
wpwrakhmm, interesting. "print" just produces a bunch of empty pages.01:12
cladamwwpwrak, aha ... i forgot to try 'print' in 3494. now trying.01:13
wolfspraulroh: schhist makes a pdf as well, automatically01:14
wolfspraul(a lot, actually)01:14
wolfspraulyes, I am wiring this up right now01:14
wpwrakprinting seems quite thoroughly broken here :-(01:15
wpwrakplot to postscript work, but it's a little awkward to turn this into something friendly01:15
wpwrakbetter to fix schhist :)01:16
cladamwwpwrak, yes, and my text are strange though. don't know why, maybe some settings needed.01:16
wpwrakcladamw: when i "print", i get a white albino circuit hiding in bleached snow ...01:17
wpwrakeven garbled text would be an improvement ;)01:18
wpwrakwell, "print" never worked very well01:18
wpwrakah good, wolfgang already updated schhist's use of the options01:19
cladamwwpwrak, yes, as i remembered. KiCad's 'print' work not well. ugly. 01:19
rohdoes it make a multipage one of everything too?01:19
wpwrakroh: it basically makes PDFs and PNGs of all possibly relevant "views". fully static content, ideal for high-traffic sites ;-)01:21
cladamw[KiCad pdf] http://downloads.qi-hardware.com/people/adam/m1/tmp/kicad/output.pdf01:24
wolfspraulwpwrak: are you saying the latest kicad is not usable for schhist?01:25
wpwrakno, it should be fine01:26
wpwrakschhist uses plot, not print01:26
wpwrakand it seems that even print isn't broken for everyone (well, beyond its regular level of brokenness)01:26
rohwhow... funny font01:55
rohor its that only for me?01:55
wpwrakroh: you don't read High Klingon ? :)02:03
wpwrakthat's just the sort of brokenness i expect from "print". "plot" is better, though.02:04
qi-botThe build was successful: http://fidelio.qi-hardware.com/~xiangfu/build-nanonote/openwrt-xburst.minimal-20120423-2016 02:13
cladamw(Power Symbols) wpwrak what does this mean ? :-) >> "Perhaps we should use project-local components in such cases. Fortunately, it will be relatively easy to propagate changes of component names to the schematics (you can simply edit the .sch files with a text editor), and changes in the hierarchy won't affect the schematics at all. "03:11
cladamwi tried to understand but can't. :)03:11
cladamwgiven an example, then i may easily understand. :(03:12
wpwrakthings like IR_PWR or ETH_A1V803:19
wpwrakthey're very specific to chip and project03:20
cladamwyes, but i still don't understand your words. Did you meant to manually edit .sch ?03:23
cladamwinstead of creating more power symbols ?03:23
wpwrakah no, just if you rename the symbols, there's a quicker way for updating the schematics than through the GUI03:26
cladamwwithout 'chip name'( i.e. don't add power symbol ) ? and directly change txt in *.sch, is that you meant ?03:28
wpwraki basically meant to say "don't be afraid of renaming things" :)03:28
cladamwlemm try to figure it out your meanings. moment... i need to see text a little. :(03:29
wpwrakthere are two topics: 1) what to do with those power symbols. 2) how to implement changes.03:29
wpwrakfor 1), i think we shouldn't have tons of chip- and/or project-specific symbols in the "global" library03:30
wpwrakbut we can still have such power symbols, namely in a project-local library03:31
wpwrake.g., one circuit may have VIDEO_IN_3V3, AUDIO_3V4, ... while another may have ANALOG_5V. and so on.03:32
wpwrakthe common library could simply have 3V3 and such. people can then copy from there and adapt for project-specific needs.03:32
wpwrakwe can make an exception for extremely common things, like USB.03:32
cladamw(ETH_A3V3) for example. i tried to use local label (black) to connect chip's pin. but in U2.38 its type is 'Power input' if I use local label to connect that pin, I'll get warning when ERC.03:33
wpwrakah ? that's strange. what i see here is a local label and the ERC passes03:34
cladamwthen I created a power symbol called also ETH_A3V3 then ERC passed, I also tried to change U2.38 origin type to like 'input' then I didn't need that power symbol though. but in the end, I found changing symbol's type IS NOT good idea. :)03:35
wpwrakbut i don't think we should have local labels for that. the power symbol is just as compact. so it's simpler to only use it, and not add something unfamilar people have to puzzle out first03:35
wpwrakno, don't change types to something wrong just to pass ERC :)03:36
cladamwwpwrak, yeah. i just didn't remove those local label, it doesn't need at all once there's power symbol(ETH_A3V3)03:36
wpwraknow i understand the history ... you could have used a local label without ETH_A1V8 etc.03:37
wpwraki mean without a power symbol called ETH_A1V803:37
cladamwwpwrak, you can try to remove that powe symbol (ETH_A3V3) and do ERC, you will get Err. :)03:37
wpwrakbut then you would have had to add the "POWERED" symbol to that net03:37
cladamwoah... really ?03:37
wpwrak"POWERED" is just a power output. this tells kicad that someone is supplying the net with power03:38
cladamwso just add "POWERED" symbol to such case ?03:38
wpwraki like the power symbols better. just make them project-local03:38
wpwrak(and get rid of the local labels. too much redundancy :)03:38
cladamwaha.... i got it. I was thought PWR_FLG is a must in KiCad to as original 'power source' . :-O03:39
cladamwokay, i go to try now and to see this Powerful "POWERED" functionality. :)03:40
wpwrakthe problem you encountered when you only had the local label but no power symbol was that there was a power input on the next, but no power output. kicad couldn't tell that the right side of L5 was in fact a power output. by i'm actually surprised that you didn't need to add POWERED somewhere. lemme see if it isn't hidden elsewhere ...03:41
wpwrakhmm funny. must work by magic :)03:43
cladamwokay. .. good, that's also i concerned that should I need to add too many "POWERED" symbol everywhere ?03:43
cladamwfunny waht ?03:43
cladamwi don't like to add also too many "POWERED" symbols everywhere if like our huge schematics. 03:44
wpwraki'm surprised you don't need a POWERED there. not sure why. but if it works, that' fine :)03:45
wpwrakanyway, ia gree with using POWERED only when ERC complains. if kicad is happy with what you have, don't add it. i find it more a bit of bureaucracy than something actually useful.03:48
cladamwNice ! passed ! so means there will be more "POWERED" symbols. :)03:48
cladamwI'll remove those application unnecessary power symbols also in kicad-lib/ 03:49
wpwraki think it's nice to have power symbol represent power supplies. just keep the project-specific ones in a libraries belonging to the project.03:51
wpwraki.e., you can make a directory in board-m1 called board-m1/components/ and put some .lib there. or just have a .lib in board-m1/r4/03:52
cladamwso i'll add powered symbol and keep using local label to connect. :)03:52
cladamw(project-specific.lib) phew~ now just realized your idea. 03:53
wpwrakthat's also useful for other "weird" things. e.g., if you have some fancy PCB structure (connector, multi-tap antenna, whatever) that's totally project-specific.03:57
cladamw(project-specific, project-local) hehe...all i discovered the discussions between you an me is that my poor english annoying me... ORZ 04:00
cladamwi appreciated that you always pay patience to me. 04:00
cladamwbtw, it's time that I have to edit such wiki page: http://en.qi-hardware.com/wiki/Rules_on_Editing_Schematics04:03
cladamwwill put useful rules in there.04:03
wpwrak(page) cool ! we should actually make a proper document, but i've always been too lazy for that ...04:04
cladamwwpwrak, lemme do this. you bet work on others. :-)04:05
wpwrakthanks ! :)04:07
qi-bot[commit] Adam Wang: ADV7125KSTZ140: changed its pin 13, 29, 30 (VAA) back to correct 'Power input' (master) http://qi-hw.com/p/kicad-libs/a59f00505:53
qi-bot[commit] Adam Wang: removed application power symbols, they could be replaced by POWERED symbol( in POWERED.lib ) (master) http://qi-hw.com/p/kicad-libs/acad34a06:03
cladamwwpwrak, *grin* just found that there were still pins are not correct. :)06:13
LunaVoraxHello everyone!06:29
LunaVoraxHow is everything going wolfspraul? :)06:29
wolfspraulwpwrak: maybe this could be something for you www.ruhoh.com (thanks Jon)06:29
wolfsprauldoing a little spring cleaning right now, removed rss2email and some backoffice mailing lists... (wiki-log, planet-log, issues, commitlog)06:30
wolfspraullet's see whether anyone complains ;-)06:30
wpwrakthese lists may be one of those case where something reaches the peak of its notoriety, up from zero, when it is removed ;-)06:52
wpwrak(ruhoh) sounds user-friendly :) if the blogging bug should ever bite me, i'll try to remember it ...06:52
cladamwwpwrak, where's the folder KiCad system share/library after installed kicad.bzr and your patches ? I don't find it like as before since I'd like to copy KiCad's original symbol. :)08:12
whitequarksuddenly my blog is on planet-log list08:58
viricwolfspraul: in what city do you reside?09:29
qi-bot[commit] Werner Almesberger: components/EXPAND: removed project-specific power symbols (ETH_PLL1V8, etc.) (master) http://qi-hw.com/p/kicad-libs/8ae9db511:26
wpwrakoh dear. now we're getting chinese dates. at least they're not in .tw years :)11:29
wpwrak(kicad updated timestamps in the local language. a nice example why this isn't always a good idea :)11:30
wpwrakactually ... perhaps we should set kicad to english when invoking it via a makefile. that'll rid us of other surprises. the timestamps are harmless but there are more sinister things that can happen.11:31
mthwpwrak: imo localized dates are a bad idea anyway in an international project; I always use yyyy-mm-dd, since that is unambiguous12:10
mthand it sorts correctly alphabetically12:10
wpwrakoh, those timestamps are quite worthless. all they truly accomplish is create commit noise.12:16
larscand the file probably already carries a modification date in the file metadata....12:21
wpwrakthe nastier things i mentioned are localized texts, some of which act as keywords in files. then anything trying to parse that output will have to guess correctly that something as understandable as ""รก12:21
wpwraklarsc: and in git ...12:21
wpwrakit's overkill with a bomb that has a chance of going off on your side as likely as on the side of the already twice dead enemy12:23
wpwrakof course, if you regularly suffer zombie attacks, then that may actually be acceptable. but ...12:23
larscwhy'd somebody come up with the idea of localizing keywords in a fileformat?12:24
wpwrakwhy do people win the darwin award ? :)12:26
larscbecause they removed themself from the genepool12:28
wpwrakexactly. X ergo X ;-)12:30
qi-bot[commit] Maarten ter Huurne: MIPS: JZ4740: Fixed value for round robin constant. (jz-3.3) http://qi-hw.com/p/qi-kernel/0de2b2b13:37
qi-bot[commit] Maarten ter Huurne: MIPS: JZ4740: SLCD: TV-out related cleanups. (jz-3.3) http://qi-hw.com/p/qi-kernel/e7de30913:37
qi-bot[commit] Maarten ter Huurne: MIPS: JZ4740: SLCD: Add PAL-60 mode for TV-out. (jz-3.3) http://qi-hw.com/p/qi-kernel/0ff196413:37
zoltanh7211Hi everyone14:06
DocScrutinizermth: +1 :-)20:53
qi-bot[commit] Werner Almesberger: b2/Makefile: add dependency generation (.d method) (master) http://qi-hw.com/p/eda-tools/9d5c69822:39
qi-bot[commit] Werner Almesberger: b2/Makefile: remove y.output on "make clean" (master) http://qi-hw.com/p/eda-tools/c274c1d22:39
qi-bot[commit] Werner Almesberger: b2/Makefile: add "spotless" target (master) http://qi-hw.com/p/eda-tools/7fe553a22:39
qi-bot[commit] Werner Almesberger: b2/: first tentative version of hierarchy parser, with example (master) http://qi-hw.com/p/eda-tools/6bb8b0322:39
qi-bot[commit] Werner Almesberger: b2/: syntax change: use = for instead of : for format declarations (master) http://qi-hw.com/p/eda-tools/75479a822:39
qi-bot[commit] Werner Almesberger: b2/HIERARCHY: add diodes, to make the example more interesting (master) http://qi-hw.com/p/eda-tools/45196de22:39
qi-bot[commit] Werner Almesberger: b2/: add C-style comments (master) http://qi-hw.com/p/eda-tools/b3c865e22:39
--- Thu Apr 26 201200:00

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