#qi-hardware IRC log for Tuesday, 2014-02-11

wpwrakheh, there's also a clone from .ve. a bit easier than yours: http://jotase.github.io/FlappyFree/#200:01
wpwraknote that they have some currency restrictions there the people really resent ;-)00:01
wpwrakhere's a collection of clones: http://www.palmentieri.it/somepx/flappyjam/00:01
dos1wpwrak: haha, submited :D00:07
whitequarkDocScrutinizer05: I see, thanks00:51
DocScrutinizer05soory, stack cleared00:51
whitequarkre grounding and wristband00:51
DocScrutinizer05aah, np00:51
whitequarkwpwrak: (salt) I'd search pubmed to substantiate my claim but I'm too lazy right now :p00:52
DocScrutinizer05a very simple tip: use a wooden desk00:52
whitequarkDocScrutinizer05: I already do00:52
whitequarkwell, it's some kind of wooden laminate00:52
whitequarkbut that should work just as well, wouldn't it?00:52
DocScrutinizer05second tip: try to always first touch the desk's metal feet before you touch anything sitting on the desk00:53
Action: whitequark nods00:53
DocScrutinizer05even better would be a metal desk of course00:53
DocScrutinizer053rd tip: always touch electronics at the ground plane or some metal frame first00:54
DocScrutinizer05it's usually not the electronic device that's electrically charged, it's usually you00:55
DocScrutinizer05ESD concerns in lab are massively exaggerated nowadays, since 99.9% of components already have quite decent ESD protection integrated00:57
whitequarkDocScrutinizer05: I'm quite bothered when I stand up and hear a loud bang from the discharge00:58
whitequarklike, it's the kind of bang you hear when you plug something with 10000uF cap into mains00:58
DocScrutinizer05that's bad! you need to do something, change the situation, eliminate the van der Graaf00:59
DocScrutinizer05laminate may have a very high part of resin resp epoxy or plastic in general01:00
DocScrutinizer05your chair shouldn't be plastic either01:00
whitequarkwpwrak: (exif) imgur strips exif data. likely because it's the default image host for reddit, and unknowingly posted exif is a huge source of personal data leaks01:02
DocScrutinizer05a desk lamp with PE-grounded metal case that you touch to switch it on, each time you sit down at your desk, may help wonders01:02
DocScrutinizer05but when you hear static noise when you stand up, then you got the van der Graaf locally at your desk, and then you *really* need to fix that01:04
DocScrutinizer05I's blame your chair01:04
whitequarkI'm pretty sure it is the chair01:04
DocScrutinizer05and your desk surface together with the clothes on your forearms01:04
whitequarknah, no clothes on forearms--using a tshirt01:04
DocScrutinizer05ok, very good01:05
DocScrutinizer05use a conductive "blanket"01:05
DocScrutinizer05on your desk01:05
DocScrutinizer05way better than wrist band01:05
whitequarkhttp://www.amazon.com/Elenco-Electronics-WS-3-23-Inch-Anti-Static/dp/B0002KXCYE/ref=sr_1_3?ie=UTF8&qid=1392080841&sr=8-3&keywords=esd ?01:07
DocScrutinizer05whitequark: yep01:43
DocScrutinizer05get one that's large enough to not annoy you01:43
DocScrutinizer05I.E. one that's so large that you don't even think of placing something _not_ on the mat01:44
DocScrutinizer05of course a metal plate works as well01:45
apeleteHi larsc10:50
apeletelarsc: did you receive the email of my documentation patch ?10:51
apeletelarsc: there seems to be encoding issues on the patch, so I was wondering what I did wrong :-(10:53
larscapelete: I got the mail. What kind of encoding issue?11:48
apeletelarsc: some characters does not show up: http://marc.info/?l=linux-doc&m=139199386820719&w=212:10
apeletelkml doesn't even show the email at all: https://lkml.org/lkml/2014/2/9/28512:10
apeletelarsc: but it's ok if you got the mail, that means it's a client issue12:11
larscit looks good in my mail client12:14
apeleteok, thanks for the feedback12:14
apeletelarsc: with the documentation done, I can now get back at debugging the dma in the mmc driver12:15
apeleteexpecting a painful experience, does not even know where to start :-(12:16
larscshould be that bad12:23
CYB3RIs it possible to boot something on Xburst from USB (without flashing to NAND)13:18
CYB3Rlarsc: what commands should I pass to jzboot?13:24
larscthat I don't know13:25
larschas been a long time that I used it13:25
larscfor anything else but flashing the nand13:25
CYB3RIs it possible to use UART or GPIO without RAM?13:30
larscif your program fits in the OCM13:31
larscaka IRAM13:31
CYB3RI will be soldering a board in a day or two, so I want to test if everything works13:32
CYB3RMaybe I'll be doing tests before soldering NAND and SDRAM13:33
nicksydneywpwrak: ping ~13:45
wpwrakqi-hw emergency response team. please state the nature of your emergency.14:10
larscnot very responsive though that response team14:28
wpwrakwell, the caller hung up ...15:06
CYB3Rjust got awesome TCA8424RHAR from TI15:45
larscyeay, systemd *celebrate*17:05
whitequarkyou mean *mourn*17:06
larscI celebrate17:11
wpwrakwhitequark: he means this kind of celebration: http://www.clipartpal.com/clipart_pd/holiday/halloween/halloweendance_10145.html17:11
larscswitching to upstart would be switching from a bad design to a bad design17:11
larscswitching to systemd is switching from a bad design to a good design17:11
whitequarkyou can't seriously call systemd a good design17:12
whitequarkI could kinda accept "we don't currently have resources for yet another better init replacement"17:12
whitequarkwhich we, unfortunately, likely don't17:12
larscI can ;)17:13
whitequarkthen you're a troll17:13
whitequarkwpwrak: http://os.colta.ru/m/photo/2011/09/22/6_5.jpg17:14
wpwrakthat's a good crowd. once they get to the vodka, the celebration will be fun17:17
wpwrak(troll) or maybe he's poettering ? :)17:17
rohlarsc: debian decided?17:17
rohwhitequark: good designs can be reimplemented properly and become better. bad designs stay shit regardless of the number of implementations.17:18
whitequarkroh: exactly, systemd can't be fixed.17:19
kyakwhitequark: why is systemd bad design?17:20
whitequarkkyak: it's a monolithic thing which tries to do everything at once. from running services to managing networks to udev to tracking console sessions to displaying qr codes17:21
kyakwell.. is it bad?17:24
whitequarkyou'd think so17:29
larscwhitequark: http://0pointer.de/blog/projects/the-biggest-myths.html17:36
whitequarklarsc: okay, after looking at 20 and 22 I'll maybe reconsider it.17:38
rohwhitequark: upstart can't be fixed17:39
rohand sysv is through.17:39
rohsystemd can be fixed. atleast its a sane concept cleaning up all the sins of the stone-age a bit better17:39
dos1the biggest pain of systemd for me is that it's Linux only17:40
rohits not perfect for sure. but its a change, which i think can help us out more than upstart and sysv together.17:40
dos1but I guess that could be fixed17:40
rohalso note.. for embedded you usually use something different.. like openwrt17:41
whitequarkdos1: actually that one probably could not. lennart said this explicitly, even17:41
larscit's very hard to fix17:41
whitequark"too much linux-specific interfaces"17:41
rohopenwrt uses its own replacement: procd and its companions netifd and some more.17:41
larscyou'd have to port stuff like cgroups to $TARGET_OS17:41
rohwhitequark: i could not care less. yes its linux specific so what.17:41
dos1I don't mean systemd implementation, but the design17:41
whitequarkroh: tat was to dos117:42
rohsolaris has another, own implementation of all that, osx too.. so why care?17:42
rohwhitequark: sorry17:42
whitequarkdos1: implementing it is a huge deal though... since it's so complex17:42
rohyes.. dos1. doesnt matter.17:42
dos1I believe that good design should be at least somehow portable17:43
rohdos1: no.17:43
dos1if POSIX lacks some features, extend it and provide well-defined standard, so anyone else can implement it if they want17:43
rohgood designs solve proplems. defined problems. when you define all problems into one, your design will be bad and never be implemented properly.17:44
rohand posix is part of the problem, not the solution.17:44
whitequarkdos1: you forget though that POSIX is defined after the fact17:44
rohsee gnu hurd. posix to the death doesnt dolve anything.17:44
whitequarklook at POSIX AIO. no one implements or uses it.17:44
rohforget classic unix. thats whats holding us back. yeah there are some good parts on that concepts. those we keep. but the legacy mindfuck can go and stay away17:45
whitequarkI'm not even sure how POSIX is useful for anything except extorting money for certifications17:45
rohin the future one will have linux with systemd and ipc via kdbus.17:45
whitequarkbecause there's as much subtly incompatible implementations as there are kernels17:45
dos1from my point of view, if there's some well documented standard (call it even "the Linux standard") I can conform to when writing my own kernel and that will be enough to make systemd working on my kernel, then it's "portable enough"17:46
whitequarkas many*17:46
whitequarkdos1: you'd be basically rewriting implementation details of Linux. that's the problem.17:46
dos1I don't care if this standard is in fact POSIX, Win32 or anything else17:47
whitequarkwhat is the point of implementing your own kernel just to badly emulate Linux with it?17:47
dos1whitequark: yes. that's why I want well defined and documented standard, not "Linux emulation"17:47
whitequarkit's a chicken-and-egg problem. you need competing implementations for a sensible standard to emerge17:48
dos1you might be right17:49
whitequarkand I honestly don't think that any kernel is going to compete with Linux in... a decade or something? at least. that's how long it takes to set things up17:49
rohdos1: there is nox _fixed_ standard17:49
rohyou are always dependant on behaviour and features of the state of versions you code against.17:49
dos1but I fear that it will stop on "custom implementation" stage, exactly for that (I honestly don't think that any kernel is going to compete with Linux in... a decade or something?) reason17:49
rohso YES, compiling something against a glibc2.something makes one depending on that or later compat.17:49
rohbut hey... you COULD just do opensource or recompile your crud from time to time. and be fine.17:50
whitequarkyou don't get standards by thinking "oh we need to make a standard for it"17:50
rohcompatibility is something one can have cheap and easy.. on the source level.17:50
whitequarkthat's how you get "standards" like that shit microsoft creates. it solves only the use case of the person writing it17:50
dos1source level is enough for me17:50
whitequarkyou need actual competing implementations people would use17:50
rohthe kernel will never be complete17:50
whitequarkand collaboration17:50
whitequarka lot of it17:50
whitequarksee: browsers17:51
rohlinux is an iterative, evolving process of an os. nothing which will ever be done, complete or finished, i think17:51
whitequarkis software ever complete? :)17:51
dos1whitequark: I understand, but I don't want Linux to became new Internet Explorer :P17:52
dos1I'm not against custom extentions, new features outside of standard etc.17:53
dos1I'm against IE mentality when doing so17:53
whitequarkwell, IE was deliberately made incompatible17:53
dos1and systemd story somehow triggers some warning light in my head17:53
dos1hopefully it's false alarm :P17:54
larscwell you can't run the BSD init system or the solaris init system or the MacOSX init system on Linux either, but I get your point17:54
dos1I don't know much about non-Linux init systems - so maybe if I did, then the same light would blink as crazy when thinking about them ;)17:56
whitequarksstemd looks quite a bit like launchd17:56
whitequark(osx init)17:56
larscsome features were inspired by launchd17:57
dos1maybe except OS X, as I simply don't care17:57
dos1I don't expect anything from Apple, as they can't expect me to be their customer ;]17:57
CYB3RAnyone knows if Jolla is using systemd or not?17:57
whitequarkit makes no sense to be ignorant about design, even if you don't use the system just because of licensing terms17:58
dos1whitequark: maybe, but if I won't use it because of licensing terms, then I don't care about it being portable and for sure never will17:59
larscit's a bit sad that the Unices and Unix-like OSs are drifting apart API wise. But systemd is not the problem17:59
larscmerly a symptom of the problem if you will so18:00
whitequarkdos1: portability != design. I may never run OSX but I sure will look into how its kernel is done18:00
whitequark(well, except its kernel is a decade behind Linux so I actually won't. but I could.)18:00
dos1whitequark: :nod:18:00
larsc"good artists copy, great artists steal" ;)18:01
wpwrak(posix) that discussion is bizarre. posix was one of the biggest steps forward in the whole history of unix. before, there was a gazillion of diverging unices. that's was the situation that got us things like autocrap.23:31
nicksydneywpwrak: was reading the mailing list about nano....when the project was started the ingenic chip was sourced from ingenic directly ?23:52
--- Wed Feb 12 201400:00

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