#qi-hardware IRC log for Wednesday, 2012-09-26

pabs3whitequark: I guess you saw it got on slashdot?03:03
pabs3good points there about FPGA openness03:03
whitequarkmth: nope03:35
whitequarkit's serving static HTML03:35
whitequarkthere were around 12k people, that's not really that lot03:36
whitequarkfunny, around 5% of traffic was ipv603:37
wpwraksmart people (i.e., those who use static content) don't fear /. :-)03:46
whitequarkwpwrak: exactly03:50
whitequarkpabs3: yeah, I monitor referers03:50
whitequarkIMO, FPGAs are more like ASIC processes. They can and will be patented themselves, but after it's done, it does not affect you or your design, and you're not locked in to a particular chip.03:51
whitequarkwpwrak: actually... do you ever sleep at all?!03:55
wpwrakwhitequark: what for ? :)04:59
wpwrakwell, truth be told, i tend to sleep quite a lot. but in an irregular pattern and often just a few hours at a time.05:33
whitequarkwpwrak: I once read a blog of man who has tried a specific pattern of sleeping for 80 days or so05:55
whitequarkhe slept the usual 8 hours or so a day, but in small intervals05:55
whitequarki.e. 2-4-2-4-2-4-2-405:55
whitequarkthere is some evidence that sleeping like that improves productivity significantly05:56
wpwrakyeah, that's somewhat close to what i do. but without pattern. i just sleep when i feel tired.05:56
whitequarkhe kind of confirmed that, but after those 80 days he quit because his wife was objecting05:56
wpwrakyeah, environmental acceptance is a bit of an issue ;-)05:56
whitequarkwell, I did that as well for months, on and off, but there are some problems I'd say05:56
whitequark(env acceptance) hehe05:56
whitequarkfirst, most (if not all) employers aren't satisfied with that05:57
whitequarkso, until I could completely dedicate my time for my own company... probably no luck with that05:57
wpwrakyeah, you need to be your own master for this kind of thing. or be really far from your boss.05:58
whitequarksecond, shops. this is especially painful if you're too lazy to cook for yourself, which I only recently overcame, but applies globally. night-time delivery and 24h shops help, but there are very few instances of both05:59
whitequark(for some reason I tend to be awake mostly at night with this pattern, or at least awake and bored enough to care about something which is not my work)05:59
whitequark(far from boss) well, our company has people working from Thailand and different eastern parts of Russia. here the problem is mainly the irregularity.06:00
whitequarkthird, you're caught in this stupid fragile body which has hormones, and if I'm only awake at night I tend to slip into quite a depressed state, which doesn't help much either.06:01
wpwraknighttime is generally best for work that requires concentration. less disturbances.06:14
wpwrakdunno if there's a strong link between nocturnal activity and depression. maybe it amplified other factors, though.06:15
wpwrakand yes, shops are a problem, particularly if you throw procrastination into the mix :)06:16
whitequarkwpwrak: I wonder why is it so for nighttime (less disturbances)06:29
whitequarkthat is, I find that to be the case too, but as my online activity includes a lot of overseas people, it's not that I distracted less by IRC or something06:30
whitequarkit is a little bit quieter overall, but I doubt that difference between 30dB and 40dB (or something like that) actually helps that much06:31
wpwrakduring daytime, you can get many interruptions, telemarketing, postal service, neighbours ringing your doorbell, people calling the wrong number, and so on.06:31
wpwrakat night, all these are usually asleep and don't bother you.06:32
whitequarkoh, I almost never get those. telemarketing, ewww. that's why I don't have a landline phone.06:32
wpwrakand yes, the ambient noise level surely makes a difference. your rodent brain feels safer if it's quiet and you could hear any predators from very far :)06:33
wpwrakgetting rid of landline helps a lot, agreed06:33
whitequark(as per postal service, I'd be happy if those bastards actually ringed my door. but no, they prefer to slack off and make me walk all the way to the postal building, sigh)06:33
whitequark(rodent brain) makes sense. I'm definitely much more comfortable in a very quiet environment06:34
wpwrakanother factor may be that there are fewer activity choices at night. so you can focus better on the few(er) things available for doing.06:34
whitequarklekernel: talking about moronic comments10:09
whitequarkonce I shared a snippet on russian site with a name roughly translated as "shittycode.ru" [govnokod.ru]10:10
whitequarkthe snippet was an excerpt from Ingenic kernel source with a comment, you might have seen it, // No risk no fun, run with interrupts on!10:11
whitequarkthe rating of the snippet instantly plunged to -20 or so10:11
whitequarkguess how was that motivated? "That's quite good piece of code, it has COMMENTS IN IT [emph. mine]"10:12
whitequarkthey were actually serious about that.10:13
whitequarkwpwrak: talking about russia and stuff. I'm not sure if you realize it or no, but it's perfectly normal for [state-financed] ambulance to refuse to go to an elderly patient with an explicitly stated reason of "we don't want to make our death statistics worse"10:17
mthwhitequark: I still wonder what the "panda out, take care!!!" comment means... (also Ingenic code)12:32
viricit seems like fun, working for Ingenic12:35
larscmth: he will eat all your bamboo!12:41
wpwrakwhitequark: (ambulance) sounds like a very good reason :) someone seems to have created dubious incentives13:17
xiangfuwpwrak, I finish the code on read the xc6slx9 configure register out. but the document confused me.14:19
xiangfuwpwrak, if you have time. please take a quick look at ug380.pdf Page:116.14:20
xiangfu[Shift configuration packets into the CFG_IN data register, MSB first.]. it says [159 Clocks(TCK)]. the correct value should be 111. right? (16 * 7 - 1 = 111)14:21
wpwrakhmm, i never really programmed low-level JTAG, so i'm not sure. the other transfers seem to suggest between n-1 and n TCK for n bits, yes.14:32
wpwrakit's a little weird. the description says "Write two dummy words to the device to flush the packet buffer.". this may explain the extra TCK cycles.14:35
wpwrakbut they may be words "d" and "e"14:37
wpwrakstep 5 is also confusing ... "Shift 32 bits out of the Status register", but only 15 cycles ?14:38
wpwrakseems that you'll have some experimenting to do :)14:39
wpwrakmaybe urjtag sources can help, too14:40
xiangfuI already finish my code. and it's working.14:40
xiangfu111 TCK, and 15 cycles when shift status register.14:40
xiangfuthe status register is 16 bits. :)14:41
wpwrakkewl. then you know all the answers :)14:41
wpwrakmaybe you should send xilinx a patch ;-)14:42
xiangfu(the purpose of the document is confuse you)14:42
wpwrakyeah, so it seems indeed ;-)14:43
xiangfuI will keep the code as it is. (111 TCK). if I meet some error I will try to add some 'dump words'. 14:47
xiangfu("Write two dummy words to the device to flush the packet buffer."). thanks. I ignore this line when read the document.14:47
xiangfu159 - 111 = 48 that is 3 dump words. 14:48
xiangfumaybe one at begin and two at the end. 14:48
xiangfuthe code is here : https://github.com/Wolfgang-Spraul/fpgatools/blob/master/mini-jtag/mini-jtag.c#L216 14:51
wpwrakthe 0x2000 already seems to be the dummy data. see also table 6-1 on page 111.14:51
wpwraki guess page 116 is simply very wrong14:51
xiangfuyes. page 111  is more clear, 2 0xffff at the begin. and 0x2000 after Sync wold. 14:58
qi-bot[commit] Bas Wijnen: use automake (master) http://qi-hw.com/p/iris/d9a122220:20
qi-bot[commit] Bas Wijnen: usb boot working again (master) http://qi-hw.com/p/iris/223ed8620:20
--- Thu Sep 27 201200:00

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