Back
[01:15:08] <rue_shop3> maybe
[03:28:25] <OndraSter> shouldn't be a run cap for a motor be right next to it?
[03:29:04] <jacekowski> no
[03:29:09] <OndraSter> ok
[03:29:30] <OndraSter> because in the washing machine it was on the opposite side
[03:29:30] <jacekowski> there is a lot of different types of motors and some don't even need a cap
[03:29:50] <OndraSter> I am not sure if the cap was for the pump (small) motor or for the main drum
[03:29:54] <OndraSter> the big motor
[03:30:00] <jacekowski> pump
[03:30:24] <OndraSter> I was thinking that the drum one has got cap on its assembly
[03:30:31] <jacekowski> nope
[03:30:35] <OndraSter> none at all?
[03:30:39] <jacekowski> none at all
[03:30:41] <OndraSter> oh
[03:31:40] <jacekowski> that main motor is a AC/DC motor
[03:31:51] <jacekowski> and doesn't need any capacitors or anything
[03:32:13] <OndraSter> I was always a n00b when it comes to motors
[03:32:27] <jacekowski> it has two winding, one on the stator and one on the rotor
[03:32:46] <OndraSter> the pump has got just one from what I saw
[03:33:10] <jacekowski> that's probably typical induction motor
[03:33:13] <OndraSter> yes
[03:33:53] <jacekowski> although, that's not very common setup
[03:34:06] <OndraSter> well it was 26 years old machine :)
[03:34:15] <OndraSter> made in czechoslovakia :D
[03:34:21] <jacekowski> in most cases they use like shaded pole motor or something
[03:39:05] <OndraSter> I need to find a use for it yet thoguh
[03:39:09] <OndraSter> though
[03:39:34] <OndraSter> hmmm PC watercooling? :D
[03:40:40] <OndraSter> it has got big (5cm approx) inlet and small (1.5cm) out
[04:08:41] <kdehl_> Wth is up with ebay, am I the only one having problems viewing pages in chrome?
[04:11:22] <OndraSter> I have got speed issues with youtube
[04:12:47] <OndraSter> I mean, even on 360p it gets stuck
[04:12:52] <OndraSter> and it even switched me to 144p few times!
[04:12:55] <OndraSter> MINECRAFT ON YOUTUBE
[04:14:16] <kdehl_> Ouch.
[04:30:49] <wakko> OndraSter: did you try proxfree ?
[04:38:25] <OndraSter> no
[05:31:04] <thethruthisoutth> h
[06:23:22] <thethruthisoutth> how do you trigger a software reset on an atmel ?
[06:25:11] <RikusW> use the WDT
[06:26:30] <thethruthisoutth> i mean the thing has a reset function
[06:26:51] <RikusW> enable the watchdog timer and let it expire
[06:26:52] <thethruthisoutth> and i don't use wdt now
[06:27:00] <RikusW> it will reset the avr
[06:27:02] <thethruthisoutth> but then it would continuously trip
[06:27:05] <thethruthisoutth> ;/
[06:27:18] <RikusW> iirc it becomes disabled after rest
[06:27:20] <RikusW> reset
[06:27:26] <thethruthisoutth> really?
[06:27:29] <RikusW> you enable in in fw not using fuses
[06:27:29] <thethruthisoutth> thanks then
[06:27:41] <thethruthisoutth> so wdton and i reset
[06:27:58] <specing> No, wdton and you wait 30 ms or so, then resey
[06:27:59] <RikusW> there should be a register somewhere to do that
[06:28:18] <RikusW> so enable wdt and do while(1);
[06:28:22] <thethruthisoutth> specing i don't use wdt now, so it will expire whatever comes :)
[06:28:38] <specing> there is a much better way to reset
[06:28:44] <specing> jump 0
[06:29:03] <specing> oh wait, that wont reset the registers
[06:29:05] <specing> nwm then
[06:29:07] <thethruthisoutth> tried goto restart; in C but gcc don't like it
[06:29:53] <specing> xmega surely has some software reset register, not sure about normal avrs
[06:30:11] <thethruthisoutth> i don't have that yet, but my next device is that
[06:30:22] <thethruthisoutth> this is just atmega
[06:31:01] <specing> I just moved onto ARM
[06:31:34] <thethruthisoutth> i don't need 300MHz core to flash leds
[06:32:42] <specing> Its only 80 MHz
[06:34:10] <thethruthisoutth> my printer has 400MHz one :)
[06:35:50] <specing> PostScript is a beast to render, I know
[06:35:55] <specing> also PDF
[06:36:24] <thethruthisoutth> especially at 1200x1200dpi
[06:37:10] <specing> Though 100+ MHz is Linux range
[06:37:41] <thethruthisoutth> old linux works on a 386
[06:38:04] <specing> except we are talking ARM here
[06:38:08] <thethruthisoutth> they removed support in latest
[06:38:27] <thethruthisoutth> arm is more powerful than 386
[06:38:31] * RikusW just installed iatkos ml2 in virtualbox :)
[06:38:36] <thethruthisoutth> and 386 was like 25MHz
[06:38:37] <specing> depends which one
[06:38:45] <RikusW> megal0maniac_afk: you around ?
[06:40:30] <RikusW> boots very slowly...
[06:41:12] <thethruthisoutth> RikusW virtualization
[06:42:14] <RikusW> XP runs quite fast
[06:43:12] <RikusW> and my CPU does have VTx
[06:48:37] <thethruthisoutth> and you have 4GB ram ?
[06:48:45] <RikusW> 8GB
[06:48:51] <RikusW> 2 ML
[06:48:55] <thethruthisoutth> thath should be fine
[06:49:02] <RikusW> 2GB for ML
[06:49:12] <RikusW> could give it 4 as well
[06:52:24] <thethruthisoutth> how do you usually implement button debounce ? if press release diff greater than t_debounce then cycle event after release? or if no release event after t_debounce then cycle event?
[06:53:02] <h4x0riz3d> to flash a bootloader (hex) on an atmega2561 using usbasp - what do i need?
[06:53:19] <h4x0riz3d> change fuses?
[06:53:25] <thethruthisoutth> yo h4x0riz3d
[06:53:45] <Horologium_> thethruthisoutth, 386sx16 here that still works...16mhz
[06:53:45] <RikusW> BOOTRST yes
[06:53:49] <RikusW> and bootsz
[06:54:20] <thethruthisoutth> Horologium_ ;) you have a local museum for display ?
[06:54:29] <Horologium_> no, just a basement full of junk.
[06:54:53] <thethruthisoutth> you could send me a few HUGE hdds
[06:55:31] <thethruthisoutth> those are treasure for robot projects
[06:55:45] <h4x0riz3d> avrdude –c usbasp –p m2561 –u –U flash:w:blah.hex ?
[06:55:58] <Horologium_> h4x0riz3d, you need programmer hardware(usbasp), programmer software(avrdude usually), the AVR, power, and a way to connect them together....depending on the fuse settings of the AVR you might need an external crystal or clock source but from factory you don't neec that.
[06:56:26] <Horologium_> thethruthisoutth, harddrives that work I keep for fun...ones that don't work get ripped apart for parts.
[06:56:33] <h4x0riz3d> afaik the usbasp supplies power
[06:56:55] <h4x0riz3d> and the actual cpu is soldered on a small board which has a crystal
[06:57:13] <Horologium_> then you got it.
[06:57:28] <h4x0riz3d> is that command gonna work?
[06:57:31] <h4x0riz3d> m2561 <-
[06:58:11] <thethruthisoutth> h4x0riz3d< also your unicode stick will not work with avrdude
[06:58:16] <thethruthisoutth> '–'
[06:58:49] <Horologium_> assuming you have the m2561 in your avrdude.conf file, yeah.
[06:59:01] <h4x0riz3d> unicode stick?!
[06:59:13] <thethruthisoutth> yeah, commands need '-'
[06:59:19] <thethruthisoutth> it is called minus
[07:00:13] <h4x0riz3d> -p m2561
[07:00:16] <MrMobius> Horologium_, i saw some site where a guy breadboarded a 486 for a univerisyt project. i wonder what the newest desktop CPU you could breadboard is
[07:01:02] <theBear> 486 is the last one i can think of in dip
[07:01:12] <Horologium_> 486 in dip?
[07:01:14] <Horologium_> where?
[07:01:14] <thethruthisoutth> '–' is 'e2 80 93'
[07:01:15] <Horologium_> show me!
[07:01:16] <MrMobius> theBear, dip???
[07:01:20] <theBear> least i think there were dip ones, around early dx25 and 33
[07:01:26] <Horologium_> it had pins but not a dip package.
[07:01:36] <thethruthisoutth> '-' is '2d'
[07:01:41] <theBear> hmm, must have 386 + 387 in my head
[07:01:42] <Horologium_> last processor I ever saw in dip was the 68000
[07:01:46] <MrMobius> yeah, i think it was the modern flat square we are used to
[07:01:56] <Horologium_> even the 286 was square.
[07:01:58] <MrMobius> 386 in dip would be cool too
[07:02:17] <Horologium_> although it might have been available in dip package.
[07:02:18] <theBear> i'm sure there were dip 386
[07:02:27] <theBear> and a 387 socket next door
[07:02:41] <MrMobius> my graphing calculator had a 286 compatible but im sure that was smd
[07:02:55] <Horologium_> 8086 and 8088 were dip.
[07:03:08] <thethruthisoutth> breadboard a pentium on an fpga
[07:03:15] <Horologium_> looks like 286 was not
[07:03:15] <theBear> i wonder... 386sx40 is about as fast as they went in clock, 68k revisions (68020 etc) went up to hmm... 30mhz ?
[07:03:21] <Horologium_> and I'm sure none of the newer ones were.
[07:04:16] <MrMobius> and there are also some chips made a lot later that include all the 386 or 486 functionality on a chip but i assume we arent talking about those
[07:04:29] <Horologium_> 68060 went to 75MHz
[07:04:42] <Horologium_> was comparable to the pentiums.
[07:05:14] <Horologium_> then it went coldfire and dragonball
[07:05:22] <theBear> mmm... those pics look familiar, is my memory really that bad ?
[07:05:34] <MrMobius> and that was DIP too? i know cisco used 680x0 variants in their routers for a long time
[07:06:36] <Horologium_> no,,,68060 was not dip..nor the newer ones.
[07:07:24] <Horologium_> 68008 was the last dip in that line.
[07:08:18] <Horologium_> 68008 was to the 68000 what the 386sx was to the 386dx
[07:08:32] <Horologium_> they shrunk and multiplexed the external busses.
[07:10:42] <MrMobius> at one point i wanted to get a DIP 68000 chip to play with but i lost interest before i ever bought one
[07:10:59] <MrMobius> other than some 8051s i want to get, i think i have no nostalgia for outdated chips any more
[07:11:45] <Horologium_> have 8 or 10 of them here.
[07:11:58] <Horologium_> and 8051 type chips are still common and current.
[07:12:11] <Horologium_> they even have higher speed cores that compare with the AVR these days clock for clock.
[07:13:03] <Horologium_> http://www.maximintegrated.com/datasheet/index.mvp/id/4078 like that one...approaching 33MIPS at 33MHZ
[07:13:58] <Horologium_> and that chip has something no AVR has....the ability to execute programs from external ram, flash, eeprom, etc..
[07:14:21] <MrMobius> yes i read up a lot on them. i have an msp430 i am making things with but i think modern 8051s can easily out perform it
[07:15:05] <MrMobius> yeah, that is convenient but all the ones i have seen are limited to 12 cycles per read from external memory so even if the CPU is pipelined the external memory is a bottleneck
[07:16:02] <h4x0riz3d> what has to be changed in order for the usbasp to be able to flash a bootloader hex?
[07:16:11] <Horologium_> h4x0riz3d, nothing.
[07:16:23] <Horologium_> are you getting an error from avrdude?
[07:16:51] <theBear> wow, i didn't know they went as high a 68060
[07:17:00] <Horologium_> a bootloader is, for flashing purposes, just a hex file, like any other, with specific memory locations in upper memory.
[07:17:22] <Horologium_> theBear, there was an amiga extension board that used the 68060 processor I think.
[07:17:31] <h4x0riz3d> Horologium_ yes, well can't the boot sector be "locked" ?
[07:17:33] <Horologium_> I know the 68040 was used in later amigas.
[07:17:37] <h4x0riz3d> Brownout disabled, JTAG enabled, SPI enabled, 4096 words, divide clock by 8, Int RC Osc Startuptime 6CK+65ms
[07:17:46] <h4x0riz3d> these are the "enabled" fuses right now
[07:17:58] <Horologium_> h4x0riz3d, if you enable the lock bits, yes...
[07:18:08] <theBear> Horologium_, probably, i only know them from mac sx/30 and various related series hack-upgrades
[07:18:10] <Horologium_> but you need to enable the bootloader fuse for it to work.
[07:18:11] <h4x0riz3d> which lock bits?
[07:18:22] <h4x0riz3d> i'm lost ;]
[07:19:15] <theBear> from memory if you put a 030 or 040 in a se/30 and maybe handwired an address line or two you could get a little more ram in there and run VERY slow linux :) but i was sure i had seen dip 386 and 486 10 mins ago, so ya know, remember i been on 24/7 painkillers most of the last 2 years :)
[07:19:45] <Horologium_> hmm..does the 2651 have lock bits?
[07:19:50] <Horologium_> heck, haven't worked with that one.
[07:20:05] <h4x0riz3d> i don't know
[07:20:07] <dpy2> http://www.programmableplanet.com/author.asp?section_id=2999&doc_id=261083
[07:20:10] <Horologium_> theBear, never seen either of those in dip.
[07:20:15] <theBear> don't all avrs have at least some lock bits (main progmem read/write at least ?)
[07:20:31] * dpy2 wonders if someone has actually tried that :D
[07:20:43] <Horologium_> tried what dpy2 ?
[07:20:46] <theBear> Horologium_, yeah, and the web agrees with you, like i say, longterm painkillers, my memory seems to be 1/10th imagination at the moment
[07:21:00] <dpy2> put an amiga 500 on that programmable SoC
[07:21:01] <MrMobius> theBear, im googling but not finding them. if you remember where you saw it that would be interesting
[07:21:01] <dpy2> :P
[07:21:07] <h4x0riz3d> i *think* there is some way to protect the bootsector from getting overwritten by the bootloader, which is enabled by default
[07:21:29] <Horologium_> I didn't think the bootloader could write to the boot sector.
[07:21:31] <h4x0riz3d> and *think* this has to be disabled when you want to put a bootloader hex
[07:21:32] <theBear> MrMobius, i googled hard and couldn't see them either, and the pics of square 386 and 486 look familiar
[07:21:49] <Horologium_> not seeing anything in the fuses
[07:21:53] <h4x0riz3d> Horologium_ it sure can, at least on the atmega162 (don't ask me how i know) ;]
[07:21:57] <Horologium_> h4x0riz3d, are you getting an error trying to write it?
[07:22:37] <h4x0riz3d> yes, hold on
[07:22:40] <theBear> musta had the 387's in my minds eye, i KNOW they were dip :)
[07:24:13] <MrMobius> theBear, hehe, yeah it looks like 8086 was the last one to be DIP
[07:24:17] <Horologium_> theBear, that's possible...
[07:24:41] <Horologium_> h4x0riz3d, look in the memory programming section of your datasheet abotu lock bits and the lock byte.
[07:24:48] <Horologium_> it's programmed like the fuses.
[07:25:15] <Horologium_> but if you do a chip erase then they are disabled and should be able to write to the thing.
[07:25:57] <Horologium_> and add -e before the -u in your avrdude command...need to erase the chip before programming it unless it is factory new.
[07:28:57] <h4x0riz3d> i suspect it's factory new
[07:29:06] <h4x0riz3d> but i'm not sure
[07:29:20] <Horologium_> still, should do erase on program.
[07:29:30] <Horologium_> if there is ever something on it you have to.
[07:29:56] <Horologium_> and, really, get rid of the -u
[07:30:18] <h4x0riz3d> okay
[07:30:56] <h4x0riz3d> currently, avrdude 5.11.1 says "no programmer has been specified on the command line or the config file"
[07:31:09] <h4x0riz3d> while the command has -c usbasp
[07:32:06] <Horologium_> then it seems your version of avrdude doesn't have usbasp setup in it.
[07:32:35] <Horologium_> or
[07:32:39] <Horologium_> you are using the wrong -
[07:32:49] <h4x0riz3d> wrong - ?
[07:32:53] <h4x0riz3d> what does that mean?
[07:32:57] <thethruthisoutth> this is rendered ;)
http://upload.wikimedia.org/wikipedia/commons/e/ec/Glasses_800_edit.png
[07:33:00] <theBear> heh, ONE person asking for a pic/reference of one in wikipedia land agrees with my 286-dip memory :)
[07:33:03] <Horologium_> avrdude -c usbasp -p m2561 -e -U flash:w:blah.hex
[07:33:13] <theBear> i've still got that board, if it exists i take a pic nexttime i go to mums place
[07:33:14] <thethruthisoutth> http://www.povray.org/
[07:33:16] <Horologium_> it means you are using the wrong character for -
[07:33:24] <Horologium_> at least, when you posted it to here.
[07:33:41] <theBear> anyone ever seen a 'enterprise 128' "keyboard computer" (c64/bbc/zx80 style) ?
[07:33:47] <Horologium_> you are somehow using a unicode character.
[07:34:01] <Horologium_> it shows up different when I copy and paste it to my command line
[07:34:04] <h4x0riz3d> oh?
[07:34:08] <h4x0riz3d> wtf
[07:34:09] <theBear> i got one of those YEARS ago, and have never found a reference or heard of ANYONE that remembers them
[07:34:49] <Horologium_> when I copy your post to my command line the - shows up as a long bar where when I copy mine over it shows up as a short bar.
[07:35:30] <Horologium_> on my keyboard the - is on the same key as the _
[07:35:39] <Horologium_> or the minus key on the keypad.
[07:35:39] <theBear> heh, povray still around eh ? i first got that on a magazine-cover-disc slightly before 486s existed i think
[07:35:47] <theBear> i bet it renders a bit faster these days :)
[07:35:48] <Horologium_> theBear, same here.
[07:36:02] <theBear> :)
[07:36:03] <Horologium_> povray rocks.
[07:36:09] <theBear> it is pretty special
[07:36:12] <Horologium_> I remember when it took days to render something.
[07:36:43] <theBear> yerp
[07:37:13] <Horologium_> and remember going from 386 to 486 and being amazed it would render so much faster.
[07:37:25] <Horologium_> I used that for testing relative speeds on processors.
[07:37:47] <theBear> yeah, weren't many options for running a cpu flatout back then
[07:38:43] <theBear> and you couldn't use most games, they run 'ok' on the old machine, on the new one yer little guy runs off the screen and dies in an instant before you can get near the keyboard :)
[07:40:09] <MrMobius> theBear, wasnt that around the time they introduced the turbo button?
[07:40:36] <Horologium_> MrMobius, turbo button was available on 286
[07:40:43] <abcminiuser> Ok, who would want a widget that plugs into an Android phone, and gives you a USART/SPI/I2C bidirectional terminal?
[07:40:49] <abcminiuser> Kinda like a bus pirate?
[07:40:53] <Horologium_> abcminiuser, me!
[07:41:00] <Horologium_> only, android tablet, not phone.
[07:41:18] <Horologium_> but, same smell either way.
[07:41:24] <abcminiuser> s/phone/device/1
[07:41:27] <theBear> MrMobius, that's when it was most obvious, it's also when the 286 i had was the only one in existance without even the option to connect one :)
[07:41:49] <MrMobius> theBear, what do you mean the only one in existance?
[07:42:10] <Horologium_> gotta love those chips that turned a 386 into a 486...the 486DLC chips.
[07:42:33] <theBear> only one i ever saw/heard of ... there was no turbo button, no keyboard hotkey, no way to slow it from 8 (or was it 12mhz?)
[07:42:43] <Horologium_> 486 chip with cut down external bus that would plug into a 386DX socket.
[07:43:00] <theBear> i liked my amd5x86 or whatever they called them, turn a 486 into something useful :)
[07:43:28] <Horologium_> same thing...cut down external bus that plugged into predecessor board.
[07:43:39] <theBear> first proc i'd ever seen that needed a heatsink too :)
[07:43:58] <Horologium_> gave you some speed upgrade but not as much as the full motherboard replacement.
[07:44:41] <Horologium_> and people(even computer techs) couldn't grasp the DX/SX for 386 was not the same as for 486....they meant different things.
[07:44:55] <h4x0riz3d> okay, now i re-wrote the avrdude command by hand and it doesn't contain any weird chars
[07:44:58] <h4x0riz3d> avrdude: verification error, first mismatch at byte 0x1e000
[07:45:15] <h4x0riz3d> it writes the bootloader hex, but.. actually it doesn't ;]
[07:45:21] <h4x0riz3d> it's like locked
[07:45:30] <Horologium_> they also couldn't understand how my 486DX-50 could outperform their 486DX2-100
[07:45:45] <Horologium_> then unlock it.....look at the datasheet about lock bits...
[07:46:30] <Horologium_> the lock byte is separate from the fuses.
[07:47:21] <Horologium_> but if you put the -e in there then the lock bits should be reset.
[07:47:28] <h4x0riz3d> oh?
[07:48:20] <h4x0riz3d> with the command i pasted, avrdude says: avrdude: NOTE: FLASH memory has been specified, an erase cycle will be performed. To disable this feature, specify the -D option.
[07:48:36] <h4x0riz3d> isn't that the same as specifying -e ?
[07:49:01] <Horologium_> that's flash memory erase.
[07:49:04] <Horologium_> the -e erases the whole chip.
[07:49:08] <h4x0riz3d> ah
[07:49:10] <h4x0riz3d> okay
[07:49:13] <Horologium_> which should clear any lock bits..
[07:49:20] <Horologium_> but it doesn't affect the fuse bytes.
[07:50:18] <Horologium_> and, time to go do something.
[07:50:20] <h4x0riz3d> well, i don't know what is preventing it from writing the bootloader hex
[07:50:31] <h4x0riz3d> is it lock bits or fuses
[07:50:53] <h4x0riz3d> last time i tried this i managed to brick the cpu :/
[07:51:09] <h4x0riz3d> so i'm afraid of touching the fuses ;]
[08:00:30] <thethruthisoutth> wow
http://www.mylinuxsoftware.com/linux_games.html thre are some
[08:08:20] * RikusW is thinking unity is an osx imitation...
[08:34:49] <Horologium_> I'm guessing the avr xml database can be found in avrstudio?
[08:35:05] <Tom_itx> probably
[08:38:18] <nevdull> Horologium: i've been parsing/converting those xml device specifications into SDL
[08:38:35] <nevdull> they're in $(ATMEL_STUDIO_HOME)/devices
[08:47:27] <Horologium_> into SDL?
[08:47:32] <Horologium_> simple directmedia layer?
[08:47:36] <Horologium_> or is there another sdl?
[08:50:37] <nevdull> Simple Declarative Language
[08:50:55] <nevdull> it is sort of a cross between YAML and JSON but much lighter than xml
[08:53:07] <Horologium_> far too many programming languages and extensions and layers these days.
[08:53:13] <nevdull> for sure
[08:53:37] <Horologium_> I remember writing in C something the professional programmers in the company were struggling to do in PERL.
[08:53:43] <Tom_itx> none seem any better than 10 yrs back either
[08:54:00] <specing> :3 lua
[08:54:03] <Horologium_> it was to parse a bunch of files and extract and reformat data.
[08:54:20] <Horologium_> they fucked with it for a week.
[08:54:27] <Horologium_> I wrote a parser in a couple of hours.
[08:54:53] <specing> then they were fired and you became the only programmer?
[08:54:56] <Horologium_> and it was cross platform capable...just compile it on whatever, linux, windows, mac, etc with your favorite compiler.
[08:55:03] <Horologium_> no, I was the IT manager.
[08:55:19] <Horologium_> that company is now defunct....
[08:55:41] <Horologium_> good idea behind it but they hired idiot programmers who couldn't follow directions and tried hardcoding things that should have been dynamic.
[08:55:48] <Horologium_> amongst other crap they did.
[08:55:52] <Horologium_> the code never worked right.
[08:57:21] <Horologium_> the idiots were hardcoding state machines to try to define a dynamic changing environment.
[08:59:40] <thetruthisoutthe> are you talking about haskell ?
[08:59:56] <Horologium_> nope.
[09:00:15] <Horologium_> I program in C almost exclusively but can do many other languages.
[09:00:35] <thetruthisoutthe> i like C too, and compile everything with gcc
[09:02:32] <Horologium_> mostly I was just chattering.
[09:02:53] <Horologium_> about the plethora of programming languages and stuff out there.
[09:03:04] <thetruthisoutthe> h4x0riz3d cool nick
[09:03:08] <Horologium_> and the fact that the old tools work just fine if you know how to use them.
[09:03:45] <thetruthisoutthe> make && make install - compile and burn atmel :)
[09:04:00] <Tom_itx> new generations have better ideas (so they think) so create better tools
[09:04:17] <Tom_itx> if they would realize their flaws we as a ppl would progress quicker
[09:04:20] <Horologium_> Tom_itx, mostly they add complexity layers
[09:05:44] <thetruthisoutthe> don't you just hate it when some program requires odd lame libraries that are probably full of bugs too ?
[09:05:52] <twnqx> Horologium_: my boss was just complaining i chose perl to do something (after we discussed i could use whatever i wanted)
[09:06:04] <Horologium_> no reason why a 3GHz computer should take longer to boot to a usable state than a 386sx20 did in the 90s.
[09:06:20] <thetruthisoutthe> twnqx choose lisp then ? :)
[09:06:30] <twnqx> no, i was supposed to use python
[09:06:35] <twnqx> i refuse to use python.
[09:06:40] <Tom_itx> i can boot dos and be halfway done doing what i was gonna do before windows gives me a splash screen
[09:06:40] <thetruthisoutthe> mee too
[09:06:41] <Horologium_> or use brainfuck
[09:07:16] <Tom_itx> i'd say the same for linux but i don't know it as well
[09:07:23] <twnqx> "why didn't you use python, the almost working example code we had was python too, noone else here knows perl"
[09:07:27] <thetruthisoutthe> Tom_itx that is fun, it is in the bootsector i think :)
[09:07:45] <thetruthisoutthe> MBR
[09:07:54] <twnqx> "because i don't know python, i hate python's structuring with a passion, and you told me i can use whatever"
[09:08:22] <Horologium_> linux in bios rocks!
[09:08:31] <Horologium_> command prompt in,,,3 seconds.
[09:08:31] <theBear> well, without real-world/actual examples, in theory a modern machine could/should take more boot time... consider that it covers all the legacy (well, a lot of it) devices/init, newer ones (hid, mouse maybe (i know, hid still), network card and various post-init stuff on it) and a whole bunch of other stuff that didn't exist in 'the olden days', timeouts for above, perhaps some things can't init/timeout in parallel for various reasons
[09:09:01] <thetruthisoutthe> Horologium_<= but linux does not configure your motherboard ;/
[09:09:16] <Horologium_> theBear, ok, go from end of bios init to functional...windows still takes longer to boot with each generation.
[09:09:22] <theBear> in the old days the slowest bit was probably reading the eprom, maybe a quick tickle of the disk controllers, scan a meg or two for option roms, *BAM*
[09:09:48] <theBear> Horologium_, i believe you, i noticed this many years ago and chose to go down a less-apparently-declining road :)
[09:10:15] <twnqx> Horologium_: efi kernel
[09:10:24] <twnqx> bios loads kernel and runs it - no bootloader or anything in between
[09:10:57] <Horologium_> wish I had the skill to design and build my ultimate computer.
[09:11:07] <twnqx> i wish i had the time and money
[09:11:39] <Horologium_> quad core processor with like 4 FPGAs on the board for custom processing.
[09:11:48] <Horologium_> and OS boot from flash.
[09:11:56] <twnqx> the problem is
[09:12:07] <twnqx> you don't have memory mapped flash these days
[09:12:14] <twnqx> even BIOSes are SPI connected
[09:12:22] <thetruthisoutthe> Horologium_:= you mean a computer with 3 cpus, and 9 ddr3 memories, and 18 sata 3 connectors, and 48 usb connectors, and 2 pc speakers, and dual link optical fiber integrated ethernet, and dual dvi monitor ?
[09:12:23] <Horologium_> can still find parallel flash.
[09:12:41] <twnqx> also, just plug the fpgas into pcie
[09:12:51] <twnqx> i'd rather say
[09:12:52] <Horologium_> thetruthisoutthe, no, a computer with on the fly programmable processors..
[09:12:57] <thetruthisoutthe> ohh
[09:13:01] <thetruthisoutthe> that is fpga card
[09:13:02] <Horologium_> so I can build the kernel in "hardlyware"
[09:13:21] <twnqx> mostly pointless, by the way
[09:13:26] <thetruthisoutthe> yes :(
[09:13:27] <Horologium_> probably.
[09:13:29] <Horologium_> but still.
[09:13:35] <Horologium_> ok..gotta go do things again..laters.
[09:13:40] <twnqx> there are very few things left where fpgas outmuscle modern cpus :/
[09:13:47] <thetruthisoutthe> now you want an industrial process control machine, or something to warez music from torrent
[09:13:54] <theBear> twnqx, now and then bios is shadowed 'instantly', it may as well be direct mapped to a prom
[09:14:04] <twnqx> sure
[09:14:06] <Tom_itx> theBear, are we better off because of all those devices though?
[09:14:11] <theBear> heh hardlyware, like my pants
[09:14:12] <twnqx> that would be my approach
[09:14:32] <twnqx> theBear: put a way larger bios chip in, put kernel to it
[09:14:43] <theBear> Tom_itx, heh, that wasn't the question :)
[09:14:43] <twnqx> have it shadowed by the chipset
[09:14:51] <thetruthisoutthe> twnqx:= real time processing? parallel processing of large amounts of data ?
[09:15:08] <thetruthisoutthe> 1ns latency
[09:15:08] <twnqx> yeha, that kind of stuff
[09:15:16] <twnqx> forget 1ns
[09:15:37] <twnqx> i am always happy when my fpgas reach 40-50mhz
[09:15:39] <thetruthisoutthe> ok 5ns pin-to-pin
[09:15:53] <twnqx> 5ns pin to pin, lol
[09:15:58] <thetruthisoutthe> :)
[09:16:00] <theBear> until very recently you still had parallel flash chips, solder in a big one and off yer go
[09:16:01] <twnqx> yeah, without any processing maybe
[09:16:19] <twnqx> it's easier to solder larger nor flashs
[09:16:31] <twnqx> pin compatible with the smaller, can replace them on any modern mainboard
[09:17:39] <thetruthisoutthe> my biggest problem is what should i get for dinner
[09:18:10] <theBear> i been messing with my 'new' core2/q35 based board recently, various quirks and issues and actual damage/problems specific to this one, but tbh i can't even workout which chip IS the flash, i gotta get a torch and look up some numbers, tho hardly worth it, they all bga or somthing i know even less about
[09:18:34] <twnqx> look for something with a 25x on
[09:18:35] <thetruthisoutthe> theBear< deadbug style bga is c00l
[09:18:38] <twnqx> or
[09:18:47] <twnqx> try to get software to do it
[09:18:54] <thetruthisoutthe> just solder cat5e wires to the balls
[09:19:02] <twnqx> there's a flash software for linux that can detect the chip on most boards
[09:19:30] <Tom_itx> personally i don't care what you hang on your balls
[09:19:38] <theBear> thetruthisoutthe, yeah, but not ideal in the middle of my 'newest' pc :)
[09:20:16] <thetruthisoutthe> theBear:= no problem, just plot some transparent 2 component epoxy on it when you are done
[09:20:50] <twnqx> http://imageshack.us/a/img96/9854/pirateh.jpg
[09:20:59] <twnqx> you can always do something like this with your flash :P
[09:21:04] <thetruthisoutthe> so you can show guys your cool hack after it cures
[09:21:17] <theBear> i'm not sure i'll ever get it going completely, since i got it i noticed the guy i got it off has a bad habit of 'abusing' pc hardware, stacking and bumping it etc etc... among various other things, half the power/startup issues (i don't have a fancy modern atx2.3 supply, but i did 'upgrade' a slightly older one <grin>) went away when i resoldered the atx connector, and a couple inductors and caps for good luck :)
[09:21:58] <twnqx> lol
[09:22:32] <theBear> i'm tempted to do something to the inductor on the graphics card too, input inductor so small (and no external power input) that i suspect the chip on top might be just enough to be faulting the psu/power during 9/10 boots, got it starting 10/10 without that card now, and it aint a super-power-hog one
[09:22:49] <thetruthisoutthe> :)
[09:23:03] <thetruthisoutthe> an LC filter helps to suppress transient loads to the ps
[09:23:27] <theBear> then the little things like early hybrid-uefi bios that they 'forgot' to include an hid driver in, and of course the ps2 ports aren't populated :)
[09:26:12] <theBear> thetruthisoutthe, yeah, and modern pcs are full of them from primary side of the psu right up to well, the graphics card etc, this little maybe 1uH or something has perhaps 1/10th of the iron on one side cracked off, and with the upgraded slightly small psu, could be just enough :) either taht or there maybe crackd joints in the pcie slot, dunno how i feel about resolding that many pins on a big thick heavy copper mobo :)
[09:27:06] <thetruthisoutthe> you apply flux, put in oven
[09:27:29] <thetruthisoutthe> if plastics can handle it it will be good
[09:27:46] <thetruthisoutthe> oh and caps may explode
[09:28:04] <theBear> mmm, i'm gonna test/experiment with that whole concept a bit on a dead netbook i was given to save it from the bin first... i can almost make it finish post if i take the mobo out and squeeze the right chips :)
[09:28:23] <Tom_itx> thetruthisoutthe, chinese plastic probably won't
[09:28:33] <thetruthisoutthe> problem with leadfree is joints cracking
[09:28:40] <theBear> i done a bunch of math, read some of the non-idiot reports on the web, i think i can pull it off on the netbook, haven't looked at this bigger board 'like that' (temp wise)
[09:28:56] <thetruthisoutthe> you may n eed to apply lead solder too
[09:29:06] <theBear> thetruthisoutthe, yeah, not so common in big thick mobos, but like i said, i think this one was abused before i got it
[09:29:41] <theBear> mmm, i resoldered all the stuff yesterday with lead, whatever solder they used isn't TOO bad, most of them just flux and leaded was enough to make good joints, couple i needed to remove some of the old solder to avoid cloudyness
[09:30:16] <thetruthisoutthe> it will most likely mix well
[09:31:39] <theBear> last year when i was working i was doing warranty for EVERYONE audio, had more leadfree repair experience than all my other years put together, a LOT of the more modern ones don't mix so well
[09:32:02] <theBear> or you gotta add a LARGE % of leaded with flux to 'turn' them
[09:32:16] <thetruthisoutthe> ou only need flux for leadfree
[09:32:52] <thetruthisoutthe> without flux leadfree will not flow
[09:33:08] <Tom_itx> nor will leaded
[09:33:28] <thetruthisoutthe> leaded will still flow
[09:33:28] <theBear> wtf are you talking about ? of course you need flux for leaded
[09:33:56] <theBear> no it won't ! it won't wet, it won't attach to even the slightest corrosion, it'll form dry nasty joints, there's a ton of reasons you need flux
[09:34:13] <thetruthisoutthe> i mean you can heat leadfree until the board catches fire and it will not flow
[09:34:24] * Tom_itx thetruthisoutthe looses 1 credit point
[09:34:45] <thetruthisoutthe> it won't wet, but will flow while leadfree wont
[09:34:58] <theBear> if you heat leaded that long it won't either, the longer you heat it without flux, the more 'objectionable' the solder bcomes
[09:35:43] * theBear considers the difference between terms wet and flow re: soldering
[09:36:31] <theBear> the flux makes it behave, the MOMENT a joint stops smoking (while hot) it starts to get worse and worse
[09:38:15] <megal0maniac> Leadfree is horrible
[09:38:25] <megal0maniac> BUT, it doesn't give me headaches :P
[09:38:50] <megal0maniac> Have a 120mm extractor fan in the pipeline. University holidays...
[09:39:03] <theBear> indeed, some more than others, but i would never use it unless i absolutely had to (for example a quick or complex repair on a board/existing solder that just doesn't mix with leaded)
[09:39:20] <megal0maniac> theBear: Chinese leadfree
[09:39:29] <megal0maniac> And the properly cheap stuff
[09:41:04] <theBear> megal0maniac, mmm, i find chinese is no different, tho they do seem to have 'bad days' in the factory sometimes, whole unsold/warranty units where it's like the solder was made with an ingredient missing or something, and i'm only familiar with a couple of alloys i've used, but there are probably 5 or 8 'common' alloys
[09:41:46] <thetruthisoutthe> Tom_itx:= i don't feel like losing anything, i just don't have flux because i do not repair
[09:42:28] <theBear> chinese stuff is actually much better than it used to be, sure it's made to a price point etc etc, but it compares nicely with other countries these days in new stuff... remember back to the future how much the 1955 doc couldn't believe that japanese made the vidcam ... i figure china now is kinda like that, do something enough, you get good at it
[09:42:57] <theBear> they ain't kicking arse in the economy scoreboards 'cos they keep making shit year after year...,
[09:43:17] <thetruthisoutthe> chinese things can get better if you give them plans to make your stuff, then they make new factory and make it rebranded
[09:47:26] <megal0maniac> And the prices are impossibly low
[09:48:15] <megal0maniac> It costs me less to get a fabricated board, with the ICs and supporting circuitry, shipped to my local post office, than to buy only the ICs locally.
[09:48:45] <RikusW> hi megal0maniac
[09:48:45] <RikusW> seems iatkos ml2 works in virtualbox 4.2.12 :)
[09:48:45] <RikusW> though quite slowly
[09:48:45] <RikusW> especially visual effects
[09:49:04] <theBear> impossibly ! even power tools that are basically coils and iron squeezed in a moulded plastic case seem to get made and sent around the world and marked up and distributed and marked up and sold for less than the metal would cost !
[09:49:26] <theBear> and they work as well and last as long as big old expensive ones !
[09:49:38] <RikusW> megal0maniac: seems iatkos ml2 works in virtualbox 4.2.12 :)
[09:49:50] <RikusW> seems my ADSL line isn't quite ok...
[09:50:43] <megal0maniac> RikusW: Nice. Now install it on a hdd :P
[09:51:01] <RikusW> /ping 3s :S
[09:51:18] <RikusW> I don't have free partitions...
[09:51:42] <RikusW> Now I want xcode on it
[09:51:49] <RikusW> another 1.6GB download....
[09:53:52] <RikusW> seems to me like unity is an osx imitation....
[09:54:36] <megal0maniac> It basically is
[09:59:30] <megal0maniac> Any obvious reason that this doesn't work will pull-down resistors, only pull-ups? I'm using 5V with 10K resistors
http://i.imgur.com/leiIvCD.jpg
[10:05:56] <thetruthisoutthe> megal0maniac<= dont uuse pulldowns unless it needs to bleed
[10:06:16] <thetruthisoutthe> and ttl ics pull up >0.5mA.
[10:06:36] <thetruthisoutthe> they need no input termination
[10:07:58] <megal0maniac> Oh.. Good to know. What do you mean by bleed?
[10:08:46] <thetruthisoutthe> like prevent accidental drive of mosfet gate when power is insufficient for the microcontroller
[10:09:02] <thetruthisoutthe> or when tristated during reset
[10:09:59] <megal0maniac> Which is not something I'd need to worry about using discreet (74xx) logic, right?
[10:10:10] <megal0maniac> i.e. Don't use pull-down for discreet logic
[10:10:15] <thetruthisoutthe> unless...
[10:10:28] <thetruthisoutthe> cmos version of 74xxx
[10:10:35] <thetruthisoutthe> those need termination
[10:11:02] <megal0maniac> It's all ALS/LS, which iirc is TTL
[10:11:07] <thetruthisoutthe> your design works with 74hc type for example
[10:11:24] <megal0maniac> Gotcha
[10:14:38] <megal0maniac> thetruthisoutthe: Just to clarify, an AND gate with floating inputs will produce a high output?
[10:17:57] <RikusW> zlog
[10:25:03] <RikusW> megal0maniac: oddly enough it the dmg didn't boot the first time, only after a few VB resets did it work... no idea why
[10:31:34] <megal0maniac> I didn't have much patience, I must admit. I'd much rather have it running natively
[10:37:45] <megal0maniac> I am amazed at how frequently packages for debian unstable are updated
[10:37:59] <megal0maniac> I'm pulling like 10+mb in packages every day
[10:38:14] <Tom_itx> it must be unstable
[10:38:18] <megal0maniac> Also, it's quite stable :P
[10:38:40] <Tom_itx> freenode is rather unstable too
[10:38:53] <megal0maniac> It heard me..
[10:40:22] <megal0maniac> Nah, it's just cool to see such active development
[11:15:44] <megal0maniac_afk> nafk
[11:16:06] <megal0maniac> Stupid keyboard
[11:52:34] <OndraSter> guys, anybody with recent LG washing machine? If we power it on, start it, the drum starts rotating but there is no water coming in...
[11:52:50] <Tom_itx> no sorry
[11:53:00] <Tom_itx> is the faucet on?
[11:53:02] <Tom_itx> :)
[11:53:14] <Tom_itx> or plugged
[11:57:32] <RikusW> OndraSter: I've got ML2 running on VBox
[12:00:30] <OndraSter> Tom_itx, of course
[12:00:34] <OndraSter> RikusW, ml2?
[12:00:44] <OndraSter> washing amchine software?
[12:00:55] <RikusW> mountain lion
[12:01:04] <RikusW> iatkos
[12:01:55] <OndraSter> oh
[12:02:39] <RikusW> graphics effects are slowish...
[12:44:03] <thetruthisoutthe> h
[12:44:25] <thetruthisoutthe> megal0maniac_afk<= if it is TTL then yes.
[12:44:56] <thetruthisoutthe> Tom_itx<= freenode is being ddosed ;)
[12:45:39] <thetruthisoutthe> OndraSter<= probably a firmware bug :)
[12:46:21] <thetruthisoutthe> new washing machines are made to die, and you need to buy another, or at least another control board
[12:48:02] <OndraSter> :)
[12:48:05] <OndraSter> this one is brand new
[12:48:14] <OndraSter> we paid even extra for 5 years warranty
[12:49:05] <thetruthisoutthe> haha
[12:49:13] <thetruthisoutthe> then claim your warranty
[12:49:46] <OndraSter> well, it might be what it is supposed to do
[12:49:55] <OndraSter> but of course their helpline does not work on weekends
[12:50:19] <thetruthisoutthe> there is always tomorrow
[12:50:25] <OndraSter> yea
[12:55:18] <specing> rule #1 of consumer devices: everything breaks on friday afternoon.
[13:03:52] <thetruthisoutthe> saturday night:?
[13:05:56] <Horologium_> if anything can go wrong it will, at the worst possible moment.
[13:06:25] <thetruthisoutthe> 8 bit mcus like uint64_t math? :)
[13:06:31] <thetruthisoutthe> using gcc
[13:06:33] <Horologium_> they can do it.
[13:06:46] <Horologium_> it's not exactly fast, but depending on what you are doing, should be sufficient.
[13:07:04] <thetruthisoutthe> :) okey
[13:07:21] <Horologium_> it's all in the libs...if the math libs and whatnot in your compiler support it then it will work.
[13:07:39] <thetruthisoutthe> i have some avr-gcc
[13:08:01] <thetruthisoutthe> avr-gcc (Fedora 4.5.0-2.el6) 4.5.0
[13:08:02] <thetruthisoutthe> Copyright (C) 2010
[13:08:04] <thetruthisoutthe> fairly new
[13:08:35] <Horologium_> 2 versions old compared to mine.
[13:08:48] <Horologium_> gcc version 4.7.2 (GCC)
[13:08:59] <thetruthisoutthe> hmm, do i miss much functionality?
[13:09:04] <Horologium_> probably not.
[13:09:13] <thetruthisoutthe> maybe larger code
[13:09:28] <Horologium_> this is just what installs by default on my debian install.
[13:09:35] <thetruthisoutthe> wow
[13:09:41] <thetruthisoutthe> you have avr-gcc by default?
[13:09:52] <Horologium_> fedora is the red headed stepchild of redhat which sold out to microsoft, after all.
[13:10:01] <Horologium_> when I apt-get install avr-gcc, yeah
[13:10:20] <Horologium_> by "default" I meant the default in the repositories.
[13:10:23] <thetruthisoutthe> yeah i know fedora is the test version of redhat
[13:11:19] <Horologium_> I used redhat back in the day..it was the shiznit...bleeding edge and things worked.
[13:11:38] <Horologium_> then they spun off fedora and things went to hell in a dvd case.
[13:11:45] <thetruthisoutthe> redhat is always outdated, meant to be used in servers
[13:11:55] <Horologium_> didn't used to be.
[13:11:56] <thetruthisoutthe> they try to patch security holes
[13:12:20] <Horologium_> but, that was,,,,12ish years ago.
[13:12:28] <Horologium_> maybe 10ish..
[13:12:30] <Horologium_> something like that.
[13:12:50] <thetruthisoutthe> oh ok, i was thinking about 205-2010
[13:13:00] <thetruthisoutthe> 2005
[13:13:36] <Horologium_> I left redhat just after they spun off fedora...think it was 2002 or 2003 maybe.
[13:13:39] <thetruthisoutthe> the RHEL4 series was stable, solid as rock
[13:13:56] <thetruthisoutthe> ofc they sell it for like $2k
[13:14:56] <thetruthisoutthe> i tried an AS enterprise version for mission critical applications
[13:15:38] <thetruthisoutthe> bad motherboard? no problem
[13:15:40] <thetruthisoutthe> ;>>
[13:15:53] <thetruthisoutthe> while the cpu clock was ticking, the system was up
[13:19:00] <Horologium_> for rock solid stability in a file server I still prefer netware 3.12
[13:19:08] <Horologium_> have never found anything that can beat it.
[13:20:19] <thetruthisoutthe> or that thing with zfs :)
[13:20:33] <thetruthisoutthe> that is my choice
[13:20:53] <thetruthisoutthe> zfs+raidz
[14:08:34] <rue_more> http://hackaday.com/2013/05/10/microcontroller-enumerates-as-usb-printer-can-be-programmed-by-printing/
[14:08:43] <rue_more> abcminiuser, good work!
[14:11:59] <thetruthisoutthe> :)
[14:12:11] <thetruthisoutthe> not a bad idea
[14:12:30] <thetruthisoutthe> enumerate as printer and print as a cnc laser cutter or mill
[14:13:23] <thetruthisoutthe> here is some circuit analysis against boredness:
http://www.analyzemath.com/applied_mathematics/electric_circuit_1.html http://users.ecs.soton.ac.uk/mz/CctSim/chap1_6.htm
[14:23:28] <abcminiuser> Cheers rue_more
[14:23:37] <abcminiuser> For some reason it's all gotten attention this week
[14:26:12] <megal0maniac> abcminiuser: "Just added"? :P
[14:27:35] <abcminiuser> Hva?
[14:29:32] <megal0maniac> Well it's been around for a good month, if not more, no?
[14:31:16] <megal0maniac> But congrats on the lack of haters. :)
[14:38:34] <abcminiuser> Yeah, it's been there for a while
[14:38:49] <abcminiuser> A ton of comments were all "Make a Mass Storage version"
[14:38:54] <abcminiuser> It's LITERALLY the next post
[14:39:11] <megal0maniac> :)
[14:39:17] <megal0maniac> Saw you replied to that
[14:39:38] * RikusW just compiled some Xlib stuff on OSX :)
[14:39:40] <megal0maniac> Interesting to read that some printers actually receive firmware like that
[14:39:51] <RikusW> using XQuartz
[14:49:01] <abcminiuser> I think I need a ALL CLASSES ARE NOT HID blog PSA
[14:52:57] <megal0maniac> Some whiney buggers on hackaday
[14:54:06] <abcminiuser> Mostly people who didn't read the next article I posted :P
[14:54:19] <abcminiuser> I guess I need to write a new blog post on the migration to Git
[14:54:23] <abcminiuser> Answer my emails
[14:54:38] <abcminiuser> Make a getting started LUFA screencast and upload it to YouTube
[14:55:01] <abcminiuser> Then start on this Android Accessory Host class terminal/bitbanger/analyzer/sig gen project
[14:55:08] <abcminiuser> Balls.
[14:56:10] <megal0maniac> It's that time of year
[14:56:11] <megal0maniac> Exams in a week
[14:56:39] <megal0maniac> Busy trying to understand what a BJT transistor is, what it does and why you'd want one
[14:56:48] <megal0maniac> And this is so not my thing
[15:02:20] <megal0maniac> abcminiuser: Starting development for Android seems like a can of worms
[15:02:36] <abcminiuser> It's OK, I've done some simple stuff in the past
[15:03:44] <RikusW> does anyone know how to the the default include dirs used by gcc ?
[15:03:53] <RikusW> and lib dirs too
[15:04:07] <megal0maniac> No
[15:04:12] <megal0maniac> BUT
[15:04:27] <megal0maniac> I've been finding find | grep *something* very helpful lately
[15:04:28] <RikusW> I know its added by -L and 0I
[15:04:30] <RikusW> -I
[15:04:41] <RikusW> env doesn't help me much
[15:04:46] <RikusW> apart from PATH
[15:04:50] <megal0maniac> On osx now?
[15:04:57] <RikusW> Mint
[15:05:07] <megal0maniac> Oh, then definitely no
[15:05:14] <RikusW> pgup/dn don't work in osx vim :S
[15:05:26] <RikusW> very annoying
[15:05:26] <megal0maniac> Install macports
[15:05:32] <RikusW> ah
[15:08:16] <RikusW> its compiling alright, just had to mess with -L and -I a bit
[15:14:45] <thetruthisoutthe> megal0maniac_afk<= if it is TTL then yes.
[15:15:31] <thetruthisoutthe> inuts will source current
[15:23:30] <megal0maniac> thetruthisoutthe: Thanks
[15:23:39] <megal0maniac> It's all working now :)
[15:23:53] <megal0maniac> Goodnight all.
[16:06:28] <Qantourisc> I'm looking at ATMEGA8-16PU but I can't quite find the UART pins ?
[16:09:41] <Qantourisc> also am i seeing it right that there are 6 ADC i can switch between I/O and ADC?
[16:11:40] <twnqx> pin 2/3 PD0RTXD and PD1/TXD
[16:11:49] <twnqx> pin 2/3 PD0/RXD and PD1/TXD are the uart pins
[16:12:25] <twnqx> and you can select one of the 6 so actually sample, there is one ADC on the chip with several selectable inputs
[16:15:49] <Qantourisc> twnqx: didn't you need 3 pins for UART ? or do you just assing a few extra pins and use those to control the read and send requests ?
[16:16:08] <twnqx> uart needs 2
[16:16:28] <twnqx> if you want flow control you need more, but you just use some IO pins, indeed
[16:17:10] <twnqx> (by uart i assume you mean rs232, not some other form of serial communication)
[16:17:24] <Qantourisc> i'll be going rs485 :)
[16:18:57] <tech2077> then that requires bit banging or rs232 to rs485 adapters
[16:19:34] <twnqx> yeah
[16:19:43] <tech2077> or a SPI to RS485 like the MAX490
[16:19:43] <Qantourisc> tech2077: or just using the UART and using an UART to RS485 + settting the send/receive pins with the IO ?
[16:19:57] <Qantourisc> or SPI if that's easy to use sure
[16:20:21] <tech2077> the max490 is probably the simplest way
[16:20:45] <tech2077> is i believe it does flow control automatically
[16:21:05] <Qantourisc> "automaticly" ?
[16:21:18] <Qantourisc> as in when you cram bytes it sends ?
[16:21:22] <Qantourisc> and when you read it reads ?
[16:21:28] <tech2077> yeah
[16:21:55] <tech2077> not sure, that's just what it appears like, haven't used rs485 before, might want to check out the datasheet
[16:22:36] <Qantourisc> tech2077: the max490 is not SPI is it ?
[16:23:10] <Qantourisc> tech2077: RS485 is like RS223 but without any flow control, and with "high impedance" mode
[16:23:12] <tech2077> hmm, my mistake
[16:23:57] <Qantourisc> but setting a bit to notify If i want to read or write is kinda ok for me
[16:24:14] <tech2077> http://www.maximintegrated.com/datasheet/index.mvp/id/2033
[16:24:29] <tech2077> that's a SPI to RS485 that looks good
[16:25:44] <Qantourisc> and one i Can't get easily :( /me check datasheet
[16:28:15] <Qantourisc> hmmm not sure
[16:28:20] <Qantourisc> looks more complicated :/
[16:30:39] <tech2077> rs232 to rs485 doesn't seem like the simplest conversion
[16:30:52] <Qantourisc> tech2077: didn't say that
[16:31:20] <Qantourisc> Just ook an MAX490(or equavilant) to the UART and control the send/receive status by IO pins
[16:31:35] <Qantourisc> that should do the trick
[16:31:38] <Qantourisc> i think
[16:32:45] <tech2077> possibly
[16:33:13] <tech2077> might want to ask over on ##electronics
[16:34:14] <Qantourisc> Electronically it's ok :)
[16:34:23] <Qantourisc> communication wise not to sure :)
[16:53:17] <RikusW> I'm seeing very odd behaviour in Linux, opening ttyACM0 fails, even though I'm in dialout, the weird part is that once opened as root and closed again I can open it normally ?!
[16:53:45] <tech2077> is it being locked by the root process?
[16:53:54] * RikusW said closed again
[16:54:06] <tech2077> ah
[16:54:15] <RikusW> right after plugging in the device it won't open
[16:54:23] <RikusW> until opened once by root
[16:54:31] <RikusW> then it suddenly works like it should
[16:54:41] <tech2077> what does dmesg say about it?
[16:56:55] <twnqx> check permissions/ownership before and after the root acess?
[16:57:38] <RikusW> first time I've seen this..
[16:57:54] <RikusW> dmesg isn't usefull, don't show anything when it fails
[16:58:33] <RikusW> trying to open file "/dev/ttyACM0"
[16:58:33] <RikusW> could not open file: Device or resource busy
[16:58:48] <twnqx> lsof | grep ACM
[17:00:34] <RikusW> now it fails even after root access
[17:00:42] <RikusW> sudo still works
[17:00:47] <tech2077> ?
[17:00:54] <tech2077> su fails but sudo works?
[17:01:14] <RikusW> I mean once accessed as root it seems to work
[17:01:20] <RikusW> normally
[17:02:20] <tech2077> ah
[17:02:51] <tech2077> do try lsof | grep ACM right after plugging in
[17:03:09] <RikusW> now that won't even work... :(
[17:03:34] <tech2077> lsof is erroring?
[17:04:20] <RikusW> gives nother
[17:04:22] <RikusW> nothing
[17:04:30] <RikusW> and hangs if I don't add -n
[17:04:34] <RikusW> lsof -n
[17:04:58] <tech2077> strange
[17:05:13] <tech2077> what are the permissions before and after access
[17:05:34] <tech2077> ls -lhXt /dev/ttyACM0
[17:05:56] <RikusW> same
[17:08:58] <tech2077> what device
[17:09:45] <RikusW> USB CDC
[17:09:59] <RikusW> it used to work just fine in Debian
[17:10:03] <RikusW> Etch
[17:10:09] <RikusW> works fine in windows too
[17:10:26] <RikusW> I hope is not a fw error
[17:11:31] <RikusW> seems like there is some kind of timeout involved...
[17:11:43] <RikusW> only if you're a normal user
[17:11:47] <RikusW> not for root
[17:13:37] <RikusW> seems like one minute or so :S
[17:19:05] <damjan> RikusW: check if modemmanager is polling the device
[17:19:24] <RikusW> will do that
[17:19:40] <RikusW> lsof don't list it as open though
[17:22:30] <damjan> my experience is that modemmanager doesn't keep it open, but polls the device from time to time
[17:25:42] <RikusW> how can I stop it ?
[17:25:50] <RikusW> killing seems ineffective...
[17:27:54] <damjan> I have a udev rule like this
[17:27:57] <damjan> SUBSYSTEM=="tty", ATTRS{idVendor}=="10c4", ATTRS{idProduct}=="ea60", ENV{ID_MM_DEVICE_IGNORE}="1", ENV{ID_MM_CANDIDATE}="0"
[17:28:50] <damjan> this is for the bus pirate afaik, it makes MM ignore the tty device
[17:30:06] <tech2077> well, if you don't use a modem, you can remove modemmanager
[17:30:10] <RikusW> damjan: thanks :) kill -s SIGSTOP 10677 works
[17:30:23] <RikusW> -9 just cause it to respawn
[17:31:14] <RikusW> if I do use a modem I'll manually configure it anyways..
[17:32:14] <tech2077> yeah, it generally is started automatically, taking up a bit of boot time, good rule of thumb is to delete it on any new computers that don't require a modem for internet
[17:33:03] <RikusW> yep
[17:35:28] <RikusW> working much better now thanks :-)
[17:36:19] <damjan> NM + MM actually do work pretty good these days
[17:36:28] <RikusW> NM ?
[17:36:33] <damjan> networkmanager
[17:36:58] <RikusW> well MM is interfering with
https://sites.google.com/site/megau2s/
[17:37:04] <damjan> MM is just so that it detect what type of modem do you have, and optionaly switch it to the "modem" functionallity
[17:37:25] <damjan> RikusW: it comes with a database of blacklisted devices
[17:39:02] <tech2077> damjan, but how many people will be using NM + MM
[17:39:20] <tech2077> if you don't own or expect to own a modem, then it isn't really needed
[17:39:45] <damjan> tech2077: it also for 3G sticks, those are popular
[17:40:19] <tech2077> hmm, that's useful
[18:00:08] <RikusW> just did -> /var/cache/apt/archives $ apt-get download modemmanager
[18:00:13] <RikusW> to keep it in cache
[18:00:29] <RikusW> if I ever need it again and cannot download it
[22:14:27] <Valen> coming out of reset PORTA=255; will leave all pins on port A as inputs pulled high yes?
[22:15:03] <jadew> yes
[22:15:05] <Tom_itx> well you need DDRA set
[22:15:21] <Tom_itx> is input default? i forget
[22:15:21] <jadew> after reset, ddra should be 0
[22:15:27] <Tom_itx> k
[22:15:35] <Tom_itx> good idea to do it anyway
[22:15:39] <jadew> yeah
[22:15:48] <Valen> I thaught input was the default state
[22:16:03] <Tom_itx> it should be but who knows who goes there
[22:16:23] <Valen> 's not a bad plan
[22:17:41] <jadew> one way that could fail is if you'd use a boot loader
[22:21:12] <Valen> don't have one of those ;->
[22:25:19] <R0b0t1`> penises
[22:50:18] <Powderhound> If I'm using about 50% of available RAM for static data, is it very likely that I'll run into stack/heap collision issues? I'm getting random lock ups during _delay_ms calls
[23:42:40] <R0b0t1`> Powderhound: Possibly, try to debug the lockups if possible.
[23:42:48] <R0b0t1`> Powderhound: You might look into storing data in flash.
[23:43:08] <R0b0t1`> (I'm assuming it's static, if not, you may need a higher-end chip.)
[23:43:11] <Powderhound> Alright, that was the path I was thinking about, but figured it was worth a shot
[23:43:16] <Powderhound> It is all static
[23:43:32] <Powderhound> though the majority of the space is held by buffers (which I can try to shrink)
[23:44:38] <Powderhound> Any tips for debugging the lockups?
[23:44:40] <Casper> Powderhound: remember that each function that you call use about 6-12 bytes of ram
[23:44:50] <Powderhound> Got it
[23:44:59] <Powderhound> I currently have an AVRISPmkII
[23:45:06] <Powderhound> and it's running on an xmega
[23:45:13] <Powderhound> so I'm limited in hardware debugging
[23:46:29] <Powderhound> Is it worth it to grab something to do hardware debugging? Or are there any tricks to see the current state of the stack?
[23:47:21] <Powderhound> I ran into
http://jeelabs.org/2011/05/22/atmega-memory-use/ which (should) allow me to print out the remaining RAM, but that's all I've found
[23:49:29] <rue_shop3> this mega162 is an interesting chip
[23:50:40] <rue_shop3> I wanted to use it up, but now that I look at it, I want to save it for a project that needs multiple uarts or a lot of pin interrupts
[23:51:02] <R0b0t1`> Hah
[23:51:19] <R0b0t1`> Just buy a bunch of some chip, and use it liberally
[23:51:34] <R0b0t1`> I've found it saves headaches for moderately increased price
[23:51:37] <R0b0t1`> prototyping, anyway
[23:51:43] <R0b0t1`> Have you seen the attiny1632?
[23:51:53] <R0b0t1`> 1634*
[23:52:28] <R0b0t1`> I2C slave and USI, two UART
[23:53:17] <rue_shop3> I thought it was a pin-for pin of the m32, but its not, I have to decide
[23:53:52] <rue_shop3> almost all my projects use m32, which, after seeing the m324 I'm gonna gradually switch over
[23:54:19] <rue_shop3> atleast if I have two of a chip I can make one stupid mistake, I only have one 162
