Back
[00:20:34] <alx741> guys
[00:21:03] <evil_dan2wik> hi
[00:21:16] <alx741> i'm trying to buy some atmega328p on ebay. Are these real?
http://www.ebay.com/itm/5-PCS-ATMEGA328P-PU-Microcontrolle-r-With-ARDUINO-UNO-R3-Bootloader-/141871645094?hash=item21083581a6:g:l~0AAOSwr81UPOa3
[01:36:51] <evil_dan2wik> alx741, hard to tell
[01:37:22] <evil_dan2wik> it will probably work either way
[01:38:56] <alx741> evil_dan2wik: Where should i bought them? (with international shipping, i live in south America :( )
[01:40:45] <evil_dan2wik> I get them off ebay, but I do not recommend that
[01:47:22] <alx741> mmhh is there any difference between fake and real ones or are there like -perfect- clones? any way to recognize them (phisically)?
[01:47:36] <alx741> *are they
[01:53:30] <evil_dan2wik> I'd assume they have to work the same
[01:53:54] <evil_dan2wik> Their calibrations and specs might differ a little (or a lot).
[01:54:28] <evil_dan2wik> aside from 2, all my atmegas have worked fine
[01:55:25] <evil_dan2wik> those 2 atmegas were miss labelled for some reason and were probably fake, they were meant to be 328p-pu but I got 328-pu
[01:55:44] <evil_dan2wik> even though the package said 328p-pu
[02:00:42] <alx741> That sounds like very fake, i had bad experiences with electronics on ebay that come from China. 32GB pendrives that have 2GB, the same for SD cards, awful.
[02:03:16] <evil_dan2wik> storage drives are easy to rig up as fake though
[02:04:00] <alx741> I guess i'll give those a try and make some tests. But really would like to get some ariginal Atmel chips. Hope local stores sell the original and not the fake ones, otherwise i will end up in the same situation for extra money, bummer. Anyways, thanks for that, i'll give ebay chips a shot
[02:11:00] <Mr_Sheesh> Can always buy off digikey, newark, mouser, the big suppliers who buy directly from the factory
[04:58:28] <Jartza> I got few fake atmegas from china years ago
[04:58:51] <Jartza> I didn't get PWM or EEPROM to work at all with them, other than that they seemed like "perfect clones"
[04:59:07] <Jartza> maybe if running something that needs to be cycle-accurate they might have failed too
[06:22:59] <LeoNerd> 8MHz CPU clock. No prescaler. If I set a timer unit to to CTC mode, period = 7, no prescaler and OCnxR = 3, should I get 1MHz on the PWM pin?
[10:34:56] <alx741> Jartza: ah! good to know that, so when i get the chips i should try the EEPROM and PWM to see
[10:48:34] <Jartza> alx741: it's still not proof they are original ;)
[10:48:42] <Jartza> this was like 5-6 years ago
[10:50:49] <alx741> yeah... but in case they doesn't have eeprom i would at least know pretty easy the're fake
[10:51:23] <alx741> i heard of fake atmega8 chips like 4-5 years ago. But never did for atmega328p
[11:02:49] <Lambda_Aurigae> they likely just upped the flash and made it look like a 328
[11:09:34] <alx741> Lambda_Aurigae: i would be what they do with pendrives flash memory, so a good test would be to write and read a 32K dummie program
[11:09:46] <alx741> *it would be
[11:10:25] <Lambda_Aurigae> several writes and reads even.
[11:12:16] <Lambda_Aurigae> fill memory with all 1111111111111111 then 0000000000000000 then alternating bits and a couple of other patterns..even doing incremental data through the memory then decremental.
[11:12:19] <Lambda_Aurigae> just to be certain.
[11:12:24] <Lambda_Aurigae> like one would do for testing ram.
[11:13:08] <Lambda_Aurigae> on a 328, if you are not sure of it's authenticity, check out various functionality too...pin change interrupt and adc and pwm..
[13:50:36] <LeoNerd> So it turns out the PORTA pullup resistor on a tiny841 is enough to make a red LED glow, so you can see it. This is handy :)
[13:53:51] <Mr_Sheesh> Pilot light basically
[13:55:54] <Emil> LeoNerd: most avrs have strong enough pullups that it indeed works
[13:55:57] <Emil> iirc
[13:57:36] <Emil> Anyone used debugWire?
[13:58:09] <Emil> LeoNerd: IO pull-up resistor is on the scale 20 to 50kOhm on attiny841
[14:00:11] <Emil> LeoNerd: Same on atmega328p
[14:00:41] <Emil> So yeah, most avrs (probably) have strong enough pull-ups
[14:04:33] <LeoNerd> Mm.. so in other news, I believe I've just made a DMX receiver chip that acts like an SPI slave
[14:04:38] <LeoNerd> :) Might have to sell these
[14:17:27] <Emil> LeoNerd: nice
[14:17:54] <Emil> LeoNerd: I have been thinking about making an esp8266 based dmx lighting controller
[14:51:52] <LeoNerd> ARTnet?
[14:52:12] <Emil> Yeah, that's one option
[15:12:31] <alx741> guys, will v-usb or lufa run on an attiny26 or so? what would be the minimum requieremts for the target chip?
[15:12:45] <Lambda_Aurigae> lufa won't
[15:12:50] <LeoNerd> v-usb is horrible and I wouldn't wish it upon my worst enemy
[15:12:52] <Lambda_Aurigae> lufa requires hardware usb.
[15:13:04] <Lambda_Aurigae> v-usb is bitbanged usb and makes a fun toy but that's about it.
[15:13:14] <LeoNerd> LUFA will run anywhere with hardware USB. Which is mostly just the megannUx chips - 16U2, 16U4, 32U4, etc...
[15:13:15] <Lambda_Aurigae> would have to look at the specs of the attiny26 for iit.
[15:13:25] <LeoNerd> I'm not aware of any ATtiny chip with a hardware USB module
[15:13:29] <Lambda_Aurigae> v-usb will run on an attiny85 or attiny45.
[15:14:18] <alx741> oops dind't know LUFA is for usb specific, LeoNerd, is there a better alternative for v-usb?
[15:14:20] <Lambda_Aurigae> attiny26 is not going to have enough flash for it.
[15:14:33] <LeoNerd> No, to talk USB you /really/ want a hardware module
[15:14:42] <LeoNerd> v-usb is a good implementation of a /terrible/ idea
[15:14:48] <alx741> Lambda_Aurigae: that's what i needed to know, thanks! :D
[15:14:50] <LeoNerd> There won't be any other better things than that, because the underlying idea is terrible
[15:15:11] <Lambda_Aurigae> v-usb is a semi-commercial mod of a hack.
[15:16:21] <Lambda_Aurigae> I have found so far that it works reliably on about 40 to 50 percent of usb ports and is functional on upwards of 80% of them.
[15:16:40] <Lambda_Aurigae> but I still constantly run into usb ports that are not compatible with v-usb.
[15:16:49] <alx741> oh.. I see... is it terrible becouse of the usb hardware specifics that v-usb solves in a hacky, but non standard compient way, am i right?
[15:17:08] <LeoNerd> USB works at 12Mbit/sec
[15:17:15] <LeoNerd> That's justabout faster than any actual AVR chip
[15:17:36] <LeoNerd> v-usb manages to work because you trick the upstream into talking at slower 1.5Mbit/sec "lowspeed" mode and then bitbang as fast as you can
[15:17:39] <LeoNerd> It still sucks
[15:17:46] <Lambda_Aurigae> v-usb is low speed only...so doing cdc is a hack that's not actually officially supported by the usb spec.
[15:17:50] <LeoNerd> Real hardware can talk /much/ faster than the AVR core can bitbang it
[15:18:17] <Lambda_Aurigae> v-usb also makes some assumptions on the usb comms that are not always correct.
[15:18:36] <Lambda_Aurigae> it has to because the chip can't handle the full usb protocol because it's not fast enough.
[15:18:41] <alx741> yeah, a USB 2.0 complient host is suposed to talk to 1.5Mbit/sec devices just fine, so I think the HID is -kind of not so bad- (?)
[15:19:30] <Lambda_Aurigae> yes, supposed to...but the way it implements that low speed protocol varies from chipset to chipset.
[15:19:46] <Lambda_Aurigae> mostly I've seen usb3 not talk with v-usb.
[15:20:25] <Lambda_Aurigae> and usb2 on about half the laptops I have.
[15:20:26] <alx741> ooh i get it... So LUFA + USB AVR would be the recomendation?
[15:20:34] <LeoNerd> Definitely so
[15:20:40] <LeoNerd> I've had much success with LUFA on the 23U4
[15:20:43] <LeoNerd> 32U4
[15:20:43] <Lambda_Aurigae> yes...
[15:20:56] <Lambda_Aurigae> I do usb on pic32 myself but that's a whole other world.
[15:21:35] <Mr_Sheesh> Sure seem to be 32u4 afficionados about :P
[15:22:01] <Lambda_Aurigae> Mr_Sheesh, 32u4 is a pretty decent chip.
[15:22:12] <alx741> great, i'll get a 32U4 to try :D thanks!
[15:22:38] <Lambda_Aurigae> alx741, one thing about the 32u4 I don't like...it's not available in a dip package.
[15:22:52] <Mr_Sheesh> I'd never heard that (snicker! All you guys say the same thing!) I'll play with one, just that 328 are OK & cheapish LOL
[15:23:36] <Lambda_Aurigae> 328 doesn't have usb though.
[15:23:37] <alx741> yeah thats a big bummer, but though abuot getting a breakbord adapter from ebay
[15:25:03] <Lambda_Aurigae> three things I wish I could get from avr and if I could I would drop pic completely
[15:25:10] <alx741> I like that PIC has for example the PIC18F4550 in DIP packge with USB 2.0 full speed... but I hate Microchip tools and everything is privative
[15:25:40] <Lambda_Aurigae> usb connection in a dip package, usb without an external clock/crystal, and ability to execute code from sram.
[15:25:49] <Lambda_Aurigae> all of those I get in the pic32.
[15:25:51] <alx741> an atmega32u4 in DIP package would be *so* nice
[15:26:31] <LeoNerd> Eh.. if anyone wants a 32U4 on a DIP breakout I'll solder you one up
[15:26:36] <LeoNerd> Post it to me and I'll post it back
[15:27:41] <Mr_Sheesh> free soldering practice, neat :)
[15:36:22] <carabia> also mcp has a wider range of devices with dma, iirc.
[15:36:41] <Lambda_Aurigae> yes
[15:36:42] <carabia> whereas with avrs i think you have avr32s or xmegas
[15:36:47] <carabia> xmega, fucking haha
[15:36:55] <Lambda_Aurigae> never touched an xmega myself.
[15:37:15] <Lambda_Aurigae> avr32 I find to be a joke for all their complexity.
[15:37:21] <carabia> failed projects like that are the reason why wankers at atmel can't run a sustainable business
[15:37:29] <alx741> i made it for a atmega8, i was lazy and didn't bother to make and print a proper schematic, a thin permanent market and pulse worked out, then i was lazy to make holes on it so i just solder the pin headers like this:
http://i.imgur.com/w1BEJUD.jpg
[15:37:34] <carabia> well, couldn't, to be exact
[15:38:02] <Lambda_Aurigae> alx741, I would have used right angle headers myself.
[15:38:13] <alx741> horrible, but worked, 32u4 wouldn't be different
[15:38:26] <Lambda_Aurigae> carabia, yeah...it'll be interesting to see how microchip handles the atmel chip line.
[15:38:30] <alx741> Lambda_Aurigae: exaclty! i dind't have any that moment
[15:39:01] <Emil> LeoNerd: come on
[15:39:02] <alx741> *at the moment (?), sorry i suck writing english
[15:39:11] <Emil> the idea is awesome and we make do with what works
[15:40:58] <LeoNerd> Emil: ?
[15:41:03] <Emil> Regarding vusb
[15:41:19] <carabia> Yeah. I don't know though. I'm going to dig into cortex-m0s myself
[15:41:28] <LeoNerd> Eh; no, the idea is to bitbang USB. You cannot satisafactorily bitbang USB from a 16MHz CPU
[15:41:32] <LeoNerd> Ain't Gonna Happen
[15:41:38] <Emil> That's not the point
[15:41:42] <LeoNerd> I could justabout see a 48MHz ARM chip talking it
[15:41:50] <LeoNerd> (48/12 = 4, a nice round number)
[15:42:13] <Emil> Anycase, anyone know of a not shitty bootloader for 32u4?
[15:42:32] <Lambda_Aurigae> might as well bitbang ethernet and UDP from an AVR
[15:42:33] <Emil> One that doesn't enumerate all the fucking time
[15:42:36] <Lambda_Aurigae> oh, right, been done!
[15:42:50] <Lambda_Aurigae> by the same guy who first bitbanged usb on an avr.
[15:43:09] <Emil> Lambda_Aurigae: his name?
[15:43:13] <Lambda_Aurigae> Igor
[15:43:19] <Emil> Ah
[15:43:22] <Emil> the legendary igor
[15:43:52] <Lambda_Aurigae> Igor Cesko
[15:44:05] <Lambda_Aurigae> unfortunately his website is down.
[15:44:11] <Lambda_Aurigae> has been for weeks at least.
[15:44:17] <Emil> yeah : /
[15:44:52] <Lambda_Aurigae> http://tuxgraphics.org/common/src2/article08101/avr309.pdf
[15:45:06] <Lambda_Aurigae> atmel must have bought/licensed it from Igor!
[15:47:33] <Emil> So now that atmel has provided an appsheet for it
[15:47:40] <Emil> is it still prohibited
[15:47:47] <Lambda_Aurigae> hehe.
[15:48:01] <Lambda_Aurigae> not prohibited..just not something I would ever rely on.
[15:48:32] <Lambda_Aurigae> and that's not v-usb but igor's earlier implementation.
[15:49:17] <Emil> How are they different?
[15:50:04] <Lambda_Aurigae> they both do the same thing..I believe v-usb was inspired by igorplug...but I would still not use either one for a real product.
[15:54:36] <Emil> Yeah
[15:54:45] <Emil> anycase, anyone used the debugwire?
[15:55:46] <Lambda_Aurigae> not I.
[15:56:09] <Lambda_Aurigae> last I knew you had to have atmel studio for that.
[15:56:17] <Emil> Oh
[15:56:27] <Emil> Strange that it has not been reverse engineered
[15:56:28] <Lambda_Aurigae> might not be the case anymore.
[15:56:38] <Lambda_Aurigae> I just never saw any linux implementation.
[15:57:37] <Lambda_Aurigae> hmm.
[15:57:50] <Lambda_Aurigae> I guess avarice and avr-gdb can do it now.
[15:58:00] <Lambda_Aurigae> http://blag.pseudoberries.com/post/16374999524/avr-debugwire-on-linux
[16:00:01] <Emil> can I use usbasp in avrdaron's place?
[16:00:13] <Lambda_Aurigae> dunno
[16:00:25] <Lambda_Aurigae> usbasp is not anywhere close to the same class as the avrdragon
[16:00:56] <Lambda_Aurigae> it would require some software changes on the usbasp.
[16:01:36] <Lambda_Aurigae> from what I'm seeing, usbasp can not do debugwire.
[16:02:18] <LeoNerd> I'm planning to get around to doing some debugWire stuff sometime
[16:02:19] <LeoNerd> Any Day Now
[16:02:28] <LeoNerd> I have RikusW's notes and decoding and stuff...
[16:02:39] <LeoNerd> My HVSP controller will probably do it
[16:03:05] <Lambda_Aurigae> usbasp is an isp programmer, not a debugger.
[16:03:18] <Emil> Lambda_Aurigae: so?
[16:03:31] <Lambda_Aurigae> so, it will need to be reprogrammed to be a debugger.
[16:03:41] <Emil> Yeah, but that's not a problem
[16:03:44] <Lambda_Aurigae> debugwire is not the same interface as isp.
[16:04:07] <LeoNerd> It's much simpler (2 wires; GND and RESET). the ISP6 connector could be pressed into use for it
[16:04:10] <Lambda_Aurigae> debugwire has not been totally reverse engineered.
[16:04:14] <LeoNerd> But it needs different hardware on the other side
[16:04:24] <LeoNerd> Also that - we have a partial understanding and some notes, but details remain vague
[16:04:30] <Lambda_Aurigae> at least, not publicly.
[16:05:09] <Lambda_Aurigae> LeoNerd, it's simpler like i2c is simpler than spi, eh?
[16:05:20] <LeoNerd> It's onewire UART
[16:05:23] <Emil> i2c is more complex than spi
[16:05:33] <LeoNerd> It literally is UART. You TX with open-collector on both sides, and give it a 10k pullup
[16:05:36] <Lambda_Aurigae> Emil, but it has fewer lines!
[16:05:47] <LeoNerd> UART at.. I think it's /128 over the fCPU of the target
[16:05:59] <Emil> Lambda_Aurigae: ssh, someone might even start to think like that because you "said so"
[16:06:06] <Lambda_Aurigae> hehe.
[16:06:12] <LeoNerd> Any AVR UART could be used to talk dW, if you just put a diode on the TX pin then common that and RX with a pullup
[16:07:09] <LeoNerd> Hm.. is a 0.5% ceramic resonator "good enough" for UART frequencies?
[16:07:18] <Emil> Lambda_Aurigae: yes
[16:07:24] <Lambda_Aurigae> LeoNerd, ummmm..try and see?
[16:07:31] <Emil> Lambda_Aurigae: *
[16:07:35] <Emil> Well, fuck
[16:07:36] <Emil> LeoNerd: *
[16:07:41] <Lambda_Aurigae> hehe.
[16:07:42] <Emil> There we go
[16:08:03] <Emil> LeoNerd: AVR's internal resonator isn't that stable iirc
[16:08:03] <carabia> you can't ask for "uart frequencies" universally
[16:08:26] <Lambda_Aurigae> Emil, because it's an RC oscillator...it actually drifts with temperature and voltage.
[16:08:30] <LeoNerd> Hrm.. or maybe I'll just use a normal xtal here
[16:08:39] <Emil> Lambda_Aurigae: indeed
[16:08:48] <Emil> LeoNerd: What kind of frequencies?
[16:09:02] <LeoNerd> Oh hang on a moment.. I remember now. Thanks for asking :)
[16:09:05] <Emil> You can tune osccal for 115200
[16:09:13] <Emil> Or, just use 250000
[16:09:19] <LeoNerd> IdeallyI'm targeting 250kbaud so I *don't* want the normal baud divisor (14.7456MHz)
[16:09:27] <LeoNerd> I specifically want a neat decimal one, like 16. :)
[16:09:35] <Lambda_Aurigae> LeoNerd, or 20.
[16:09:39] <LeoNerd> tiny841
[16:09:43] <Lambda_Aurigae> oh.
[16:09:58] <LeoNerd> (because two UARTs :) )
[16:10:22] <Emil> LeoNerd: simply use a different value in
[16:10:26] <Emil> was it UBRR
[16:10:28] <LeoNerd> Hrm.. do I go murata grainofrice resonator, or a normal xtal + 2 caps. I'm starting to like those resonators more lately
[16:10:31] <LeoNerd> Yes; UBRR
[16:10:35] <LeoNerd> I'm well aware how to talk UART, thankyou :)
[16:10:53] <Emil> LeoNerd: then you know the internal clock is stable enough ; )
[16:11:02] <Lambda_Aurigae> I've used external crystal oscillator to eliminate one pin for the crystal.
[16:11:12] <Lambda_Aurigae> oscillator cans are awesome.
[16:11:20] <LeoNerd> Hrm... that's a thought
[16:11:34] <Lambda_Aurigae> that's how Jartza did the octapentaveega
[16:11:36] <LeoNerd> How are they for size though? I mean, have you -seen- how tiny the murata cst series is?
[16:11:50] <LeoNerd> I call them rice-grains for a reason :)
[16:11:50] <Lambda_Aurigae> external oscillator feeding 3 chips at once to keep them in sync.
[16:12:08] <LeoNerd> ohoho speaking of rice I have to do an urgent grocery shop run. brb
[16:12:50] <Lambda_Aurigae> http://www.kr4.us/crystal-smd-16mhz.html?gclid=CN3Wlfa4ss0CFQuPaQod5UIMew
[16:13:22] <Emil> Lambda_Aurigae: solderable
[16:13:26] <Lambda_Aurigae> 5mmx3.2mm
[16:13:38] <Lambda_Aurigae> and I've seen smaller..that was just the first one I found.
[16:13:49] <Lambda_Aurigae> oops
[16:13:53] <Lambda_Aurigae> that's a crystal.
[16:13:56] <Lambda_Aurigae> not the oscilator can.
[16:16:01] <Lambda_Aurigae> http://www.digikey.com/product-detail/en/sitime/SIT8021AI-J4-18S-16.000000E/1473-1358-2-ND/5877331
[16:16:05] <Lambda_Aurigae> there we go.
[16:16:21] <Lambda_Aurigae> 16MHz oscillator.
[16:16:35] <Lambda_Aurigae> 1.54mm x 0.84mm
[16:16:37] <learath> that's both 1. super expensive, and 2. tiny
[16:16:49] <Lambda_Aurigae> LeoNerd wnted small.
[16:16:50] <Lambda_Aurigae> hehe
[16:17:00] <learath> and small is, amusingly, expensive
[16:17:05] <Lambda_Aurigae> yeah.
[16:17:10] <Lambda_Aurigae> it's a full oscillator though.
[16:17:12] <Emil> now, that's something I'd rather not solder by hand
[16:17:19] <Lambda_Aurigae> however, it's only 1.8v
[16:17:26] <learath> bah, it's bigger than a grain of sand :P
[16:18:35] <Emil> Hmm, what would happen if you drove that with 5V :D
[16:18:42] <Emil> Or even just 3.3
[16:18:45] <Lambda_Aurigae> http://www.digikey.com/product-detail/en/abracon-llc/ASEM1-16.000MHZ-LC-T/535-9760-2-ND/1873206
[16:18:55] <Lambda_Aurigae> it would probably fry.
[16:19:03] <Emil> :ooo
[16:19:05] <Lambda_Aurigae> that one is 3.3V and a bit easier to handle.
[16:19:09] <Emil> Check those current differences
[16:19:12] <Lambda_Aurigae> 3.2x2.5 mm
[16:19:17] <learath> the smallest size only does 3.3v max
[16:19:50] <learath> oh wow ,that one is *giant* :P
[16:19:51] <Emil> 130uA vs 10mA
[16:19:55] <learath> 3mm on a side? :P
[16:20:20] <Emil> The 3.3v one also has half the error
[16:20:27] <Lambda_Aurigae> yup.
[16:20:34] <Lambda_Aurigae> neither are crystal based though it seems.
[16:20:40] <Emil> But dat power consumption, though
[16:20:43] <learath> Really? Interesting
[16:20:48] <learath> What do they use instead?
[16:21:04] <Emil> learath: inverters probably
[16:21:07] <Emil> it's a cmos design
[16:21:10] <Lambda_Aurigae> MEMS oscillator...silicon inverter.
[16:21:39] <Lambda_Aurigae> http://www.digikey.com/product-detail/en/fox-electronics/FOX924B-16.000/631-1071-2-ND/1024663
[16:21:41] <Emil> Fun story, you can take a simple mems inverter chip and make fm \o/
[16:21:49] <Lambda_Aurigae> larger, 6mA, crystal.
[16:22:07] <Lambda_Aurigae> 5mm x 3.2mm
[16:22:13] <Emil> Even moooore accurate
[16:22:19] <Lambda_Aurigae> yup.
[16:22:20] <Emil> First one was +-100ppm
[16:22:24] <Lambda_Aurigae> crystal is far more accurate.
[16:22:25] <Emil> Seconds one 50ppm
[16:22:29] <Emil> this one is 2.5ppm
[16:22:39] <Lambda_Aurigae> and it's 10 times the size of the first one.
[16:22:45] <Lambda_Aurigae> well,,,5 times.
[16:22:47] <carabia> The sitime osc isn't all _that_ expensive. price break for 1 is $2.29. learath ?
[16:23:24] <Emil> It would drift ~1kHz at 433Mhz
[16:24:02] <carabia> And you're running at 433 MHz, now how often was it again?
[16:24:17] <Emil> carabia: PLL, my friend
[16:24:27] <carabia> For what?
[16:24:39] <Emil> carabia: radio, of course!
[16:24:40] <Lambda_Aurigae> to make 433mhz from 16mhz maybe
[16:25:14] <carabia> oh, too long didn't read. well get an xtal for the pll and stop crying
[16:25:35] <Emil> carabia: eh :D
[16:25:58] <carabia> still didn't read.
[16:27:00] <LeoNerd> Oh; yah.. 5V operation would be good
[16:27:15] <LeoNerd> It's OK.. I think I'll go for one of these resonators to be honest. I don't /need/ the extra IO pin
[17:03:21] <alx741> has anyone tried rust (language) on avr yet?
[17:21:07] <Casper> alx741: I don't think that any other language than assembler and C is good on avr. Even C++ has been found to be pushing your luck
[17:21:35] <LeoNerd> I expect that at some point Rust might actually become viable
[17:22:09] <LeoNerd> It really doesn't need anything beneath it to support it - not like the GC of Java or Go or any of that crowd
[17:59:40] <Tom_itx> wonder why ppl feel the need to keep changing languages
[18:00:27] <Emil> Tom_itx: Because innovation and future take change
[18:00:54] <LeoNerd> because other langauges let me do things C doesn't
[18:00:57] <Tom_itx> re'inventing the wheel?
[18:01:20] <Tom_itx> does it make the avr more capable?
[18:01:22] <LeoNerd> In C I can't have a link-time array of, say, 5 elements, where each element is actually notated in five different .c source code files
[18:01:23] <alx741> i would love to see Rust as a real feaseble atlernative for C
[18:01:31] <LeoNerd> I'd love if I *could* do that
[18:02:45] <Tom_itx> you can't make an external reference to it?
[18:03:10] <LeoNerd> "it"?
[18:03:19] <Tom_itx> the array elements
[18:03:21] <LeoNerd> I want an array of "Task" elements
[18:03:30] <LeoNerd> The static initialiser for each element should be in the .c file that implements /that/ task
[18:03:46] <LeoNerd> I want to initalise the tasks[0] element in task0.c, the tasks[1] element in task1.c, ...
[18:04:09] <LeoNerd> Instead, what I have to do is initalise them all in one .c file and refer out to nonstatic task functions that are linked across them
[18:04:11] <LeoNerd> it sucks
[18:04:17] <LeoNerd> It means hunting around in source code for things
[18:04:33] <LeoNerd> Instead, I'd like each .c file to be self-contained and actually.. er.. linked, at link time
[18:04:36] <LeoNerd> like a linker is supposed
[18:05:04] <Tom_itx> that's not a linker issue?
[18:05:24] <Tom_itx> or a limitation of the language itself?
[18:05:35] <LeoNerd> Sadly not. All the linker can do is patch up addresses of named symbols
[18:05:41] <Tom_itx> just asking...
[18:05:45] <LeoNerd> It's a limitation both of the ELF linker, and of the C language
[18:08:14] <Tom_itx> i've been trying to wrap my head around oops a bit lately
[18:56:22] <inflex> lo folks
[20:52:20] <rue_shop3> anyone have favorite serial bigbang tx code for avr?
[20:52:31] <rue_shop3> for async serial
[20:52:39] <rue_shop3> at some baud rate
[21:54:30] <inflex> yeah, I've got one that works
[21:54:34] <inflex> lemme go find
[21:54:45] <inflex> so I'm guessing you just want to pump out debug/data from the AVR blindly?
[21:58:22] <inflex> rue_shop3,
http://ctpc.biz/blind-tx.c
[21:59:09] <inflex> rue_shop3, I use that on my mega48 a lot, hope it gives you a starting point
[23:32:10] <rue_shop3> tiny13 and tiny26's reporting data
[23:33:28] <rue_shop3> inflex, ooo asm...
[23:33:53] <inflex> aaah... t13 stuff... hrmm... haven't tried it on those
[23:35:09] <rue_shop3> what speed did you use this on?
