Back
[00:14:31] <serp_> why did they reject it?
[00:16:37] <CapnKernel> serp_: tl;dr version:
http://www.chinese-forums.com/index.php?/topic/36854-check-dates-carefully-or-you-may-lose-your-visa/
[00:20:13] <serp_> man that sucks
[00:23:40] <CapnKernel> yes
[00:34:02] <Casper> but could be worse...
[00:34:14] <Casper> I guess they could have jailed you for illegal stay
[00:40:20] <CapnKernel1> Yes absolutely. Could have been much worse! I could have had my very own "tugboat".
[00:40:57] <Casper> but it does suck
[00:42:24] <CapnKernel1> since this is a family channel, I will restrain myself to expressing how I feel through the use of nested for loops containing the f word.
[00:51:36] <Casper> what about: while true; do echo $fword > /dev/null; done :D
[00:54:47] <CapnKernel1> Must you consign my pain to the bit bucket so callously?
[00:54:56] <CapnKernel1> Shorter: yes $fword > /dev/null
[00:55:16] <CapnKernel1> See I bet you didn't know that.
[00:56:24] <Casper> yes I do, but didn't tought fo that
[00:56:52] <Casper> it's not something I used, I know it exist, but never used it
[00:57:02] <Casper> but
[00:57:11] <Casper> I started using e4defrag :D
[00:57:18] <Casper> it's now mainstream!
[00:57:43] <CapnKernel1> LOL, just in time for people to stop caring so much (as btrfs will save us all!)
[00:59:43] <Casper> torrents kill all filesystems :(
[01:00:02] <CapnKernel1> Oh?
[01:01:37] <Casper> yeah...
[01:02:42] <Casper> e4defrag is surprisingly fast
[01:02:44] <Casper> but
[01:02:49] <Casper> it do not fully defrag
[01:02:55] <Casper> as it do not move away the other files
[01:03:09] <Casper> but still better than nothing
[01:03:11] <CapnKernel1> return true; is also surprisingly fast :-)
[01:04:40] <CapnKernel1> return 22; is correct for certain cases
[01:04:59] <CapnKernel1> (and if you have no idea what I mean by that, see the discussion for point 5. priceless:
http://www.joelonsoftware.com/articles/fog0000000043.html )
[01:05:14] <Landon> Casper: I don't think I've ever had to defrag my ext* filesystems o.O do you not preallocate torrents?
[01:05:28] <Casper> Landon: I do
[01:05:33] <Casper> using sparse
[01:05:40] <Casper> whish should do the same
[01:05:47] <CapnKernel1> sparse doesn't preallocate
[01:05:57] <Casper> hmmm
[01:06:01] <CapnKernel1> with sparse, you'll end up with fragmented data
[01:06:07] <CapnKernel1> But honestly, does it matter?
[01:06:10] <CapnKernel1> For a torrent?
[01:06:37] <CapnKernel1> For peering, it won't make any difference, since access is essentially random, even if a file is completely contiguous
[01:06:56] <CapnKernel1> The only time it's going to matter is when you take the whole file and use it or copy it to somewhere else
[01:07:26] <CapnKernel1> And I could be wrong about this, but there is code in the Linux kernel which detects that you're making large sequential reads, and kicks in read-ahead algorithms.
[01:07:45] <Casper> . . . actually vuze removed the sparse option... no wonder...
[01:08:55] <CapnKernel1> http://kerneltrap.org/node/6642
[01:09:59] <Kevin`> I do my torrents on nfs. such things get interesting
[01:10:15] <Casper> I do big torrents...
[01:10:15] <CapnKernel1> If a file will start out sparse but ultimately end up completely filled out, the only advantage to sparse is that creating the sparse container is near-instant
[01:10:22] <Casper> I had one of 182G loaded
[01:10:49] <CapnKernel1> Casper and his astronomical data :-)
[01:13:20] <Landon> heh lame, I can't find the car torrent anymore
[01:13:33] <Landon> was a couple hundred TB or something ridiculous
[01:14:08] <Landon> as disturbing as it is, it looks like someone backed up geocities though
[01:16:40] <Casper> I also saw some idiotic torrents, like the whole collection of wii games up to that time in one single torrent, all non-scrubbed
[01:17:23] <Casper> it was also a few TB
[01:17:47] <Casper> when a scrubbed collection would have been like 500-600G
[02:19:35] <cyanide> hello folks
[02:20:09] <OndraSter> mornin
[02:20:25] <OndraSter> my bus goes in minutes lol
[02:20:26] <OndraSter> 11
[03:34:35] <mitsakos> Hello, i'm sending characters on Atmega162 via uart using hyperterminal, and i display the character values on 8 (bit) leds (PORTA). When i send number character especially characters with ascii code < 0x40 i receive the ascii code correctly, when i try to send a character like 'A' (0x41) i receive two bytes one is '?' (0x3F) and the other is 0x01
[03:35:21] <Landon> any chance we can see your character sending code?
[03:35:30] <mitsakos> for ascii codes larger than 0x3F looks like it losts the first nibble
[03:36:05] <mitsakos> i have checked the ascii code sent from hyperterminal using a logic analyzer and it is correct. The problem is on the avr
[03:36:31] <Landon> ah ok, so it's going computer->uc
[03:36:53] <mitsakos> yes
[03:39:04] <ziph> mitsakos: How about sending from the AVR to a PC?
[03:39:53] <mitsakos> For ox40 i receive 0x3F & 0x00 for 0x41 i receive 0x3F & 0x01 for 0x42 i receive 0x3F & 0x02
[03:40:28] <ziph> So the corruption is both ways?
[03:41:08] <mitsakos> no this was just some more info, i haven't tried yet to send something i'm trying to configure the receive mode which is what i'm interesting for
[03:41:36] <ziph> You could try sending from the AVR to double check the bit timing.
[03:42:23] <ziph> (Since the baud rate generator on the AVR is tied to both RX and TX)
[03:43:12] <mitsakos> ok let me try it
[03:46:23] <mitsakos> nice i get messed characters
[03:51:00] <ziph> mitsakos: Sounds like a timing problem then.
[03:51:46] <ziph> mitsakos: Or a cable/driver problem, but you can check that with a CRO fairly quickly.
[03:52:33] <mitsakos> yes me too..
[03:52:41] <mitsakos> CRO ?
[03:52:50] <ziph> oscilloscope.
[03:53:07] <mitsakos> i put it on the logic analyzer
[03:53:32] <mitsakos> i receive A and then a gramming error A and a framming error and agani
[03:54:06] <ziph> Yeah, it should be fine if it looks good on that, but there's always a chance the receiver on your LA is seeing differently to the RS232 receiver.
[03:54:54] <ziph> Have you turned the LA resolution right up and checked the duration of a single bit?
[03:56:00] <mitsakos> yes
[03:56:03] <mitsakos> it's 104uS
[03:56:58] <ziph> That should be fine.
[03:57:06] <ziph> Is that for AVR->PC?
[03:58:56] <mitsakos> yes
[03:59:21] <mitsakos> and it is the same for both packets with frame error and not
[04:01:01] <mitsakos> i also checked the cpu clock by enabling CKOUT fuse and it is correct
[04:03:51] <ziph> What's your stop bit/parity set up?
[04:06:36] <mitsakos> UCSR0C |= (1<<URSEL0)| (1<<UCSZ00) | (1<<UCSZ10); //Data 8-bit / 1-stop bit / Parity Disabled / Asynchronous mode
[04:07:13] <mitsakos> o!
[04:07:32] <mitsakos> (1<<UCSZ10) i think this is wrong it should be (1<<UCSZ01)
[04:08:00] <mitsakos> f..king duble uart
[04:08:03] <mitsakos> *double
[04:08:06] <ziph> :)
[04:08:12] <mitsakos> let me check
[04:09:32] <mitsakos> hah
[04:09:40] <mitsakos> yes that's it
[04:09:41] <mitsakos> pf..
[04:11:12] <ziph> My next suggestion was to send 0x55 which should give you a nice pattern of 5 positive bits (on the RS232 line, or 5 negative bits on the TTL/CMOS side)
[04:12:41] <mitsakos> yes.. but i could saw that the problem was with larger bits and i have focused at the data transmittion but i had checked the registers lots of times but never saw that bit :P
[04:16:35] <ziph> Yeah, that kind of thing is annoying.
[04:17:31] <ziph> My most recent one like that was when TI changed the default value of a single bit of a register between two supposedly identical CODEC's. :)
[04:19:04] <mitsakos> yes.. and atmel hasn't double codes on every avr. i used before atmega8 and it was with single bit
[04:19:28] <ziph> Oh, it was the speed doubling bit?
[04:20:22] <ziph> (I might be thinking of SPI on some IC.)
[04:23:04] <mitsakos> the matter is that after 3 hours searching what is wrong i found it.. :P enough hour for lost
[04:23:17] <mitsakos> for a single bit
[04:25:51] <Tom_itx> could have been days
[04:26:10] <mitsakos> that's true :P
[04:27:37] <cyanide> lol digikey fail
[04:27:56] <cyanide> i found a sensor. digikey sells it for 24. arrow sells it for 9.90
[04:28:36] <ziph> Tried Findchips?
[04:28:55] <cyanide> talking to me??
[04:29:46] <ziph> cyanide: Yeap.
[04:30:06] <cyanide> hadn't heard of it until now. let me check
[04:30:21] <cyanide> whoa what
[04:30:26] <cyanide> why did i not know about this
[04:31:06] <cyanide> i love you man
[04:31:21] <ziph> Heh.
[04:31:31] <candrian> if you want check my blog i have posted some of my projects and i explain how they work. i have used avr. www.candrian.gr
[04:31:48] <cyanide> spent 2 usd more than another distributor, but no problem. just 8 bucks more
[04:31:53] <candrian> i want your opinions
[04:31:56] <cyanide> bought 4 of them sensors
[04:32:08] <ziph> cyanide: What kind of sensor?
[04:32:41] <cyanide> hall effect sensor for gearteeth
[04:32:48] <cyanide> http://www.hamlin.com//product-detail.cfm?productid=80
[04:33:14] <cyanide> candrian, nice website.
[04:33:35] <candrian> thnx
[04:33:45] <cyanide> and nice workbench :)
[04:33:53] <candrian> :)
[04:34:57] <ziph> cyanide: Looked at Austria Micro Systems for that kind of thing? They might have something cheaper.
[04:36:10] <cyanide> cant find any hall effect sensors
[04:38:50] <ziph> cyanide:
http://www.austriamicrosystems.com/Products/Magnetic-Encoders
[04:40:48] <cyanide> hmm
[05:36:04] <candrian> do you have any idea how to send a specific ascii via uart using screen on mac os
[05:36:14] <candrian> e specific hex number
[05:37:33] <ziph> Sounds like you need a little python app to do it. :)
[05:37:52] <candrian> you can't do it by just typing something?
[05:38:02] <candrian> i want to send some escape sequencies
[05:39:33] <ziph> It can be nice and simple like:
[05:39:33] <ziph> ser = serial.Serial('/dev/ttyS1', 19200, timeout=1)
[05:39:34] <ziph> ser.write("\x55")
[05:39:52] <candrian> i have no idea of python that's the matter :P
[05:40:25] <candrian> i guess 19200 is the baud rate?
[05:40:31] <ziph> Yeap.
[05:40:39] <candrian> and i put this code on .py file?
[05:40:41] <ziph> That's using pySerial.
[05:40:51] <candrian> it looks easy
[05:40:55] <ziph> Yeah, it would go into a .py file and you just need to install pySerial.
[05:41:06] <candrian> i think i have to experience with it
[05:41:14] <ziph> If you've got X-Code installed it should just be a matter of doing "python setup.py install".
[05:41:16] <candrian> how? :P noob
[05:41:24] <ziph> http://pyserial.sourceforge.net/
[05:41:25] <candrian> ok
[05:42:03] <candrian> when i run python setup.py install it says it can't open setup.py
[05:42:13] <ziph> http://pyserial.sourceforge.net/pyserial.html#installation
[05:42:25] <ziph> You do that in the directory that you've extracted pySerial into.
[05:51:00] <candrian> and i run it? :P noooob
[05:51:10] <candrian> ./file.py ?
[05:51:22] <candrian> or it needs compile?
[05:54:19] <candrian> ok.. python file.py :P
[09:27:42] <jaeckel> 'lo
[09:29:26] <jaeckel> A question about the AVR 8-bit uC architecture... how is unaligned memory access handled if I want to read a uint16/32/64 from memory?
[09:30:05] <jaeckel> I know that there can be problems on severeal uCs but can this occur here too?
[09:30:16] <ziph> jaeckel: How do you do an unaligned read on an 8-bit uC?
[09:30:33] <jaeckel> sorry, I didn't work before on this, but ran several times in this sort of problem on 32bitters..
[09:30:57] <ziph> All the busses an peripherals in the AVR are 8 bit.
[09:31:03] <ziph> and peripherals rather.
[09:31:34] <ziph> So every possible address is aligned.
[09:31:37] <jaeckel> ok
[09:31:39] <jaeckel> nice :)
[09:31:41] <jaeckel> that sounds good
[09:32:27] <ziph> Unaligned reads on 32 bit processors are fairly easy to avoid if you've got the right habits.
[09:34:47] <jaeckel> and these habits are?
[09:35:40] <jaeckel> NB: I work mostly with networking and depending on the protocol, I can't assure that memory is aligned...
[09:36:45] <ziph> With network header processing I usually just copy the buffer into a struct on the stack.
[09:36:57] <asteve> what's an unaligned read?
[09:37:22] <ziph> asteve: When you ask for (for example) a 32 bit integer from address 0x1
[09:38:05] <ziph> asteve: Most memory and peripherals on a 32 bit processor deal with 32 bit words and just can't return bytes 0x1, 0x2, 0x3 and 0x4 in a single read request.
[09:39:29] <jaeckel> the single 32 bit uP I used that handles unaligned reads is the NEC v850
[09:39:45] <ziph> asteve: Processors get around it either by doing multiple read cycles on the bus and shuffling everything back into the correct 32 bit word or by just throwing a processor exception when it happens.
[09:40:05] <ziph> asteve: Intel x86 just does multiple reads and shifts everything.
[09:40:25] <ziph> ARM throws a fault but also has an instruction specifically for unaligned reads.
[09:40:47] <asteve> return memory locations 0x1 or bytes 0x1?
[09:41:06] <ziph> They'll return a 32 bit word from address 0x1-0x04
[09:41:38] <ziph> So if your memory contains "H e l l o W o r l d" you can on an intel do a single fetch for 0x1 and get 'ello'.
[09:42:07] <ziph> Except the CPU is quietly doing a read at 0x0 and 0x4 and shifting the bytes around to give you your 0x01.
[09:42:53] <ziph> jaeckel: You can tell GCC that your structures may sit at unaligned addresses too by the way.
[09:43:02] <wollw> was there an older attin10 before the 6-pin current one?
[09:43:33] <asteve> ziph: ok, similar situation if you ask for a read at 0x5?
[09:43:47] <ziph> asteve: Yeap, any address that isn't divisible by 4.
[09:43:48] <jaeckel> ziph: really? I didn't know that
[09:43:51] <asteve> ziph: got it
[09:43:53] <asteve> thanks
[09:44:16] <wollw> Because I have a personal fork of AVRA I'm using with the six-pin attiny chips and I'd like to contribute it back
[09:46:02] <jaeckel> well.. afk
[09:46:03] <jaeckel> and thanks
[09:46:06] <jaeckel> :)
[09:46:11] <ziph> jaeckel: __attribute__((packed)) it is.
[09:46:24] <ziph> jaeckel: It forces GCC to read the structure members byte by byte.
[09:46:55] <jaeckel> ah OK, didn't know, that packed does this
[09:47:06] * wollw answers his own question: http://www.e-lab.de/downloads/DOCs/Tiny12.pdf
[09:47:19] <jaeckel> gone...
[10:07:27] <OndraSter> mornin
[10:16:28] <ziph> Yeah, I love Friday mornings.
[11:15:55] <dirty_d> hello thar
[11:18:24] <SilicaGel> avrdude: Device signature = 0x000000
[11:18:51] <SilicaGel> I'm stumped. It has power, 10K on the reset to 5V, and I see activity on RST, SCK, MISO, and MOSI when I try to erase the chip
[11:19:21] <SilicaGel> all the power and ground pins look fine
[11:20:01] <mrfrenzy> does it work if you try a new chip from the package?
[11:20:19] <SilicaGel> I did replace the IC and the result was the same
[11:20:37] <SilicaGel> This is with avrdude and an avr dragon
[11:21:19] <mrfrenzy> sorry then I have no idea
[11:21:55] <SilicaGel> I do have the clock pins open ... but I think that shoudl be fine as it should use the internal oscillator by default as the fuses are shipped by atmel
[11:23:10] <OndraSter> haven't you switched some pins?
[11:24:17] <SilicaGel> switched with what?
[11:24:52] <OndraSter> like SCK vs MOSI
[11:24:53] <OndraSter> or MISO
[11:25:01] <SilicaGel> oh oh i see
[11:25:15] <SilicaGel> hmm I don't *think* so ... I buzzed it out with a meter and it seemed to match. I'll double check.
[11:26:38] <dirty_d> hey are you guys familiar with IIR filters?
[11:27:16] <dirty_d> if throttle_filter = 0.95*throttle_filter + 0.05*throttle; is a single pole IIR filter, whats a 2 pole one look like?
[11:35:42] <SilicaGel> nope, pin to pin it buzzes out exactly the sme as the working board from sparkfun.
[11:35:53] <SilicaGel> I wonder if it's really a problem for it to try to program it off the internal clock
[11:47:24] <specing> CapnKernel: yo :D
[11:48:17] <CapnKernel> specing: hi there!
[11:48:56] <specing> hey
[11:52:29] <CapnKernel> So here I am in HK. My visa application was rejected today, so unless a miracle occurs, I'll be going home to Australia tomorrow.
[11:52:45] <OndraSter> ey
[11:52:49] <OndraSter> doh
[11:52:51] <specing> No more Shenzen? D:
[11:52:52] <OndraSter> that's bad
[11:52:53] <OndraSter> very bad
[11:53:04] <CapnKernel> specing: Not for a bit
[11:53:14] <OndraSter> how permanent is that?
[11:53:15] <CapnKernel> I can still do the PCB business, I have good people in place to do that.
[11:53:32] <CapnKernel> I will focus on the website side of things, to make it as easy as possible for folks to get good, affordable PCBs.
[11:54:54] <CapnKernel> Probably not a good idea to apply for another three months
[11:55:02] <CapnKernel> And I will likely change my passport
[11:56:51] <SilicaGel> you know what? it's som ekind of speed problem.
[11:57:13] <SilicaGel> I did -B 1000000 .. and it's slower than molasses, but it says:
[11:57:18] <SilicaGel> avrdude: Device signature = 0x1e950f
[11:57:18] <SilicaGel> avrdude: erasing chip
[11:57:24] <dirty_d> CapnKernel, bummer
[11:58:05] <SilicaGel> well a million is the wrong number
[11:58:07] <SilicaGel> but -B 10
[11:58:08] <SilicaGel> and it works
[11:58:17] <SilicaGel> assuming 0x1f950f is correct
[11:58:29] <OndraSter> CapnKernel, change to what?
[12:02:51] <specing> KernelCapn, obviously; the chinese are bound not to notice
[12:03:29] <OndraSter> why the cheapest and fastest CPLD from atmel from the series is NOT on stock on mouser, while the other ones are? :(
[12:04:58] <CapnKernel> Claim I lost my old passport, and get a new one with a different number.
[12:05:13] <OndraSter> http://eu.mouser.com/ProductDetail/Atmel/ATF1502BE-7AU44/?qs=sGAEpiMZZMuJNuO2s1hGZJOLqYzZpioSKN2y2OxcByk%3d
[12:05:18] <OndraSter> why these ones are not on stock? :(
[12:05:32] <OndraSter> but these are
[12:05:32] <OndraSter> http://eu.mouser.com/ProductDetail/Atmel/ATF1502ASV-15JU44/?qs=sGAEpiMZZMuJNuO2s1hGZJOLqYzZpioSv8F6tY2YHgM%3d
[12:05:34] <OndraSter> ? :(
[12:10:47] <Steffanx> I read on that china forum that's also not without risks CapnKernel ..
[12:11:06] <Steffanx> As the form you have to fill in has questions about previous visa's etc.
[12:11:29] <CapnKernel> It seems I don't have a very good memory.
[12:11:36] <specing> lol
[12:12:10] <CapnKernel> You will never get "points for honesty" with these guys, so if you think something might help (and that trick has helped many others), do it.
[12:12:46] <OndraSter> eh
[12:34:43] <_zap_> hi. does anyone know a board with two usb device ports?
[12:50:25] <ziph> _zap_: So that you can plug in two USB devices?
[12:52:30] <ziph> CapnKernel: Thinking of doing an aggregation service? Or just panel per customer still?
[12:52:49] <CapnKernel> Er...
[12:52:54] <CapnKernel> Customer just see their own boards.
[12:53:02] <CapnKernel> Behind the scenes though...
[12:53:34] <ziph> Yeap, so you'd for instance take an order for 2-3 a 2x2inch board's at a reasonable price?
[12:53:46] <CapnKernel> I believe the customer shouldn't need to worry about batching or any of the mechanics of scale
[12:54:05] <CapnKernel> min qty 5, but the prices are very attractive
[12:54:24] <ziph> Would you be comparable to something like BatchPCB I mean.
[12:55:28] <CapnKernel> I like BatchPCB's user interface quite a lot
[12:56:03] <CapnKernel> But the idea of waiting 4 day for the batch to be full, then BatchPCB getting boards back TWO WEEKS later, THEN they ship them to you.. well, fortunately it's easy to do a lot better and I'm doing that right now
[12:56:18] <ziph> Yeap.
[12:56:18] <CapnKernel> I do quotes without the need for gerbers
[12:56:24] <CapnKernel> If you like the price, send me the gerbers.
[12:56:39] <CapnKernel> They go to the factory immediately, for a manufacturability audit
[12:56:54] <CapnKernel> They usually tell me within 12 hours, 24-hours a day
[12:57:05] <CapnKernel> Then I tell you, you pay, and I green light it to the factory
[12:57:19] <ziph> So they do the panelising?
[12:57:25] <CapnKernel> 3-4 days the boards are back from them, and I ship them out the next day
[12:57:33] <CapnKernel> They tell me whether they are manufacturable.
[12:57:53] <CapnKernel> You might find this video interesting:
http://iteadstudio.com/other/how-itead-make-your-pcb-prototype/
[12:58:22] <CapnKernel> And no, I don't use the same factory as Seeed, and I don't use the same factory as Itead. I think their boards are inferior.
[12:59:01] <CapnKernel> How can this be 100% e-test?
http://dangerousprototypes.com/forum/viewtopic.php?f=2&t=3468
[12:59:41] <CapnKernel> The factory I use does features down to 4mil, which is needed for the multi-layer mobile phone boards they make
[13:05:29] <jadew> youtube is down:
http://www.youtube.com/
[13:05:59] <CapnKernel> Seems you're right
[13:06:10] <RikusW> why ?
[13:06:15] <jadew> it's back up
[13:06:22] <CapnKernel> My God, people will just have to do work or something!
[13:06:32] <CapnKernel> <html><body><b>Http/1.1 Service Unavailable</b></body> </html>
[13:07:47] <Casper> 6 video!
[13:08:30] <Casper> . . . noooooo
[13:08:33] <Casper> 502. That’s an error.
[13:08:46] <Casper> they are probably pushing an update
[13:08:51] <Casper> and it crashed
[13:08:52] <jadew> I think I broke it
[13:08:57] <jadew> had too many tabs opened
[13:10:26] <Casper> how many?
[13:10:52] <jadew> 5!
[13:11:30] <Casper> just that?
[13:11:54] <Casper> add a zero and that's not even what I have open
[13:12:20] <jadew> I was kidding of course :P
[13:12:49] <jadew> I usually end up with 100+ tabs as well, but not youtube related..
[13:13:51] <jadew> I think that's a bad habbit
[13:14:04] <jadew> making you think you need more ram
[13:14:11] <asteve> i have close to 50 tabs open right now
[13:14:17] <RikusW> I rarely get past 20 tabs :-P
[13:14:28] <asteve> my chrome also consumes 3gb of ramories
[13:15:09] <jadew> did they fix that thing with chrome shrinking the tabs to the point where you can't read the title of the page?
[13:16:09] <CapnKernel> This Firefox window: 91 tabs
[13:16:13] <OndraSter> CapnKernel, 4mil traces and 4mil spacing?
[13:16:14] <OndraSter> cool :)
[13:16:16] <CapnKernel> Yes
[13:16:21] <Landon> fix? don't most browsers shrink them to the point of favicon?
[13:16:25] <CapnKernel> These guys are pros
[13:16:29] <jadew> Landon, nope
[13:16:42] <OndraSter> how much for 10x10cm boards and 5x5cm boards (and even smaller, like 2.5cm x 5cm)?
[13:16:55] <jadew> FF will allow the text and favicon to be displayed, while giving you the option to scroll trough them, I think same happens for IE
[13:17:08] <Landon> opera shrinks to favicon
[13:17:17] <Landon> but with widescreen monitors I have no idea why people even have tab bars on the top anymore :)
[13:17:25] <CapnKernel> Depends on the quantity, and other factors such as board finish, thickness, colour etc. I can't give you a one-size-fits-all price
[13:17:35] <OndraSter> oh
[13:17:57] <CapnKernel> And you can have funny board shapes, like slots and rounded corners
[13:18:15] <OndraSter> quantity - 5 or 10 pieces usually, board finish - HASL is fine, no need for goldening, thickness - regular 1.6 or 1.2mm, colour - regular green is enough. You should make some chart(s) :)
[13:18:26] <CapnKernel> Anyway, unfortunately I do have to go to bed now, 2am
[13:18:37] <OndraSter> bb
[13:18:43] <CapnKernel> OndraSter: The website with instant quoting will be here soon
[13:19:19] <OndraSter> ok :)
[13:22:24] <_zap_> ziph: exactly, in order to plug in two USB devices
[13:22:37] <ziph> _zap_: Not with AVR8's.
[13:23:30] <ziph> _zap_: If you can find a chip that supports Host mode with hubs you could put a hub IC on the board.
[13:24:26] <_zap_> ziph: i would like to implement two USB devices that communicate with each other
[13:24:40] <ziph> _zap_: Why would that require two ports?
[13:24:51] <_zap_> ziph: two host PCs would connect to the ports
[13:25:23] <_zap_> ziph: so for the host PC it looks like a separate USB device
[13:25:30] <_zap_> ziph: i mean for each host PC
[13:26:08] <_zap_> ziph: [USB host]---[USB device µC USB device]---[USB host]
[13:26:42] <_zap_> ziph: [PC USB host]---[USB device µC USB device]---[USB host PC]
[13:28:27] <ziph> _zap_: Any reason for not using a $0.50 CAT5 cable? :)
[13:29:08] <_zap_> ziph: just toying :)
[13:29:38] <specing> _zap_: just use a null modem rs232 cable...
[13:29:43] <_zap_> specing: too slo
[13:29:48] <_zap_> specing: w
[13:30:07] <ziph> _zap_: Something like
http://www.prolific.com.tw/eng/Products.asp?ID=18 will do it.
[13:30:09] <specing> even at 10Mbaud?
[13:30:10] <RikusW> unless you use hi-speed devices USB will also be slow
[13:30:35] <ziph> Or
http://www.prolific.com.tw/eng/Products.asp?ID=72 for 2.0
[13:31:35] <_zap_> is it possible to get 10Mbaud with rs232?
[13:31:45] <_zap_> ziph: thx a lot. i'll have a look
[13:31:59] <specing> _zap_: with the right controller, yes
[13:32:20] <specing> everything is possible!
[13:32:23] <_zap_> specing: could you kindly give me a pointer
[13:32:32] <ziph> I don't know of any RS232 driver with a slew rate that high.
[13:32:52] <ziph> And the crosstalk through the average RS232 cable would be horrid. :)
[13:33:25] <specing> _zap_: I think my ARM could go up to 48 Mbaud, other than that, I do not know :/
[13:33:52] <ziph> _zap_: If you're looking for something to play with find something interesting and within your capabilities.
[13:34:02] <specing> the distance would have to be really close
[13:34:32] <specing> I was transferring Linux with zmodem to it many times
[13:36:54] <_zap_> ziph: it does not seem to be open in the sense that i can hack it
[13:37:05] <_zap_> ziph: i mean the two products that you mentioned
[13:37:08] <ziph> _zap_: Yeap.
[13:37:26] <_zap_> ziph: i know that there are ready solutions from usbgear.com etc
[13:37:26] <ziph> _zap_: An FPGA and two USB phy's would do it.
[13:37:38] <_zap_> ziph: i am sure :)
[13:37:40] <ziph> _zap_: They probably contain that IC. ;)
[13:37:47] <_zap_> ziph: i am just looking for a board that already supports that
[13:38:06] <_zap_> ziph: an alternative would be two boards each with a single usb device port and ethernet
[13:38:23] <_zap_> ziph: but i'd prefer one board with 2 usb device ports
[13:38:52] <ziph> There might be some expensive STM32 board that brings out both the FS and HS USB.
[13:39:11] <ziph> The cheap ones don't without additional hardware.
[13:39:22] <asteve> never give up hope
[13:39:23] <beepbeep> never surrender!
[13:39:37] <OndraSter> never gonna give up!
[13:39:48] <OndraSter> specing, I prefer booting over PXE alltogether :)
[13:39:58] <_zap_> ziph: i was also looking for quite some time and could not find one
[13:40:07] <_zap_> but high speed rs232 might be a viable alternative
[13:40:14] <_zap_> i did not look into that
[13:40:50] <ziph> _zap_: The other issue is getting your OS to deal with is.
[13:40:59] <ziph> _zap_: The average PC UART has tiny FIFO's.
[13:41:24] <_zap_> ziph: i thought of a usb2high_speed_rs232
[13:41:42] <_zap_> ziph: maybe there is such a thing. i never checked that
[13:48:16] <RikusW> at32uc3a3256 got hi-speed usb
[13:48:27] <RikusW> its xplain board cost $29
[13:50:34] <_zap_> RikusW: but only 1 usb transceiver if i am not mistaken
[13:50:42] <Kevin`> _zap_: 10mbit is possible with FS usb, no? although you might need some very tight code to do it
[13:52:00] <RikusW> you'll need 2 and a SPI connection
[13:52:27] <_zap_> right
[13:53:01] <_zap_> fs usb would be fine
[13:53:10] <_zap_> however something like
http://www.mev.co.uk/pages/Products/USB-Serial-Converters.html
[13:53:17] <_zap_> might do the trick as well
[13:53:37] <OndraSter> ft232 :)
[13:54:00] <RikusW> how about FT2232H
[13:54:21] <OndraSter> ft232 will push barely 2Mbaud
[13:54:25] <OndraSter> maybe 3MBaud with a bit of luck
[13:54:29] <OndraSter> (not with regular 16MHz atmega)
[13:57:20] <_zap_> hm, maybe there is one that does 10mbaud
[13:57:41] <_zap_> i am just wondering whether for rs232 1 baud = 1bit/s
[13:58:22] <Kevin`> _zap_: it is, there's no encoding except for the start/stop and such
[13:58:44] <_zap_> then 10 mbaud would be fine
[13:59:48] <LoRez> _zap_: not exactly. 9600baud is 960 characters/second
[14:00:15] <LoRez> you lose 20% of the bits
[14:01:32] <_zap_> LoRez: thx. still 10 mbaud would be find under these constraints
[14:05:43] <_zap_> it seems rs485 or rs422 might be something to look at
[15:41:35] <cyanide> HELLO LADIES
[15:42:14] <asteve> attack of the beepbeeps!
[15:42:21] <asteve> beepbeep!
[15:42:21] <beepbeep> beepbeep beepbeep!
[15:42:29] <dirty_d> vegas baby, vegas
[15:42:30] <dirty_d> lol
[15:53:06] <Landon> cyanide: hello chemicals!
[16:00:52] <cyanide> CapnKernel, you there?
[16:05:23] <OndraSter> no
[16:05:25] <OndraSter> he is asleep
[16:51:34] * OndraSter is checking out PLL to use with cheap and simple CPLDs that do not have PLL built in
[17:19:10] <Metalsutton> Hello all, is anyone avalible to throw me a bit of advice?
[17:19:51] <Metalsutton> I am quite new to the AVR game. I have made a few working basic circuits on a breadboard. USB to SNES apaptors, USB to Sega genesis adaptors.
[17:20:55] <Landon> well, spit it out!
[17:20:59] <Landon> ;)
[17:21:58] <Metalsutton> The problem is now when I have recreated one of the board on a prototypeing stripboard, it no longer wants to work, and I am sure it is something basic that I am missing, I have checked all solders and I have a feeling it has something to do with the way I have routed some paths. Can anyone recommend me a good way to troubleshoot the problem? Like software that displays a strip boards paths, or
[17:21:58] <Metalsutton> something? I dont know where to start.
[17:23:20] <Metalsutton> I have checked the VCC/+5 paths with a multimeter and everything checks out fine. The chip is programmed correctly, however it still detects as a malfunctioning usb device.
[17:25:39] <OndraSter> "my car doesn't want to start. Do you have any tips?"
[17:25:44] <OndraSter> give us something to work with :-)
[17:26:43] <Metalsutton> Like what? The only thing I could possibaly do would be to draw up a diagram of the current circuit. .... Well, I guess I better get started on that then :)
[17:31:31] <vectory> if you re_created it, the protoboard never worked in the first place
[18:30:13] <chupas> anyone ever use an Xmega for USB virtual serial?
[19:09:01] <Tom_itx> CapnKernel, boards look good
[19:09:21] <Tom_itx> routed edges, not vscored
[20:56:50] <learningc> anyone good with programming LCD screens?
[20:59:00] <Roklobsta> i have 2 lcd's for programing. A 24" and the laptop's 15"
[21:00:34] <Landon> hohoho
[21:00:41] <Landon> learningc: what kind?
[21:01:16] * Landon is using an hd44780 screen in his latest project
[21:01:28] <learningc> graphic kind
[21:02:19] <learningc> hi Roklobsta :)
[21:02:43] <Roklobsta> SPI interface?
[21:02:59] <Roklobsta> hd44780 is LCD with training wheels
[21:03:09] <Roklobsta> <so rude!>
[21:03:19] <learningc> not sure yet, those TFT kind you can find on ebay..
[21:03:53] <learningc> I think they have rgb input?
[21:11:37] <SianaGearz> you can find everything and then some on ebay.
[21:12:06] <SianaGearz> if you mean laptop screens, hello, they have lvds inputs -.- it's suicide.
[21:12:59] <learningc> what kind of input those small 3.2 inches tft screens have?
[21:13:15] <SianaGearz> oh. dunno. depends. show?
[21:13:29] <learningc> is there a standard in lcd inputs?
[21:13:35] <SianaGearz> probably something manageable.
[21:13:49] <SianaGearz> no, for those kinds of things, there isn't.
[21:14:18] <SianaGearz> well manageable with a PLD, not sure whether manageable with a microcontroller :P
[21:14:34] <learningc> http://www.ebay.ca/itm/230684823376?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2649
[21:14:41] <Roklobsta> just find one with an SPI input.
[21:14:55] <Roklobsta> how manyXbyY pixels?
[21:15:09] <learningc> qvga I think
[21:15:51] <SianaGearz> see controller data sheet. ta duuum.
[21:16:07] <Roklobsta> for exampl:
http://www.ebay.com.au/itm/132x132-pix-Serial-Color-Graphic-LCD-SPI-PIC-ARM-AVR-/400078962224
[21:17:11] <SianaGearz> haaa, nokia displays, a classic.
[21:17:23] <SianaGearz> they are awesome to work with.
[21:18:49] <Roklobsta> learninc: if it's a 16 bit data bus you just need to multiplex with a latch. it might have an 8 bit mode though. i'd say it'd be fairly simple/
[21:19:11] <Roklobsta> some have some 2d drawing routines in them.
[21:19:47] <Roklobsta> have a look at the range here for some ideas:
http://www.4dsystems.com.au/
[21:22:49] <Roklobsta> learncing: the pain will be coming up with a nice graphics lib
[21:26:29] <SianaGearz> the ebay qvga display linked has 8-bit interface and and a full onboard framebuffer, so it's microcontroller-able, but it will take a good bit of work.