| 01:00.21 | *** join/#edev Dr{Who} (n=mathewss@dev.null.nutech.com) |
| 02:10.44 | *** join/#edev purl (i=ibot@pdpc/supporter/active/TimRiker/bot/apt) |
| 02:10.44 | *** topic/#edev is a channel for the discussion of hacking small consumer electronic devices, mostly ARM based; Hammer Development boards now available at http://www.tincantools.com/ |
| 02:23.57 | *** join/#edev Dr{Who} (n=mathewss@dev.null.nutech.com) |
| 02:57.31 | *** join/#edev purl (i=ibot@pdpc/supporter/active/TimRiker/bot/apt) |
| 02:57.31 | *** topic/#edev is a channel for the discussion of hacking small consumer electronic devices, mostly ARM based; Hammer Development boards now available at http://www.tincantools.com/ |
| 04:25.17 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 04:26.30 | *** join/#edev redblue (i=star@ppp039.108-253-207.mtl.mt.videotron.ca) |
| 04:55.19 | *** join/#edev harpal (n=harpal@124.125.76.195) |
| 06:14.31 | *** join/#edev steliosk (n=Stelios@athedsl-4396544.home.otenet.gr) |
| 06:28.44 | *** join/#edev Madsy (n=madsy@ti500710a342-1583.bb.online.no) |
| 06:41.28 | *** join/#edev harpal (n=harpal@124.125.76.195) |
| 07:24.04 | *** join/#edev harpal (n=harpal@124.125.76.195) |
| 07:26.40 | *** join/#edev hw_ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 07:58.36 | *** join/#edev linac (n=lin@121.41.36.133) |
| 08:22.26 | *** join/#edev howtoo (n=howtoo@195.159.62.66) |
| 08:27.32 | *** join/#edev MothaWork (n=monmotha@mail.forwardengineering.com) |
| 08:30.07 | *** join/#edev MothaWork_ (n=monmotha@mail.forwardengineering.com) |
| 09:30.36 | *** join/#edev Madsy (n=madsy@ti500710a342-1583.bb.online.no) |
| 09:30.45 | *** join/#edev lyakh (n=lyakh@p57BD28B0.dip0.t-ipconnect.de) |
| 10:06.28 | *** join/#edev redblue (i=star@ppp039.108-253-207.mtl.mt.videotron.ca) |
| 10:28.16 | *** join/#edev DrZimmerman (n=theo@tbaumgartner.broker.freenet6.net) |
| 10:59.14 | *** join/#edev Madsy (n=madsy@ti500710a342-1583.bb.online.no) |
| 11:37.37 | *** join/#edev sjhill (n=sjhill@real.realitydiluted.com) |
| 11:37.37 | *** mode/#edev [+o sjhill] by ChanServ |
| 11:47.20 | *** join/#edev ico2 (n=ico2@ACCB654F.ipt.aol.com) |
| 12:07.22 | *** join/#edev madsy_ (n=madsy@ti500710a342-1583.bb.online.no) |
| 12:52.49 | *** join/#edev DrZimmerman (n=theo@tbaumgartner.broker.freenet6.net) |
| 13:17.36 | *** join/#edev MothaWork__ (n=monmotha@mail.forwardengineering.com) |
| 13:27.31 | *** join/#edev gandhijee (n=root@host-66-202-34-165.spr.choiceone.net) |
| 13:36.43 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 13:37.14 | *** join/#edev chouimat|work (n=dieu@209.217.106.98) |
| 14:06.54 | *** join/#edev kajaca (n=rakhus@32.137.71.21) |
| 14:22.17 | prpplague | cbrake: http://blog.makezine.com/archive/2008/03/eyefi_dissected.html |
| 14:22.23 | prpplague | cbrake: ar6k based sd card |
| 14:26.11 | cbrake | prpplague: neat, thanks! |
| 14:26.30 | prpplague | cbrake: which chipset did you end up going with? |
| 14:27.29 | cbrake | prpplague: my involvement with that project is fairly minimal as the customer has an internal engineer up to speed now -- actually a relief for me as it was WinCE |
| 14:27.44 | cbrake | prpplague: last I heard the AR6K is the only one they have working, and it has bugs |
| 14:28.17 | prpplague | cbrake: inteeresting |
| 14:28.18 | cbrake | prpplague: at least with the 1.1 firmware. Basic functionality is working fine, but with wifi, there are way too many senarios and some don't work |
| 14:28.38 | cbrake | prpplague: apparently the 2.0 firmware is out for the AR6K, but it requires driver changes |
| 14:28.41 | prpplague | cbrake: we've gone with the 8385 on one project and the wi2wi 8686 based device on the other |
| 14:29.23 | cbrake | prpplague: customer is still working the the CE driver for the Marvell chips |
| 14:29.30 | chouimat|work | prpplague: I wonder if the same applied to engineer http://www.nytimes.com/2008/03/18/science/18beer.html?_r=2&oref=slogin&oref=slogin |
| 14:29.31 | cbrake | prpplague: what driver you using by now for your projects? |
| 14:30.28 | cbrake | prpplague: also, did you need any special test software for FCC certification? |
| 14:30.37 | prpplague | cbrake: the libertas driver in the main kernel |
| 14:30.45 | cbrake | prpplague: cool |
| 14:30.47 | prpplague | cbrake: we outsource our fcc testing |
| 14:31.04 | prpplague | cbrake: no special software |
| 14:31.30 | cbrake | prpplague: do you have to put the radio in cont transmit mode or anything like that? |
| 14:32.05 | prpplague | cbrake: no all we were require to do is connect to an app send send/receive some data and then disconnect |
| 14:32.56 | cbrake | is still confused by FCC certification |
| 14:33.27 | cbrake | prpplague: I hear different things from different people :-) |
| 14:33.57 | prpplague | cbrake: me too, but since i don't have to deal with it, i don't worry |
| 14:59.24 | *** join/#edev CosmicPenguin (n=nobody@163.181.251.103) |
| 14:59.24 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 15:00.22 | prpplague | CosmicPenguin: greetings |
| 15:00.28 | prpplague | hands CosmicPenguin a lefthand milk stout |
| 15:01.45 | CosmicPenguin | ohh - tasty |
| 15:01.46 | chouimat|work | is it good? |
| 15:01.55 | CosmicPenguin | best beer in america, yo |
| 15:02.03 | prpplague | amen |
| 15:02.07 | *** join/#edev KaV (n=kvirc@host86-138-79-90.range86-138.btcentralplus.com) |
| 15:02.21 | prpplague | CosmicPenguin: i'd actually say its the best beer i've ever tasted |
| 15:02.34 | CosmicPenguin | Lefthand brews are good from top to bottom |
| 15:02.52 | prpplague | just very very very expensive |
| 15:03.05 | CosmicPenguin | They had a special beer this winter that was only on tap at the brewery called the Double Smoked Goosinator - just damn good |
| 15:03.22 | prpplague | CosmicPenguin: oh nice sounding |
| 15:03.42 | CosmicPenguin | Expensive though - 30 bucks for a growler |
| 15:04.01 | prpplague | CosmicPenguin: growler? |
| 15:04.04 | CosmicPenguin | gallon |
| 15:04.12 | prpplague | ahh |
| 15:04.13 | chouimat|work | not bad |
| 15:04.34 | CosmicPenguin | I have two growlers at home - I go in every month or so and get them filled |
| 15:04.42 | *** part/#edev KaV (n=kvirc@host86-138-79-90.range86-138.btcentralplus.com) |
| 15:04.46 | prpplague | CosmicPenguin: 750ml bottles of lefthand stuff is $15.99 and the six packs are $25.99 |
| 15:04.57 | prpplague | CosmicPenguin: those are illegal in texas |
| 15:05.11 | CosmicPenguin | man - if you knew how tiny the lefthand operation is, you would be surprised you could get bottles at all |
| 15:05.26 | prpplague | CosmicPenguin: i'm sure |
| 15:05.48 | prpplague | CosmicPenguin: it takes alot of effort to get a beer into texas, so i'm totally amazed |
| 15:06.27 | prpplague | CosmicPenguin: they have those growlers in the seattle area as well |
| 15:06.37 | prpplague | CosmicPenguin: used to see people getting them at the brew pubs |
| 15:07.13 | prpplague | CosmicPenguin: i think most people in texas would freak out if they saw someone getting a gallon of beer in a jug |
| 15:08.02 | *** join/#edev landley (n=landley@adsl-68-93-84-18.dsl.hstntx.swbell.net) |
| 15:08.02 | *** mode/#edev [+o landley] by ChanServ |
| 15:08.19 | CosmicPenguin | prpplague: yeah - growlers rock |
| 15:08.20 | prpplague | landley: holy houston batman |
| 15:09.10 | prpplague | CosmicPenguin: http://www.rahrbrewing.com/beer.html |
| 15:09.16 | chouimat|work | prpplague: http://en.wikipedia.org/wiki/Beer_bottle |
| 15:09.16 | prpplague | CosmicPenguin: only brewery in the area |
| 15:10.30 | CosmicPenguin | prpplague: really? So where is Shiner Bock and the other breweries? closer to San Antonio? |
| 15:10.49 | CosmicPenguin | I think I've had the Ugly Pug before |
| 15:11.20 | CosmicPenguin | AT the GABF, so its likely I don't remember it very clearly |
| 15:11.37 | prpplague | CosmicPenguin: yea mostly around austin |
| 15:12.06 | prpplague | CosmicPenguin: shiner is actually mostly owned by AB now |
| 15:12.40 | prpplague | CosmicPenguin: there are a couple of breweries in the houston area |
| 15:12.52 | CosmicPenguin | The bottle opener on my keychain is a "key" to the shiner bock brewery |
| 15:12.55 | prpplague | CosmicPenguin: but all the beers i've had from texas are very very very week on flavor |
| 15:12.56 | CosmicPenguin | I got it at my first GABF |
| 15:13.01 | prpplague | hehe |
| 15:13.37 | prpplague | st.arnolds, shiner, rahr |
| 15:13.46 | chouimat|work | prpplague: poor you |
| 15:13.58 | prpplague | chouimat|work: hehe |
| 15:14.09 | prpplague | chouimat|work: well fortunately i have a good beer store in the area |
| 15:14.38 | CosmicPenguin | The flying saucer? |
| 15:14.42 | prpplague | chouimat|work: the thing that always makes me laugh is that any beer from outside of texas is marked "imported" |
| 15:14.44 | chouimat|work | prpplague: and I guess they are glad you are there because you help them stay in business ;) |
| 15:15.05 | prpplague | CosmicPenguin: yea the FS has a really nice selection, but you can't buy beer to take home from the FS |
| 15:15.18 | CosmicPenguin | the important thing is that it gets in your belly |
| 15:15.26 | prpplague | CosmicPenguin: indeed |
| 15:15.51 | prpplague | CosmicPenguin: well with the FS i'd still have to drive home |
| 15:15.55 | chouimat|work | feels like having a smoke meat sandwich and a guinness for lunch |
| 15:16.32 | prpplague | http://www.ratebeer.com/beer/breweries/brewers-directory-43-213.htm |
| 15:16.39 | prpplague | check out how many are out of business |
| 15:17.57 | *** join/#edev linac (n=lin@121.41.36.133) |
| 15:18.12 | prpplague | wow there is a new one in grapevine |
| 15:18.15 | chouimat|work | prpplague: those one are nice http://www.ratebeer.com/beer/brick-waterloo-dark/7131/ http://www.ratebeer.com/beer/hockley-valley-dark/31037/ |
| 15:19.58 | prpplague | CosmicPenguin: holy cow, bigbuck brewing has special dispensation to use growlers |
| 15:21.30 | CosmicPenguin | get on down there |
| 15:21.39 | prpplague | CosmicPenguin: hehe up there actually, hehe |
| 15:22.03 | CosmicPenguin | all texas is down there to me |
| 15:23.27 | prpplague | hehe |
| 15:37.31 | *** join/#edev T0mW (n=Tom@24.115.218.135.res-cmts.sth.ptd.net) |
| 15:43.58 | landley | prpplague: hiya. |
| 15:57.33 | *** join/#edev Madsy (n=madsy@ti500710a342-1583.bb.online.no) |
| 16:03.41 | *** mode/#edev [+o prpplague] by ChanServ |
| 16:04.13 | prpplague | T0mW: greetings? |
| 16:04.27 | T0mW | prpplague: good morning |
| 16:04.56 | T0mW | prpplague: just messing with the Zipit today. |
| 16:05.02 | prpplague | T0mW: fun fun |
| 16:05.16 | prpplague | T0mW: i've been meaning to pick up one to tinker with the wifi |
| 16:05.39 | T0mW | prpplague: yeah, I can get wifi working but ssh keeps telling me that setlogin failed: function not implemented. |
| 16:06.40 | T0mW | prpplague: it is related to the angstrom build, I can build another distro type and it works fine. |
| 16:06.48 | T0mW | something to do with glibc, I think |
| 16:12.20 | *** join/#edev ico2 (n=ico2@modem-4048.cougar.dialup.pol.co.uk) |
| 16:19.14 | *** join/#edev ico2 (n=ico2@modem-4048.cougar.dialup.pol.co.uk) |
| 17:02.52 | *** join/#edev ico2_ (n=ico2@modem-2537.chameleon.dialup.pol.co.uk) |
| 17:06.37 | *** join/#edev Redhatter (n=vk4fsjl@2001:388:f000:0:0:0:0:279) |
| 17:10.27 | *** join/#edev landley (n=landley@adsl-68-93-84-18.dsl.hstntx.swbell.net) |
| 17:10.27 | *** mode/#edev [+o landley] by ChanServ |
| 17:28.05 | prpplague | well shoot, gonna have to file an amended tax return |
| 17:28.26 | prpplague | i just got a 1099 for some interest i got on an account, just $125 |
| 17:28.41 | prpplague | they sent it to my old address :( |
| 17:29.10 | prpplague | wonders if he will get audited because of the missing $125 |
| 17:35.46 | *** join/#edev lyakh (n=lyakh@p57BD28B0.dip0.t-ipconnect.de) |
| 17:36.06 | landley | prpplague: no. |
| 17:36.26 | landley | The IRS computer system is toast, audits are random. |
| 17:36.46 | landley | My grandfather was involved in trying to implement a new computer system for the IRS just before Y2K. |
| 17:36.54 | prpplague | landley: thats what i figured, but i figured having a 1099 that wasn't declared would be a red-flag |
| 17:37.02 | landley | The project was one of the more spectacular failures washington has seen, which is saying something. |
| 17:37.11 | prpplague | hehe |
| 17:37.15 | landley | The 2000 tax returns were literally done on paper, and beyond that I dunno what they're doing. |
| 17:37.40 | landley | My grandfather pulled out a few months before the end (telling me how badly it was going to crater and how he just couldn't get the bureaucrats to listen) and retired to florida... |
| 17:38.14 | landley | It's possible that in the 8 years since then they've gotten something coherent installed. But unlikely. |
| 17:38.17 | landley | I haven't heard about it if so... |
| 17:38.53 | landley | prpplague: sorry I didn't get to test tinycc on the nail last night. |
| 17:38.59 | landley | I took a nap and woke up in the morning. |
| 17:39.09 | landley | (11 hours of sleep. Apparently, I needed it.) |
| 17:39.11 | prpplague | landley: np, i have a serious case of F-its right now |
| 17:39.34 | landley | It's too bad, the nail is a nice little board. |
| 17:40.07 | prpplague | landley: yea, i'll be more motived next week to finish up what needs to be done for final production |
| 17:40.16 | prpplague | landley: things at the office have me down |
| 17:40.20 | landley | If they got some VC or something behind it and tried to go for maybe a $79 price point... |
| 17:40.30 | landley | The big differential between success and failure here is going to be price. |
| 17:41.03 | landley | I'll be demoing the thing as part of my cross compiling tutorial at CELF next month, by the way. |
| 17:41.04 | prpplague | indeed |
| 17:41.13 | prpplague | landley: ohh cool |
| 17:41.29 | prpplague | landley: yea we'll be doing the starting price at $189 |
| 17:41.40 | landley | Alas. |
| 17:41.43 | prpplague | landley: i've got mine set up to x sessions over ssh |
| 17:41.51 | landley | While that's cool, it seriously limits the volume you can expect. :( |
| 17:42.12 | landley | prpplague: is that running from the initrd, or do you have an actual filesystem on the flash? |
| 17:42.18 | Crofton | <PROTECTED> |
| 17:42.21 | prpplague | landley: indeed, i've been talking to the boss about how we can lower the costs |
| 17:42.24 | Crofton | oops wrong channel |
| 17:42.44 | prpplague | landley: i'm using the initrd to start up then chroot to a jffs2 root on the flash |
| 17:43.48 | prpplague | landley: tinkering with a small OLED board to go into the case with the nail |
| 17:43.56 | landley | hmmm? |
| 17:44.03 | prpplague | landley: just a small display |
| 17:44.06 | landley | This would use the expansion slot? |
| 17:44.15 | prpplague | landley: sjhill wanted a small display |
| 17:44.16 | prpplague | landley: yea |
| 17:44.36 | landley | I don't suppose there's any way to put an rtc on the oled board? :) |
| 17:44.40 | prpplague | landley: you can see how i have the lcd in the clear case at the bottom of the page http://www.elinux.org/Nail_Board |
| 17:44.50 | prpplague | landley: yea, i've been looking at that too |
| 17:45.12 | landley | A nice advantage of doing a clear case anyway. :) |
| 17:45.23 | prpplague | indeed |
| 17:45.49 | prpplague | ha funny, just as i was talking about the OLED display the quotes for the parts just came in |
| 17:46.28 | landley | What driver would that display use? |
| 17:46.34 | landley | (Framebuffer?) |
| 17:47.16 | landley | It occurs to me you could plug a USB keyboard into the nail's downstream port and put a text mode display on the framebuffer... :) |
| 17:47.26 | prpplague | landley: yea, it would use the standard framebuffer interface |
| 17:47.35 | prpplague | landley: yep |
| 17:47.55 | landley | What driver would provide said framebuffer? |
| 17:48.25 | prpplague | landley: i'd have to modify one of the existing drivers |
| 17:48.41 | prpplague | landley: there are a number that already exist for that style of interface |
| 17:48.51 | prpplague | landley: nothing major |
| 17:51.26 | landley | Again, needs to go upstream... :) |
| 17:51.44 | landley | At least cc: me on 'em so I can acked-by: the things. |
| 17:52.30 | prpplague | landley: yea, alot of that stuff doesn't get pushed upstream by developers, they consider it too specific for inclusing in the main tree |
| 17:53.35 | prpplague | landley: generally in the embedded environment most people just post the code somewhere and people do "whatever" with it |
| 18:00.06 | landley | If your device can work with a stock kernel, it's perceived as both more widely used and more general purpose than otherwise. |
| 18:00.38 | landley | Besides, it makes upgrading to newer kernel versions relatively painless if the company behind the device goes away or loses interest. |
| 18:00.52 | landley | So from a "selling enough to bring the price down" perspective, getting your patches in the kernel is kind of important. |
| 18:00.58 | landley | (It's a weirdly technical marketing thing.) |
| 18:01.20 | landley | (The "besides" above equals "safer".) |
| 18:03.55 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 18:06.47 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 18:08.35 | RobotGuy | Interesting buildroot situation.. It is not adding Python, PHP, or TCL to my rootfs for some reason. It makes me wonder what else is not being put in. |
| 18:08.51 | landley | Any opinions on http://www.itwire.com/content/view/17183/53/ ? |
| 18:08.54 | prpplague | landley: yea, i'll submit the patches, but in most cases like with AML, it's simply not worth our time to submit patches |
| 18:09.00 | prpplague | looks |
| 18:09.38 | prpplague | landley: yea i was reading about that |
| 18:09.56 | RobotGuy | Why is it not worthh the time to submit patches?? |
| 18:10.02 | prpplague | landley: problem is right now there just aren't any flash or sdram that operate at the 0.3v level |
| 18:10.54 | landley | I'm still waiting for cereal boxes with an active display on the front. |
| 18:10.56 | prpplague | RobotGuy: because most companies like AML who produce a device rarely upgrade the kernel or update versions of the existing software |
| 18:11.00 | landley | Disposable computing is just a matter of time... |
| 18:11.19 | prpplague | RobotGuy: once the device "works" companies don't spend any more time than they have to on the device |
| 18:11.31 | RobotGuy | prpplague: Oh, OK. Too bad. |
| 18:11.41 | prpplague | RobotGuy: the only time they get upgraded is if a bug is found or a new feature requires it |
| 18:11.44 | RobotGuy | It makes sense though, unfortunately. |
| 18:11.51 | landley | Nail is different. You're offering general purpose computing infrastructure for third parties to do their own thing with. |
| 18:11.56 | landley | That's gotta be merged to matter. |
| 18:12.00 | prpplague | landley: indeed |
| 18:12.05 | prpplague | landley: and i agree |
| 18:12.18 | RobotGuy | I can't get a 2.6.24.3 kernel to build. |
| 18:12.32 | prpplague | RobotGuy: why not? |
| 18:13.14 | RobotGuy | prpplague: I don't know. I will pastebin the error. I just added the hammer hardware .c file and that one line to include it. |
| 18:14.03 | prpplague | landley: by going to such a low voltage on the core, power savings with be dramatic |
| 18:14.36 | RobotGuy | prpplague: http://www.pastebin.ca/949249 |
| 18:15.06 | chouimat|work | cool cell phone and laptop with a 2 years battery life between charges |
| 18:17.56 | landley | chouimat: unlikely. |
| 18:18.17 | landley | The cell phone still needs to transmit a certain signal strength, and that eats an amount of power determined by the laws of physics. |
| 18:18.27 | SpeedEvil | No it's not |
| 18:18.39 | SpeedEvil | It's very weakly determined |
| 18:18.49 | landley | Maybe you could manage a 2 year standby time... |
| 18:18.57 | SpeedEvil | For example, beamforming antennas at 60GHz and mesh networks utterly change things. |
| 18:19.16 | SpeedEvil | (in terms of absolute fundamental power usage) |
| 18:19.21 | landley | Mesh networks have security implications unless you're encrypting all your data. |
| 18:19.42 | SpeedEvil | Sure. End-end encryption'd be pretty much required, as would micropayment |
| 18:19.44 | prpplague | RobotGuy: looks like its trying to compile with gcc not arm-linux-uclibc-gcc |
| 18:19.44 | landley | And directional antennas are a bit hard to do in a moving device with current technology. (I'm all for it, but show me how you direct the beam in software without moving parts.) |
| 18:19.52 | SpeedEvil | with current technology |
| 18:19.57 | landley | This is speaking as someone whose father worked on the Aegis phased array radar system that _did_ manage to do it. |
| 18:19.59 | SpeedEvil | landley: easy - not cheap |
| 18:20.03 | RobotGuy | prpplague: Hmmmmm. I set ARCH = arm in the Makefile. |
| 18:20.17 | prpplague | RobotGuy: did you set the compiler ? |
| 18:20.19 | landley | SpeedEvil: any solution that involves micropayment is immediately disqualified. |
| 18:20.23 | landley | It will not work, for social reasons. |
| 18:20.29 | landley | People keep sucking on that pipe dream... |
| 18:20.39 | SpeedEvil | landley: there was recently a single chip that beamformed 32 outputs of RF |
| 18:20.45 | SpeedEvil | at 16(?) GHz |
| 18:21.17 | SpeedEvil | Micropayment in terms of some fair mesh scheduling algorithm. |
| 18:21.49 | SpeedEvil | Scary GaAS stupidly expensive prototype, but it's first silicon. |
| 18:21.55 | RobotGuy | prpplague: That's the problem alright. It's basing the compiler to be used on the host arch. |
| 18:22.03 | landley | Chouimat: anyway, I'd expect increases in standby time but not that much in talk time. |
| 18:22.18 | landley | SpeedEvil: neither of us will live to see micropayments matter. Period. |
| 18:22.33 | chouimat|work | landley: :) |
| 18:22.39 | SpeedEvil | landley: bittorrent implements it to some degree |
| 18:23.08 | landley | Also, as I was trying to say; laptops which need a backlit display will have a bit of a problem with 2 year battery life. |
| 18:23.27 | landley | SpeedEvil: it doesn't matter. |
| 18:23.31 | SpeedEvil | landley: I mean micropayment in the loosest sense - any way of sharing a contested resource that resists individual nodes doing bad stuff. |
| 18:23.43 | landley | SpeedEvil: so you want to redefine micropayment. |
| 18:23.44 | chouimat|work | landley: I know and it was a sarcastic comment :) |
| 18:24.09 | SpeedEvil | landley: it all comes down to cash in the end if you're redistributing a paid-for service. |
| 18:24.29 | SpeedEvil | landley: even if the user never sees that cash directly. |
| 18:24.35 | landley | You want to redefine micropayment to mean resource contention. |
| 18:24.44 | SpeedEvil | I picked the wrong term. |
| 18:25.10 | landley | Resource contention is a very active field, and yeah bittorrent does some interesting stuff there. |
| 18:25.26 | landley | Resource contention in mesh networks is an area that's hardly scratched the surface. |
| 18:25.38 | landley | OLPC might give us some mesh networks big enough to experiment with. |
| 18:25.57 | landley | There's a chicken and egg problem with mesh networks: until they reach a critical mass they don't work. |
| 18:26.02 | prpplague | i'm totally surprise WIAV hasn't gone after OLPC |
| 18:26.17 | prpplague | i guess they are just getting their war chest stocked first |
| 18:26.20 | landley | What's a WIAV? |
| 18:26.24 | prpplague | ~WIAV |
| 18:26.25 | purl | somebody said wiav was a patent troll with information at http://trolltracker.blogspot.com/2008/01/4-interesting-new-cases-from-last-2.html and http://www.setexasrecord.com/news/205753-texas-company-seeks-judgement-to-stop-potential-patent-suit |
| 18:26.27 | RobotGuy | I just don't know how to 'fix' it to be ARM instead of i686. |
| 18:26.34 | SpeedEvil | No, I was using it in the sense of a way of resolving costs through a contended network using some sort of money-like token to 'bid' for connectivity. |
| 18:27.01 | landley | prpplague: it would be a PR nightmare for them. |
| 18:27.11 | landley | Patent trolls have to optimize money received vs attention attracted. |
| 18:27.14 | prpplague | RobotGuy: CROSS_COMPILE?= arm-linux-uclibc- |
| 18:27.30 | SpeedEvil | landley: it's indeed horribly complex - mesh networks aren't going to work without at least several different layers at diffeerent scales, or local uplinks to the internet. |
| 18:27.49 | prpplague | landley: yea, thats why i figured they were waiting till they have a good amount of money first |
| 18:27.53 | landley | The last thing they want is a new york times editorial urging people to write their congressmen to get the patent office to review the validity of their patents. |
| 18:28.04 | SpeedEvil | landley: for global connections that is - if you only ever want to use them to talk to people 200m away, they work well. |
| 18:28.30 | prpplague | landley: the thing is, the patents apply 100% to the OLPC mesh networking |
| 18:28.35 | landley | I doubt they'd go after OLPC period. The attention+outrage vs potential cash ratio is all wrong for troll bait. |
| 18:29.02 | prpplague | landley: or atleast not for awhile |
| 18:29.02 | SpeedEvil | But if you don't defend a patent against one infringer, isn't it invalid? |
| 18:29.12 | landley | prpplague: probably not ever. There's nothing in it for them. |
| 18:29.12 | RobotGuy | prpplague: Ahhhh! I had to set ARCH=arm and CROSS_COMPILE |
| 18:29.13 | prpplague | SpeedEvil: ?? |
| 18:29.13 | SpeedEvil | an infringer that you know of. |
| 18:29.47 | prpplague | SpeedEvil: sorry not understanding you |
| 18:29.51 | landley | SpeedEvil: to talk to people 200m away 801.11b works fine. |
| 18:30.04 | landley | SpeedEvil: that's trademarks. |
| 18:30.08 | SpeedEvil | I mean - don't you have to take action against infringers of your patent, or you have limited recourse against those that you choose to take action against - you cant |
| 18:30.09 | SpeedEvil | oops |
| 18:30.11 | landley | Patents you don't have to defend at all and they remain valid. |
| 18:30.17 | landley | Copyrights too. |
| 18:30.22 | SpeedEvil | landley: not in urban environments |
| 18:30.36 | RobotGuy | prpplague: Still won't build and I just don't have the patience to mess with it. :( :( |
| 18:30.39 | SpeedEvil | landley: congested wireless environments that is. |
| 18:30.41 | RobotGuy | sighs |
| 18:30.42 | landley | Trademark law got seriously weakened during World War II when we grabbed Bayer's patents as reparations from the germans and rationalized it as the patents not being defended. |
| 18:30.54 | SpeedEvil | Ah |
| 18:31.42 | prpplague | SpeedEvil: not sure i understand you, but the main problem with the WIAV patents are that they are trying to state they apply to all 802.11, which is incorrect |
| 18:31.54 | prpplague | SpeedEvil: but they _do_ apply to most mesh networks |
| 18:32.14 | SpeedEvil | Hmm. From memory, I thil so. |
| 18:32.18 | SpeedEvil | think |
| 18:32.29 | prpplague | SpeedEvil: think so what? |
| 18:32.38 | SpeedEvil | prpplague: agree with yuou |
| 18:32.42 | prpplague | SpeedEvil: oh |
| 18:33.05 | prpplague | well we are going to settle with WIAV so they can go and attack others |
| 18:33.21 | prpplague | real shame, but it just doesn't make any business sense |
| 18:33.32 | SpeedEvil | prpplague: I was meaning earlier that I thought that you couldn't let someone infringe your patent without taking action, or you lost the right to attack later infringers. However, I was thinking of trademarks. |
| 18:33.45 | chouimat|work | prpplague: ouch |
| 18:34.24 | prpplague | chouimat|work: yea well, management did the math, and its just not worth our time and money to defend everyone else |
| 18:34.38 | prpplague | chouimat|work: we've had zero help from the OSS community |
| 18:34.39 | SpeedEvil | :/ |
| 18:34.40 | chouimat|work | prpplague: I agree with them |
| 18:35.04 | chouimat|work | prpplague: OSS community is not really helpful most of the time |
| 18:35.10 | SpeedEvil | prpplague: there was almost zero accessible info unfortunatley. |
| 18:35.27 | prpplague | we get the "you are an evil business, you've got money defend yourself" |
| 18:35.34 | SpeedEvil | prpplague: in the sense of a page that you could point someone to, and they'd grasp the issue in 2 mins. |
| 18:35.54 | SpeedEvil | So it's tricky to motivate if they have to parse the patent themselves. |
| 18:36.09 | prpplague | SpeedEvil: yea well we approached just about every OSS legal orgination there is |
| 18:36.14 | SpeedEvil | :/ |
| 18:36.39 | SpeedEvil | I mean getting random linux users interested. /. et al. |
| 18:36.51 | prpplague | SpeedEvil: yea, i'm talking more of groups |
| 18:38.32 | chouimat|work | prpplague: as the great poet RMS once said: Business are evil |
| 18:39.03 | chouimat|work | and human can live with nothing but fresh air |
| 18:40.13 | prpplague | chouimat|work: yea, that seems to be the attitude when it comes to stuff like WIAV |
| 18:40.58 | SpeedEvil | The buisness is evil mindset works fine if you never want the buisnesses to do nice things for you. |
| 18:41.10 | chouimat|work | true |
| 18:41.13 | SpeedEvil | Like document their hardware, or ... |
| 18:41.43 | chouimat|work | SpeedEvil: this same mindset produce "good" thing like GPLv3 |
| 18:42.29 | SpeedEvil | Yeah. |
| 18:42.50 | chouimat|work | hates v3 |
| 18:43.13 | SpeedEvil | I'm really unsure how I feel on the locked firmware issue. |
| 18:44.23 | SpeedEvil | For example, in cases where the hardware is subsidised by later software sales, I think it's not unfair. |
| 18:44.30 | SpeedEvil | As long as the user can pay to unlock it. |
| 18:45.10 | chouimat|work | SpeedEvil: true but there some case where law prevent you of giving non locked firmware |
| 18:45.25 | SpeedEvil | ? |
| 18:45.38 | SpeedEvil | You mean hardware drivers? |
| 18:45.56 | SpeedEvil | Or stuff like car ABS systems running linux. |
| 18:46.08 | chouimat|work | SpeedEvil: second |
| 18:46.09 | prpplague | or stuff like firmware for radio devices |
| 18:46.27 | chouimat|work | prpplague: that too ... |
| 18:46.44 | SpeedEvil | That can somewhat be addressed by binary drivers for the radio |
| 18:47.48 | SpeedEvil | But the ABS case pretty much can't. |
| 18:48.04 | SpeedEvil | Other than virtual machines that the user can play in. |
| 18:49.45 | chouimat|work | SpeedEvil: but I guess some "mechanics" will complain about this fact |
| 18:49.55 | SpeedEvil | holds up his megasquirt. |
| 18:50.01 | SpeedEvil | (which I've still not assembled) |
| 18:50.07 | SpeedEvil | DIY ECU with source. |
| 18:51.24 | SpeedEvil | And it's deeply annoying that you can't hack on the dashboard of a modern car, as it means if you want to swap the ECU, you've pretty much got to destroy the existing dashboard and recreate it from scratch. |
| 18:53.01 | ds2 | SpeedEvil: I think in the US, the EPA disallows that |
| 18:54.45 | SpeedEvil | In the UK, it's allowed, as long as you don't screw up the emissions. |
| 18:55.11 | *** join/#edev bbradley (n=bbradley@82-32-35-42.cable.ubr05.azte.blueyonder.co.uk) |
| 18:55.11 | ds2 | but how do you prove that? |
| 18:55.34 | SpeedEvil | Actually, I don't think there is any requirement to have it in a low emission condition all the time. |
| 18:56.07 | SpeedEvil | The requirement is that it meets ppms of unburned hydrocarbons, CO, NOX, ... at idle and fast idle |
| 18:56.12 | SpeedEvil | when tested once a year |
| 18:56.57 | SpeedEvil | And I think there is a 'no visible smoke' requirement on a test drive |
| 18:58.18 | ds2 | so by that line of reasoning, you guys can take out the cat? |
| 18:58.37 | SpeedEvil | I'm not completely sure. |
| 19:00.21 | SpeedEvil | In general our vehicle testing is very lax. Basically, it's see if any bits of the suspension or brakes are clearly unsafe, on a visual inspecion only without dissasembly |
| 19:00.56 | ds2 | sounds like one can pass inspection by using plastic "metalized' parts ;) |
| 19:01.05 | SpeedEvil | For example 'structural rust within 30cm of a suspension component' or something is a fail |
| 19:01.18 | SpeedEvil | In germany, pretty much any rust anywhere on the car, and it's a fail. |
| 19:01.46 | prpplague | landley: your nail tube case should ship today |
| 19:03.33 | *** join/#edev ico2__ (n=ico2@modem-1602.cheetah.dialup.pol.co.uk) |
| 19:03.33 | landley | prpplague: woot. |
| 19:04.12 | prpplague | landley: someone be there friday to get the package? |
| 19:04.19 | landley | ds2: requirements vary by state. |
| 19:04.30 | landley | In california, if it burns gasoline at all, it's kind of suspect. |
| 19:06.00 | landley | prpplague: yeah, Fade. |
| 19:06.04 | landley | Actually, friday I'll be there. |
| 19:06.13 | landley | Friday's a holiday at work. I forget why. |
| 19:06.16 | landley | Don't have to be here. :) |
| 19:06.37 | chouimat|work | I think it's Good Friday :) |
| 19:08.21 | prpplague | landley: good friday |
| 19:09.11 | chouimat|work | is happy: 4 days week, a 4 days weekend and another 4 days week |
| 19:10.56 | landley | has jury duty next week. |
| 19:11.05 | landley | I wonder if I can bring my laptop? |
| 19:11.16 | landley | (It's just the selection part...) |
| 19:12.11 | prpplague | landley: how did you end up with jury duty? |
| 19:12.24 | prpplague | landley: you register to vote? |
| 19:12.37 | prpplague | landley: iirc thats how they get most people |
| 19:13.32 | prpplague | landley: ok package is out the door 2nd day UPS |
| 19:13.34 | chouimat|work | prpplague: I think the other way is to fill a tax form ;) |
| 19:13.38 | prpplague | landley: so you should get it friday |
| 19:14.14 | prpplague | chouimat|work: not in the US, tax form isn't information that can be released to either federal or state judicial agencies |
| 19:14.26 | ds2 | ewwwwwwwwwwwwww XP on XO |
| 19:14.33 | prpplague | chouimat|work: each state has their own methods |
| 19:14.39 | chouimat|work | prpplague: :) |
| 19:14.43 | prpplague | chouimat|work: some states use their driver licenses |
| 19:14.54 | prpplague | chouimat|work: others also use voter registration |
| 19:15.01 | ds2 | Tax forms doesn't ahve the info to determine if you can serve |
| 19:15.03 | chouimat|work | prpplague: in canada they voter registration |
| 19:15.26 | RobotGuy | I have been corresponding with this guy from Sprint who apparently works in their WiMax division. Talking to him about Hammer too. |
| 19:15.28 | prpplague | chouimat|work: ahh |
| 19:15.36 | prpplague | RobotGuy: oh yea? |
| 19:15.36 | chouimat|work | ds2: no date of birth? |
| 19:15.42 | RobotGuy | prpplague: Uh huh. :) |
| 19:15.48 | ds2 | chouimat|work: no citizen info |
| 19:16.00 | ds2 | even green cards have to file tax returns |
| 19:16.04 | prpplague | RobotGuy: to what end? he doing robotics? |
| 19:16.08 | RobotGuy | prpplague: The guy wants to make a mobile WiMax demo robot. |
| 19:16.15 | prpplague | ahh |
| 19:16.24 | prpplague | ds2: exactly |
| 19:16.37 | ds2 | if you do that, why not just use the phone book |
| 19:17.05 | RobotGuy | prpplague: I came up with an idea that takes his idea a few steps further. :) |
| 19:17.26 | prpplague | nice, good luck with it |
| 19:17.31 | ds2 | if they want a real demo... use WiMax as the link for a RC airplane ;) |
| 19:17.52 | ds2 | then fly it through a maze of fragile stuff (expensive wine bottles, etc) |
| 19:18.25 | RobotGuy | ds2: My idea is a swarm of WiMax and WiFi equipped robots running around a neighbor hood or much larger area, using WiMax for inter node links and WiFi B/G to connect people to the net. |
| 19:18.53 | ds2 | RobotGuy: but that says nothing about the latency characteristics |
| 19:18.54 | SpeedEvil | steals RobotGuys robots. |
| 19:19.24 | *** join/#edev neuromancer (n=neuroman@91.145.193.247) |
| 19:20.01 | RobotGuy | ds2: I don't know enough about WiMax to really talk much about it, but I would not mind latching on to a WiMax adapter if it is supported by Linux. :) |
| 19:20.47 | RobotGuy | ds2: I could network my whole apartment complex with a few roving access points. :) |
| 19:20.57 | RobotGuy | And more! |
| 19:21.24 | RobotGuy | It's just a crazy idea. |
| 19:21.30 | landley | prpplague: here in texas they take it from driver's license info. |
| 19:22.08 | RobotGuy | My idea actually has use though, like to make a mobile network link for rescue crews, even military. |
| 19:22.22 | prpplague | landley: is that state wide or does it depend on the county? |
| 19:22.29 | landley | state wide. |
| 19:22.43 | neuromancer | greetings! |
| 19:22.50 | landley | (I think so, anyway. The licenses are state-wide, and the state government has the info...) |
| 19:22.55 | landley | neuromancer: hiya. |
| 19:24.01 | RobotGuy | sighs |
| 19:24.23 | RobotGuy | Somebody will eventually DO my ideas, just like what always happens. |
| 19:25.29 | RobotGuy | Hammer could work well in a setup like my idea. |
| 19:26.14 | prpplague | landley: yea, looks like it is state wide for DL and voter registration |
| 19:26.34 | RobotGuy | Bye |
| 19:26.35 | prpplague | landley: if you are registered to vote and have DL you are more likely to get called |
| 19:26.39 | *** part/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 19:34.02 | landley | prpplague: *shrug* |
| 19:34.14 | landley | They've sent me things before but this is the first time they managed it to the right address. |
| 19:34.27 | landley | They seemed somewhat annoyed about the previous attempts, seems reasonable to just go with it. |
| 19:34.38 | prpplague | landley: i wonder how i never got called up for jury duty when i was over seas |
| 19:34.46 | prpplague | landley: i still had a texas DL |
| 19:35.01 | landley | If it keeps me away from C++ code that creates a bitfield to represent a char, puts it in a union with a structure that has a char member named "byte", and then leaves half the bitfield fields uninitalized... |
| 19:35.29 | landley | I've been in texas for 10 years. Jury duty happens sooner or later. |
| 19:35.52 | *** join/#edev neuromancer (n=neuroman@91.145.193.247) |
| 19:36.19 | prpplague | landley: yea i've been called 13 times since i turned 18 |
| 19:36.22 | prpplague | landley: never served |
| 19:37.18 | prpplague | landley: but it does make me wonder how i was called on for jury duty while i was overseas, maybe i'd been called enough already |
| 19:38.43 | landley | Let's see: I'm a member of mensa, intellectual property law is a hobby of mine, I've been a plaintiff in four civil suits in the past year, I considered becoming a paralegal as recently as 2005... |
| 19:38.59 | landley | Oh yeah, prime juror material. Right at the top of who they select. |
| 19:39.11 | landley | Easily predictable, easily controllable, somebody likely to take lawyers' arguments at face value... |
| 19:39.35 | *** join/#edev ico2 (n=ico2@modem-1602.cheetah.dialup.pol.co.uk) |
| 19:40.27 | prpplague | landley: hehe, exactly why i've never actually served |
| 19:40.45 | prpplague | landley: the last time back in 2003, the DA told me i'd never sit on a jury |
| 19:42.50 | chouimat|work | prpplague: and you were happy about this wonderfull news? |
| 19:44.08 | prpplague | chouimat|work: mixed, i really wanted to serve on a jury atleast once, but i really had better things to do, yet it makes me totally sad that the legal system depends on uneducated people who are easily swayed |
| 19:44.58 | chouimat|work | prpplague: I agree it's sad ... |
| 19:50.30 | neuromancer | landley: searching a solution extermal memory implementation |
| 19:51.46 | neuromancer | landley: need something of few MB, but dont know whether to use sram |
| 19:51.56 | neuromancer | landley: mostly becouse of it cost |
| 19:53.29 | prpplague | neuromancer: few mb has a wide range of solutions |
| 19:53.54 | prpplague | neuromancer: its going to depend on what I/O you have available, when you need to access the info, and at what speed |
| 19:54.26 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 19:54.28 | neuromancer | prpplague: well, a few i mean not actualy a few, but something like 32, preferably 64 or even more |
| 19:54.48 | RobotGuy | just created #tincantools and set it up as much like here as possible. |
| 19:55.23 | RobotGuy | For focused discussions of any of TinCanTools embedded products or projects. |
| 19:56.12 | RobotGuy | prpplague: You have maximum access to #tincantools now. |
| 19:56.26 | prpplague | RobotGuy: okie dokie |
| 19:56.30 | neuromancer | well, is it ok to use sdram controllers - do you use them? |
| 19:56.47 | prpplague | neuromancer: 32 or 64 is not a few, hehe |
| 19:57.08 | prpplague | neuromancer: you needing permanent storage or just ram? |
| 19:57.37 | neuromancer | prpplague: just ram, not eeprom, if you mean that |
| 19:57.43 | prpplague | oh |
| 19:57.55 | prpplague | neuromancer: well sdram is about your only choice due to cost |
| 19:58.02 | prpplague | neuromancer: for the size of 32 or 64mb |
| 19:58.17 | prpplague | neuromancer: your processor have a sdram controller? |
| 19:59.30 | neuromancer | well, i just can't get a processor with sdram conroller - the local sellers don't have any, so, i'm just confused what shall i do |
| 20:00.01 | prpplague | neuromancer: oh so you don't have an existing design |
| 20:00.18 | prpplague | neuromancer: what is it you are trying to do? |
| 20:00.22 | *** join/#edev linac (n=lin@121.41.34.232) |
| 20:00.43 | neuromancer | prpplague: well, at a time trying to figure out what it will be |
| 20:01.01 | prpplague | neuromancer: i'm sorry i did not understand your statement |
| 20:03.20 | chouimat|work | prpplague: http://www.cbc.ca/nextprimeminister/blog/2008/03/canadas_next_great_prime_minis.html |
| 20:04.50 | neuromancer | prpplague: i want to create a system with arm (at91sam7) as a core, maybe some adc/dac, usb, etc, and run linux on it |
| 20:05.27 | prpplague | neuromancer: unless you are experienced at doing pcb layout, it is highly unlike you will be able to do that level of work |
| 20:05.41 | prpplague | neuromancer: might i suggest you use a cpu module |
| 20:05.47 | neuromancer | prpplague: but dont know how to handle a problem with unsufficient ram |
| 20:06.13 | prpplague | neuromancer: the sam7 serious does not support the addition of external ram |
| 20:06.19 | prpplague | s/serious/series |
| 20:06.52 | prpplague | neuromancer: there are plenty of cpu modules that you can build around |
| 20:07.45 | neuromancer | prpplague: but there is sam7se32, se512 and se256 with external bus interface |
| 20:08.05 | prpplague | neuromancer: external bus is not the same as support for external sdram |
| 20:08.14 | neuromancer | prpplague: and maybe, some example, please |
| 20:08.20 | landley | neuromancer: what do you mean by "external memory implementation"? What should it do? |
| 20:08.41 | prpplague | landley: he's talking about adding something like 32mb of sdram to a device |
| 20:09.16 | RobotGuy | BRB |
| 20:09.31 | prpplague | neuromancer: http://www.armkits.com/ and http://www.tincantools.com |
| 20:09.40 | landley | A device that starts with how much? |
| 20:09.58 | prpplague | landley: around 256k internally |
| 20:10.05 | *** join/#edev nhoman (n=homann@89.100.83.14) |
| 20:10.19 | landley | Hmmm... Can you add more ram to nail via that expansion slot thingy? |
| 20:10.34 | chouimat|work | hmmm sound lot like a lpc2xxx |
| 20:10.43 | prpplague | landley: no, it would need to be added to the sdram controller bus |
| 20:10.44 | SpeedEvil | landley: no |
| 20:11.05 | prpplague | landley: adding sdram is very delicate in the aspect that the length of the traces are very critical |
| 20:11.14 | neuromancer | prpplague: just didn't meet any yet (cpu modules) |
| 20:11.20 | prpplague | landley: sdram support isn't for the inexperienced |
| 20:11.26 | landley | So it would require modifying the hammer module, and couldn't be done via a carrier like nail? |
| 20:11.36 | landley | prpplague: the inexperienced use SIMMs. :) |
| 20:11.37 | prpplague | landley: that is correct |
| 20:11.58 | prpplague | landley: even with laying out SIMM connectors the design is very very very delicate |
| 20:11.59 | neuromancer | landley: yep |
| 20:11.59 | neuromancer | landley: What should it do?: well, i'd like to make some bios or whatewer booting program to configure external memory to be used just like internal, from the executive application point of view |
| 20:12.14 | landley | And technically, in linux I could plug my 500 gig external USB drive into the nail and mmap the sucker, it just wouldn't help much. :) |
| 20:12.30 | SpeedEvil | Actually, it does. |
| 20:12.43 | *** join/#edev neuromancer (n=neuroman@91.145.193.247) |
| 20:12.53 | SpeedEvil | swap to usb1.1 to a fast storage media can be faster than swapping to hard disk |
| 20:12.54 | landley | SpeedEvil: dog slow. |
| 20:13.12 | landley | It _is_ a hard disk. |
| 20:13.27 | SpeedEvil | I mean to a fast RAM or flash disk |
| 20:13.39 | landley | Also, beyond a certain point adding more swap just ties up more of your internal ram to track the swap and you get diminishing returns on small systems. |
| 20:13.46 | SpeedEvil | yeah |
| 20:13.56 | landley | You wind up spending all your ram on page tables. It gets ugly. |
| 20:14.19 | prpplague | landley: the sdram bga package on the hammer is compatible with several different sizes |
| 20:14.29 | landley | Note sure where the sweet spot would be with hammer, though. I'd have to go do math... |
| 20:14.38 | prpplague | landley: if we want to make a version with larger sdram, we could change out the part |
| 20:14.47 | landley | What are the size options? |
| 20:15.05 | landley | And by "change out the part" you mean run a new line of hammer modules? |
| 20:15.08 | landley | Can't be done aftermarket? |
| 20:15.13 | SpeedEvil | yes |
| 20:15.14 | prpplague | landley: 32mb,64mb and 128mb |
| 20:15.23 | SpeedEvil | but it requires someone who's done it before |
| 20:15.29 | prpplague | landley: after market could be done, just not cost effective |
| 20:15.32 | SpeedEvil | and has the stuff |
| 20:15.44 | SpeedEvil | you basically need to hot-air remove the existing BGA - easy. |
| 20:15.52 | SpeedEvil | Clean the pads off - not too bad. |
| 20:15.59 | chouimat|work | waits for the hard part |
| 20:16.13 | SpeedEvil | Stencil on new solder paste - again not really hard, but requireing +-1mm registration or so |
| 20:16.16 | SpeedEvil | .1mm |
| 20:16.38 | SpeedEvil | then place the bga chip precisely on and reflow with the heat gun again |
| 20:16.49 | prpplague | landley: flash and sdram interfaces are delicate work, which is why there are so many cpu modules, the folks do all the hard dev on the module, allow other developers to work more with the peripherals which are not as sensative to specific design needs |
| 20:16.54 | chouimat|work | SpeedEvil: sound time consuming |
| 20:17.03 | SpeedEvil | chouimat|work: not really if you have the stuff. |
| 20:17.08 | SpeedEvil | chouimat|work: 5-10 mins |
| 20:17.16 | SpeedEvil | chouimat|work: if you know what you're doing. |
| 20:17.19 | prpplague | its way behind the average persons capability to do |
| 20:17.31 | chouimat|work | SpeedEvil: I see but since I'm really bad with my hands I will find someone to do it ;) |
| 20:17.32 | prpplague | both in cost and skills |
| 20:17.45 | SpeedEvil | I'd disagree with that - they could be trained to use the equipmentl. |
| 20:17.59 | SpeedEvil | But a vanishingly small number of people will have hte skils and equipment |
| 20:18.28 | SpeedEvil | Well - vanishingly small is a bit pessimistic, but not that far. |
| 20:18.29 | prpplague | SpeedEvil: you over estimate the skills over the average developer |
| 20:18.40 | landley | Ok, so if tincan wants to come out with nail boards with 128 megs of ram as an option, it wouldn't be too hard to belt out 100 of them at a time. |
| 20:18.46 | SpeedEvil | prpplague: you can train monkeys to use some of the reflow ones. |
| 20:18.55 | SpeedEvil | prpplague: but they are much more investment |
| 20:19.04 | prpplague | landley: that is correct |
| 20:19.05 | SpeedEvil | But I suspect you know this. |
| 20:19.08 | landley | Possibly it could even be an ordering option for $x dollars more, on the web form. |
| 20:19.15 | prpplague | landley: correct |
| 20:19.21 | landley | Good to know. |
| 20:19.31 | SpeedEvil | Is a 128M chip really that much more than 32? |
| 20:19.46 | landley | How does that affect the price, and the volume/price curve thingy? |
| 20:19.50 | prpplague | landley: about 50% more |
| 20:20.41 | prpplague | sorry i meant that the cost of a 128mb chip is 50% more than a 32 |
| 20:20.41 | landley | Roughly what volume would you have to sell to make the version with 128 megs come in around $89? |
| 20:20.52 | prpplague | landley: 10k |
| 20:21.02 | landley | Hmmm... |
| 20:21.13 | prpplague | landley: 10k is the magic number for the build |
| 20:21.15 | landley | So if you got one 10k order, you could offer it to everybody for $89? |
| 20:21.34 | prpplague | landley: at that qty, the price breaks make it a big difference |
| 20:21.38 | prpplague | landley: yea |
| 20:21.49 | landley | Hmmm... |
| 20:21.58 | SpeedEvil | I'd order 3 at $89 |
| 20:22.17 | prpplague | if we got an order for 10k , we'd probably do an actual build of 15k |
| 20:22.17 | ds2 | so we would all come out ahead if we can put together an average of 1.5 hammers per person |
| 20:22.34 | SpeedEvil | That's what - 8 billion? |
| 20:22.37 | ds2 | and people can throw away the 0.5 hammer they don't need |
| 20:22.38 | prpplague | hehe |
| 20:22.39 | SpeedEvil | Maybe that's optimistic. |
| 20:23.21 | CosmicPenguin | I have .5 of a hammer - I call it a stick |
| 20:23.31 | prpplague | hehe |
| 20:23.48 | landley | Well, if I get a 3 day weekend I may be able to finally get the documentation I'm doing unblocked and install FWL on the nail I've got, along with tinycc. |
| 20:23.59 | landley | I still need the kernel patches, though. |
| 20:24.07 | ds2 | so in reality, you only need about 4.5K people ordering it to acheive that |
| 20:24.09 | landley | And if you do a write up on getting X working remotely on the hammer, I'm interested. |
| 20:24.22 | prpplague | landley: yea, i'll be off friday, i'll see about getting time to get that to you |
| 20:24.25 | landley | Oh, and and I need to do a write-up on using gdb through the serial console on hammer. :) |
| 20:24.33 | landley | (Yes, it can be done.) |
| 20:25.01 | landley | There's a gdb assistant server thingy you can run on the far end. |
| 20:25.08 | landley | But the assistant thingy wants to be talked to via the network. |
| 20:25.18 | prpplague | landley: yea, you can also use openocd to do gdb as well |
| 20:25.26 | landley | So what you do is you install netcat on the hammer, run netcat via the serial console to talk to loopback. |
| 20:25.44 | landley | Use gdb's pipe functionality, I did this last year and can work out the details again... |
| 20:25.52 | landley | prpplague: that might be easier, of course. :) |
| 20:26.01 | landley | Except then you're debugging the OS, not an application. |
| 20:26.22 | landley | But a write-up on that is cool. There might be on in my bookmarks... |
| 20:26.30 | landley | Been ~2 weeks since I've looked at this. |
| 20:30.35 | *** join/#edev cajaka (n=none@cpe-24-58-193-200.twcny.res.rr.com) |
| 20:32.03 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 20:33.19 | RobotGuy | prpplague: OK, #tincantools is setup. You can make whatever changes you want now. |
| 20:45.32 | RobotGuy | I'll be back later - got to go see if I can get my power chair fixed |
| 20:52.09 | *** join/#edev neuromancer (n=neuroman@91.145.193.247) |
| 21:02.56 | *** join/#edev Madsy (n=madsy@ti500710a342-1583.bb.online.no) |
| 21:04.04 | *** join/#edev ico2_ (n=ico2@modem-953.chimpanzee.dialup.pol.co.uk) |
| 21:12.56 | *** join/#edev dijenerate (n=dijenera@72.51.66.248) |
| 21:22.52 | *** join/#edev neuromancer (n=neuroman@91.145.193.247) |
| 21:39.50 | *** join/#edev JonHylands (n=hylandsj@CPE001c102333bd-CM0014e88f58e0.cpe.net.cable.rogers.com) |
| 21:40.06 | JonHylands | hey prpplague - you there? |
| 21:42.47 | *** join/#edev prpplague^2 (n=dave@mail.americanmicrosystems.com) |
| 21:49.10 | JonHylands | prpplague^2: are you there? |
| 22:07.31 | *** part/#edev JonHylands (n=hylandsj@CPE001c102333bd-CM0014e88f58e0.cpe.net.cable.rogers.com) |
| 22:16.58 | *** join/#edev CosmicPenguin (n=nobody@163.181.251.103) |
| 22:16.58 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 22:29.45 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 22:35.20 | *** join/#edev chouimat (n=dieu@r2351064.cidc.net) |
| 23:05.35 | *** join/#edev ico2_ (n=ico2@modem-953.chimpanzee.dialup.pol.co.uk) |
| 23:07.53 | *** join/#edev gandhijee (n=user@mail.win-ent.com) |
| 23:39.09 | ds2 | and the wheels of bluez turns once again |
| 23:51.50 | *** join/#edev prpplague (n=dave123_@ppp-70-244-160-94.dsl.rcsntx.swbell.net) |