| 00:00.02 | ds2 | I'd think those are acceptable reasons for delays... having a file sit around cuz there is a business logic exception is, IMO, not acceptable |
| 00:00.51 | davidc__ | meh, use it or don't use it.. Doesn't really bother me. All I'm saying is that the SFE people are all honest people, and they do try and get things out as fast as possible |
| 00:01.00 | adyer_ | ds2: you get what you pay for... |
| 00:01.55 | ds2 | davidc__: if it sounds like I don't trust them, that's not what I intend to convey. |
| 00:02.25 | ds2 | adyer_: all I really want is to avoid having to setup echant thanks for some simple boards. |
| 00:04.00 | SpeedEvil | For single sided non-soldermasked boards, at relaxed spacing, it's pretty easy to do at home. |
| 00:04.13 | ds2 | SpeedEvil: messy! |
| 00:04.25 | adyer_ | ds2 lol have lpkf (pcb milling) come and do a 'demo' using your files :-) |
| 00:04.40 | ds2 | heh |
| 00:05.38 | RobotGuy | Why is GCC build still having problems with makeinfo? |
| 00:07.00 | RobotGuy | I think I found it - still had gcj enabled. Turned that off. |
| 00:09.25 | *** join/#edev thedaniel (n=thedanie@38.98.11.186) |
| 00:10.27 | RobotGuy | thedaniel: Greets |
| 00:10.39 | thedaniel | RobotGuy: cheers |
| 00:12.18 | *** join/#edev CosmicPe1guin (n=nobody@163.181.251.103) |
| 00:12.21 | RobotGuy | Now it gets quiet, after I tell you it's been busy all day.. |
| 00:12.36 | thedaniel | haha |
| 00:12.46 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 00:13.08 | thedaniel | i've never done embedded development (unless simple atmega by way of arduino counts) |
| 00:13.21 | RobotGuy | thedaniel: Those are embedded also. |
| 00:13.27 | *** join/#edev chouimat|busy (n=dieu@r2351064.cidc.net) |
| 00:13.48 | thedaniel | i'm kind of interested in nintendo ds homebrew, just ordered my flash kit |
| 00:13.53 | RobotGuy | I also do PICs and dsPICs, but no AVR yet. |
| 00:14.36 | RobotGuy | I wonder if I could use a DS as a console to Hammer, including graphics screens. |
| 00:14.57 | RobotGuy | Not quite a big enough screen to work on though. |
| 00:21.15 | RobotGuy | The Hammer is my first TRUE embedded system. |
| 00:21.24 | RobotGuy | With a real OS and such. |
| 00:21.37 | CosmicPenguin | to the extent that anything that runs full linux can be considered 'embedded" |
| 00:22.03 | RobotGuy | I'm not talking about PCs. |
| 00:22.03 | SpeedEvil | Sure it can. |
| 00:22.04 | MonMotha | eh, I'd say there are certainly applications of full blown Linux which are "embedded" |
| 00:22.19 | MonMotha | anything with application specific hardware is probably a good candidate for that word |
| 00:22.31 | MonMotha | even if it's PC based |
| 00:23.03 | MonMotha | if the hardware was developed specifically for the application, and the software carries out functions limited to the application, then I'd classify it as "embedded" |
| 00:23.52 | ds2 | MonMotha: so a Appliance running on stock PC hardware would qualify under your def? |
| 00:23.58 | RobotGuy | MonMotha: That would not necessarily hold true for many robots though, if they use all off the shelf parts with no custom hardware. |
| 00:24.26 | RobotGuy | Yet, those robots would still be embedded systems. |
| 00:24.31 | MonMotha | ds2: not if the hardware is stock |
| 00:24.38 | MonMotha | I'm saying custom PC based hardware |
| 00:24.53 | ds2 | would a stock MB + a custom PC card qualify then? |
| 00:24.55 | MonMotha | PC104 type stuff would even be a stretch unless it includes application speciic hadware |
| 00:24.56 | ds2 | PCI |
| 00:24.58 | RobotGuy | Embedded systems do not have to have any custom hardware. |
| 00:24.59 | MonMotha | ds2: I'd say it could |
| 00:25.15 | MonMotha | ok, custom is a bad word. Application-specific hardware |
| 00:25.40 | MonMotha | I remember trying to come up with a good definition in #edev and #elinux a long time ago. IIRC, we gave up |
| 00:25.47 | RobotGuy | They don't even have to have any application specific hardware. Servo controllers are not application specific, for instance. |
| 00:26.21 | MonMotha | it's application specific in that PCs are not normally equipped with such hardware, and it is being used to interface with a specific, integral part of the applicaiton |
| 00:26.40 | RobotGuy | MonMotha: What such hardware are they not normally equipped for? |
| 00:26.55 | RobotGuy | Serial Servo controllers are not application specific. |
| 00:27.07 | RobotGuy | Like the Lynxmotion SSC-32, for instance. |
| 00:27.19 | ds2 | so then per this definition, embedded really goes against the grain of using COTS hardware? |
| 00:27.22 | MonMotha | unless you're hooking your servo up to the parallel port, there's some sort of interface hardware |
| 00:27.47 | MonMotha | ds2: I'd say that a system utilizing entirely COTS hardware without anything application specific (even if it's COTS) is not embedded |
| 00:27.56 | RobotGuy | MonMotha: Of course there is, but it is NOT necessarily application specific. |
| 00:28.36 | RobotGuy | You can use an SSC-32 in many different applications. |
| 00:28.43 | MonMotha | RobotGuy: it's application specific in that it's something used only in this or similar applications and is not something a PC is normally equipped with |
| 00:29.13 | RobotGuy | MonMotha: Well, I disagree and will just leave it at that. |
| 00:29.17 | MonMotha | by application specific, I mean hardware which is not normally a part of the PC and which is used for a part of your embedded application that other embedded applications won't necessarily have |
| 00:29.53 | MonMotha | that said, dangling a servo controller of a PC serial port might be a stretch for "embedded" anyway |
| 00:30.22 | RobotGuy | It's the application that is embedded, not necessarily what hardware you use to accomplish the goal. |
| 00:30.32 | MonMotha | it is actually really hard to come up with a good definition of embedded |
| 00:31.02 | ds2 | heh.. tell that to the guys running stock PC hardware w/windows install |
| 00:31.16 | SpeedEvil | 'Can the end user install software on it' |
| 00:31.23 | MonMotha | true, but then you start talking about the software or hardware side and it may not look "embedded" anymore. A digital signage Kiosk running some custom full-screen software on a stock install of WinXP is not embedded in my mind |
| 00:31.34 | MonMotha | SpeedEvil: even that fails sometimes |
| 00:31.40 | ds2 | SpeedEvil: but then a lot of phones fail that test |
| 00:31.44 | MonMotha | what if the device is end-user programmable ala a PC |
| 00:31.50 | MonMotha | and yes, phones/PDAs |
| 00:31.53 | RobotGuy | MonMotha: I think that is indeed embedded - an embedded APPLICATION. |
| 00:32.13 | SpeedEvil | ds2: so? |
| 00:32.30 | MonMotha | RobotGuy: perhaps, but from a software/hardware point of view, the issues that crop up are very different from normal embedded applications |
| 00:32.31 | ds2 | so what would an appliance like a NetApp fileserver be considered if it not embedded? |
| 00:32.34 | SpeedEvil | ds2: a phone that doesn't isn't really embedded. |
| 00:32.37 | SpeedEvil | IMO |
| 00:32.44 | MonMotha | this again falls down to the issue of it being really hard to come up with a good definition |
| 00:32.48 | RobotGuy | MonMotha: How so? |
| 00:32.55 | ds2 | it is fun debating ;) |
| 00:33.30 | MonMotha | RobotGuy: most embedded systems have custom thing that need twiddled or some other sort of interfacing that needs to be done, either software or hardware (or both). That doens't come up in a digital sign kiosk. |
| 00:33.30 | RobotGuy | MonMotha: Embedded covers a much wider range of things than some might want to admit. |
| 00:33.35 | MonMotha | it does |
| 00:33.54 | MonMotha | but I would not classify a digital sign running some app that plays movies as embedded. It's a PC playing movies |
| 00:34.02 | RobotGuy | MonMotha: Every application has custom things that have to be twidded, whether in software, hardware, or both. |
| 00:34.19 | RobotGuy | MonMotha: It's still embedded. :D |
| 00:34.53 | MonMotha | I'm not really sure about that, but whatever |
| 00:35.05 | *** join/#edev Redhatter (n=stuartl@gentoo/developer/redhatter) |
| 00:35.11 | RobotGuy | MonMotha: I just have a wider definition of what embedded is. |
| 00:35.35 | MonMotha | perhaps |
| 00:35.41 | MonMotha | it does seem to vary from person to person |
| 00:35.45 | Mirell | *yawn* |
| 00:35.54 | MonMotha | being a hardware/driver guy, embedded to me means I have work to do :) |
| 00:36.20 | RobotGuy | MonMotha: Same for a more software oriented person. |
| 00:42.16 | *** join/#edev Redhatter (n=stuartl@gentoo/developer/redhatter) |
| 00:48.04 | Mirell | Mmm |
| 00:50.50 | prpplague | ho ho hum |
| 00:51.25 | Mirell | Indeed. |
| 00:51.34 | Mirell | I get paid this Friday. |
| 00:51.39 | Mirell | 'sall I'm looking forward to. |
| 00:53.36 | prpplague | Mirell: bills bills and more bills |
| 00:54.45 | Mirell | Indeed. |
| 00:54.52 | Mirell | Well, right now, I have $40 in the bank account |
| 01:06.26 | *** join/#edev dijenerate (n=dijenera@66.54.118.16) |
| 01:11.34 | prpplague | landley: hehe i see you've found the /bin/sh issue in ubuntu |
| 01:14.30 | Mirell | prpplague: He's ranted about that for months |
| 01:23.29 | prpplague | hehe |
| 01:25.50 | prpplague | ho ho hum |
| 01:28.44 | landley | prpplague: YES! |
| 01:28.46 | landley | Grrrrrrrrr. |
| 01:28.53 | landley | I've actually been ranting about that for most of a year. |
| 01:28.57 | prpplague | hehe |
| 01:29.18 | prpplague | landley: yea it showed up over the weekend with some of the hammer build stuff |
| 01:33.21 | *** part/#edev thedaniel (n=thedanie@38.98.11.186) |
| 01:35.34 | prpplague | landley: i was looking at the expansion port i've got setup on the nail, we could add a rtc chip and small watch battery on the back side of the pcb if it looks like people really want it |
| 01:41.47 | landley | prpplague: never use the /bin/sh symlink. |
| 01:41.49 | landley | Its broken on ubuntu. |
| 01:42.05 | landley | prpplague: I'm not sure it's worth $4. |
| 01:42.51 | landley | Collect potential upgrades for the second round, list that, but I'm guessing the first batch of prototypes is already pending with an existing design? |
| 01:44.09 | SpeedEvil | If there is space on the bottom - a pad suitable for soldering on a 1-wire chip, connected to another GPIO? |
| 01:44.21 | prpplague | landley: yea, but the thing is, a good design always has enough expansion and i/o that mods and such can be added without too much of on issue |
| 01:45.12 | prpplague | SpeedEvil: there is a 2x10 header that has the i/o |
| 01:45.22 | landley | It's more a question of sending a questionaire with the prototypes. |
| 01:45.32 | landley | "Is a battery backed up clock worth another $4"... |
| 01:45.36 | prpplague | landley: yep |
| 01:45.46 | landley | Today I'm too blah to really have a coherent opinion. :( |
| 01:45.55 | SpeedEvil | 'would you prefer a $5 board that can be connected like this' |
| 01:46.15 | SpeedEvil | Though it probably couldn't be $5 |
| 01:47.07 | prpplague | SpeedEvil: no it would be more like $20 |
| 01:47.10 | prpplague | SpeedEvil: http://www.elinux.org/Image:Nail_profile.jpg |
| 01:47.16 | prpplague | SpeedEvil: you can see the option port |
| 01:47.46 | prpplague | SpeedEvil: the option port is what is use to place the leds and push buttons inside the case |
| 01:49.30 | landley | Would it fit inside the case? |
| 01:50.17 | prpplague | yea the case is design for that |
| 01:55.38 | prpplague | landley: surface mounted leds and buttons don't work well from the outside of a case without alot of plastic parts |
| 01:56.49 | *** join/#edev chouimat|away (n=dieu@kde/developer/chouinard) |
| 01:56.53 | prpplague | landley: so we've add this header to serve two purposes |
| 01:56.56 | landley | prpplague: so you could offer an expansion board thingy providing a battery backed up realtime clock. |
| 01:57.05 | prpplague | landley: yep |
| 01:57.18 | prpplague | landley: or a wide range of other items |
| 02:00.46 | prpplague | landley: add something like this - http://www.sparkfun.com/commerce/product_info.php?products_id=308 |
| 02:01.07 | prpplague | landley: and then you can swing you nail board like a light saber and make noises on the buzzer, hehe |
| 02:01.15 | SpeedEvil | That's annoying and silly. |
| 02:01.23 | SpeedEvil | Better to use a digital out one |
| 02:01.24 | landley | prpplague: cord isn't long enough. |
| 02:01.40 | SpeedEvil | I mean annoying to interface. |
| 02:02.14 | prpplague | landley: hehe, you can always get a longer usb cable, but that's just an example of something you could add |
| 02:02.26 | SpeedEvil | http://www.sparkfun.com/commerce/product_info.php?products_id=753 is easier - i2c/spi |
| 02:03.21 | prpplague | landley: even add something like this - http://www.sparkfun.com/commerce/product_info.php?products_id=8291 |
| 02:04.59 | landley | prpplague: make a list. :) |
| 02:05.05 | landley | However, right now I wouldn't focus on that. |
| 02:05.19 | SpeedEvil | The sparkfun website is already mentioned. |
| 02:05.20 | landley | I brought up the rtc because the SOC is documented as supporting it. The hammer, however, does not. |
| 02:05.30 | SpeedEvil | At least half of the stuff is worth connecting. |
| 02:05.46 | landley | SpeedEvil: where is it mentioned, and who is really going to go through and dig to find out what's there based on a passing mention? |
| 02:05.55 | SpeedEvil | true |
| 02:06.12 | prpplague | landley: right, with just 40 pins you have to select what most people are going to use |
| 02:06.50 | prpplague | landley: actually i've been working like crazy on finishing the specs and layout for the nail |
| 02:07.05 | prpplague | landley: i've got pretty much everything done except the case |
| 02:08.16 | landley | Cool. |
| 02:08.56 | landley | Did you send the broken up kerne patches my way? |
| 02:09.11 | prpplague | landley: most of the folks that hang in #edev are pretty close to the hardware dev, thats why having your input on the software level is important , it really adds some strength |
| 02:09.29 | prpplague | landley: i got the patches done, but i was swampped today |
| 02:09.39 | prpplague | landley: i'll get them out to you in the morning |
| 02:10.05 | prpplague | landley: we have a new product in the pipe and the EE hasn't left me much time to debug stuff |
| 02:11.43 | prpplague | landley: i'm basically doing two jobs |
| 02:11.59 | prpplague | landley: main job for aml and then all the testing and dev for tincantools |
| 02:12.52 | landley | no rush. |
| 02:13.12 | landley | Try grad school, consulting, and being a convention chair sometime. :) |
| 02:13.37 | landley | Or back in the 90's: day job programming, teaching night courses at ACC, and writing Motley Fool columns. |
| 02:13.48 | landley | Ah, sleep deprivation. My old friend... |
| 02:13.55 | prpplague | hehe |
| 02:14.08 | prpplague | landley: yea, i did alot of overtime when i work for abcsinc |
| 02:14.11 | MonMotha | I promised myself a minimum of 6 hours of sleep per night when I graduated from college |
| 02:14.29 | MonMotha | I'm still recovering from 4 years of ~3 per night |
| 02:15.01 | prpplague | landley: one time i had 110 billable hours in a 7 day period |
| 02:15.35 | landley | I generally try not to top 90. |
| 02:15.41 | landley | My efficiency goes _way_ down after that. |
| 02:15.56 | MonMotha | shite... |
| 02:16.04 | MonMotha | 60 seems to be about my limit |
| 02:16.10 | landley | (Yes, I know this from personal experience.) |
| 02:16.14 | landley | At three different companies... |
| 02:16.19 | prpplague | hehe indeed |
| 02:16.26 | MonMotha | I can still do stuff after that, but it's about 2 hours to do 1 hour's worth of work |
| 02:16.59 | landley | It gets to the point where going away and sleeping results in _faster_ completion, net. |
| 02:17.21 | landley | And working through three consecutive weekends has a similar effect. |
| 02:18.10 | prpplague | landley: my service area was the caribbean, back in 2000, a hurricane came through there, so i was totally busy for almost 45 days |
| 02:18.26 | prpplague | landley: indeed |
| 02:20.27 | Mirell | Mmm |
| 02:20.49 | prpplague | Nnn |
| 02:20.54 | landley | Ok, that's a good excuse. |
| 02:20.59 | landley | Mirell: food? |
| 02:22.17 | Mirell | landley: Where? |
| 02:22.33 | landley | Well, I'm thinking of going to chick-fil-a. |
| 02:22.40 | Mirell | Ahh |
| 02:22.43 | landley | I still have that calendar to give you, if you wanted to meet me there. |
| 02:22.48 | landley | (For me, it's right down the street...) |
| 02:22.50 | Mirell | That could work. |
| 02:23.06 | Mirell | landley: There's a Chick-Fil-A at Parmer and Mopac, so there's on close to me too. |
| 02:23.15 | prpplague | landley: http://www.sparkfun.com/commerce/product_info.php?products_id=8474 |
| 02:23.21 | prpplague | landley: can add bluetooth |
| 02:23.24 | landley | I'm kind of sick today. Really don't feel up to a long drive. |
| 02:23.32 | Mirell | landley: The one near you works fine. |
| 02:23.33 | landley | I've been meaning to go out for an hour now,b ut haven't left the couch. |
| 02:23.39 | landley | (Fade's sick too. Dunno why...) |
| 02:23.42 | Mirell | landley: Just saying that for work stuff. |
| 02:23.46 | Mirell | landley: Cedar fever? |
| 02:23.53 | landley | No idea. |
| 02:24.04 | landley | General blah in my case, highly upset stomach in hers. |
| 02:24.06 | Mirell | A lot of my coworkers are suffering from it. |
| 02:24.15 | prpplague | a ton of people at work are sick |
| 02:24.20 | landley | How exactly _does_ one ork a cow? |
| 02:24.26 | landley | Scott Adams didn't really go into details... |
| 02:25.44 | landley | Mirell: meet you at the chick-fil-a near me in 20 minutes? |
| 02:25.50 | Mirell | landley: A'ight |
| 02:25.54 | landley | Woot. |
| 02:26.22 | prpplague | hehe |
| 02:28.11 | *** join/#edev RobotGuy (n=RobotGuy@dsl093-038-120.pdx1.dsl.speakeasy.net) |
| 02:32.11 | *** join/#edev dijenerate (n=dijenera@72.51.75.84) |
| 02:33.41 | prpplague | dijenerate: hello wanna |
| 02:35.07 | prpplague | dijenerate: we don't tolerate no foolishness in #edev, ya heaaar? |
| 02:36.19 | RobotGuy | prpplague: I modified my autobuild script to use some environment variables to allow setting it up for any buildroot. |
| 02:36.37 | prpplague | cool |
| 02:36.53 | RobotGuy | I will put it on the Wiki eventually. |
| 02:38.57 | RobotGuy | prpplague: I am thinking of doing this: mount thumb drive on my PC at /src/hammer/buildroot/project_build_arm/walter/root/usr and then building the software I want on my Hammer. |
| 02:39.23 | prpplague | that works |
| 02:39.35 | RobotGuy | prpplague: The only thing that concerns me is the initramfs - that is still going to want to build stuff in /usr. I want to mount my thumb drive on /usr on the Hammer. |
| 02:40.59 | prpplague | RobotGuy: just a reminder, initramfs and initrd are two different things |
| 02:41.08 | RobotGuy | prpplague: I understand that. |
| 02:41.22 | prpplague | <RobotGuy> prpplague: The only thing that concerns me is the initramfs - that is still going to want to build stuff in /usr. I want to mount my thumb drive on /usr on the Hammer. |
| 02:41.28 | prpplague | you said initramfs |
| 02:41.30 | RobotGuy | prpplague: They do seem somewhat similar though and easily confused. |
| 02:41.40 | RobotGuy | prpplague: I meant initrd |
| 02:42.02 | prpplague | RobotGuy: i know, just didn't want you to develop a bad habit , hehe |
| 02:42.16 | RobotGuy | prpplague: I try to use the correct terms. |
| 02:42.42 | prpplague | RobotGuy: i know, just teasing |
| 02:42.48 | RobotGuy | prpplague: :D |
| 02:43.25 | prpplague | dijenerate: hey did you know errol barrow served in the RAF during WW II ? |
| 02:43.34 | RobotGuy | prpplague: So what happens when I want to load an initrd? Does it go to a special place in flash at the start where I defined the length in .size? |
| 02:44.07 | prpplague | RobotGuy: yes and no |
| 02:44.22 | RobotGuy | prpplague: I love a decisive fellow |
| 02:44.39 | prpplague | RobotGuy: the .size you are refering to is the total size of the bootloader, bootloader settings, kernel and initrd image |
| 02:45.19 | RobotGuy | prpplague: My real question is about the /usr stuff I build for the initrd and what I will build for software that goes on my thumb drive as /usr. Do they get mixed up at all? |
| 02:45.26 | prpplague | most embedded developers just group that area as a read only space in the kernel to prevent it from being damaged from inside userland |
| 02:45.48 | RobotGuy | The .size area? |
| 02:46.30 | prpplague | yea, if you look at that struct, its basically a parition table |
| 02:46.39 | prpplague | you have two areas listed |
| 02:46.42 | RobotGuy | OK, I thought it looked familiar. |
| 02:46.43 | prpplague | one is the system area |
| 02:46.51 | prpplague | the other is a the jffs2 area |
| 02:47.00 | RobotGuy | OK |
| 02:48.08 | RobotGuy | I am using ext2 on my thumb drive. Should I use jffs2 instead? |
| 02:48.20 | MonMotha | in general, on |
| 02:48.23 | prpplague | no ext2 is fine on your thumbdrive |
| 02:48.27 | RobotGuy | OK |
| 02:48.31 | MonMotha | in fact, you usually can't without going to efforts that do no good |
| 02:48.46 | RobotGuy | MonMotha: Can't what? |
| 02:48.59 | MonMotha | can't use jffs2 |
| 02:49.01 | prpplague | RobotGuy: the thumbdrive's controller handles all the functions that jffs2 does |
| 02:49.12 | RobotGuy | prpplague: OK, good enough. |
| 02:49.16 | prpplague | RobotGuy: you'd be doing everything twice |
| 02:49.18 | MonMotha | USB drives show up as regular block devices, while jffs2 has to be able to talk directly to the flash (mtd block device) |
| 02:50.15 | *** join/#edev chouimat|busy (n=dieu@r2351064.cidc.net) |
| 02:52.20 | RobotGuy | prpplague: The initrd is a completely different /usr then, right? |
| 02:52.41 | prpplague | RobotGuy: its what ever you put on it |
| 02:53.13 | RobotGuy | prpplague: That is not what I meant. It's /usr is different from what I put on my thumb drive to mount at /usr |
| 02:53.52 | prpplague | RobotGuy: based on what you've described to me , yea |
| 02:56.28 | RobotGuy | Since I am going to have my thumb drive mounted at <project>/root/usr I can just build everything that way because none of it will be in the initrd, right? |
| 02:57.20 | RobotGuy | Or am I way off base again? |
| 02:59.40 | RobotGuy | Isn't the initrd build going to want to put everything into the rootfs image? |
| 03:02.52 | MonMotha | we have a Marvell rep trying to sell us on their stuff. I'm trying to determine if we're being sucked into a whole or if this is the real deal |
| 03:02.54 | MonMotha | *hole |
| 03:03.05 | MonMotha | they have some neato stuff, certainly |
| 03:03.35 | MonMotha | sadly, we've already been told no on wifi driver info |
| 03:03.42 | MonMotha | even under NDA, it's not available |
| 03:03.46 | MonMotha | (!) |
| 03:03.59 | SpeedEvil | How many'd you be integrating? |
| 03:04.10 | MonMotha | not enough to get the specs, it seems |
| 03:04.11 | RobotGuy | MonMotha: Don't they want their chips USED by lots of companies? |
| 03:04.14 | MonMotha | (and not many, anyway) |
| 03:04.28 | MonMotha | RobotGuy: you can get a binary blob from one of their "strategic partners" |
| 03:04.51 | MonMotha | but that locks you to a specific kernel |
| 03:05.07 | RobotGuy | MonMotha: Go get somebody else's chips then. |
| 03:05.16 | MonMotha | well, we're looking |
| 03:05.22 | MonMotha | but their SoCs are very attractive otherwise |
| 03:05.35 | MonMotha | especially for multimedia-centric devices |
| 03:05.50 | RobotGuy | What about Freescale? The i.MX chips |
| 03:06.10 | MonMotha | I don't currently have anybody repping freescale for me |
| 03:06.15 | MonMotha | *to me |
| 03:06.26 | MonMotha | I'm sure I could find somebody |
| 03:06.36 | MonMotha | I do have an Atheros rep |
| 03:06.40 | MonMotha | they're looking at things for s |
| 03:06.42 | MonMotha | us |
| 03:06.50 | RobotGuy | I would love to have a board with an Freescale i.MX31 or TI DaVinci |
| 03:07.00 | MonMotha | DaVinci is licensing hell |
| 03:07.06 | MonMotha | and you need serious volume just to get started |
| 03:07.14 | MonMotha | it's certainly neat, but kinda messy |
| 03:07.30 | MonMotha | Marvell's media stuff is much simpler from a licensing point of view. Just pay MPEG-LA |
| 03:07.33 | RobotGuy | MonMotha: That sucks. I got samples from Freescale of the i.MX21, i.MX27, and i.MX31. |
| 03:07.53 | RobotGuy | They are the dreaded BGAs though. :( |
| 03:08.12 | MonMotha | eh, BGAs aren't too much of a problem |
| 03:08.17 | MonMotha | we don't hand-assemble very much stuff anyway |
| 03:08.26 | MonMotha | just not worth paying an engineer to do assembly |
| 03:08.40 | MonMotha | and we aren't big enough to have dedicated assembly techs |
| 03:10.03 | prpplague | and the davinci's are min 50k order |
| 03:10.13 | MonMotha | like I said, serious volume |
| 03:10.29 | MonMotha | 10k can be pulled on many projects, but 50k for a first order is kinda high volume |
| 03:10.44 | MonMotha | our current applications are even lower, 1k-5k |
| 03:11.03 | MonMotha | I'm surprised Marvell is willing to even entertain that |
| 03:11.11 | SpeedEvil | Hoping to get to 10-100. |
| 03:11.32 | SpeedEvil | Probably hope you'll grow. |
| 03:11.38 | prpplague | MonMotha: yea we are usually in the 5-10k range |
| 03:11.57 | prpplague | MonMotha: 50k is just way too many |
| 03:12.01 | MonMotha | prpplague: and let me guess, if there was a SPI Wi-Fi chipset that was supported by Linux 2.6, it would get designed in immediately |
| 03:12.28 | prpplague | MonMotha: there are 5 chipsets that support spi |
| 03:12.51 | MonMotha | prpplague: supported by Linux and available (reliably) in small quantity? |
| 03:13.09 | MonMotha | small meaning ~1k |
| 03:13.17 | MonMotha | plus samples for proto |
| 03:14.08 | prpplague | MonMotha: only the libertas chipset from marvell is partially supported under linux |
| 03:14.25 | prpplague | MonMotha: it should be fully supported in another few weeks |
| 03:14.33 | MonMotha | see, that's the kicker |
| 03:14.40 | RobotGuy | prpplague: Is it the bootloader that sets the Hammer up and initializes the hardware devices? |
| 03:14.40 | prpplague | MonMotha: the zipit v2 uses the libertas in spi mode |
| 03:14.58 | prpplague | MonMotha: if you are willing you can always buy the reference driver from marvell |
| 03:14.59 | MonMotha | hum, I see |
| 03:15.07 | prpplague | MonMotha: it'll cost you a few k |
| 03:15.17 | MonMotha | prpplague: I've asked. They won't sell it at any price NDA or no with source. |
| 03:15.20 | prpplague | RobotGuy: only partially |
| 03:15.23 | MonMotha | binary blob only |
| 03:15.43 | prpplague | MonMotha: uh are you sure? |
| 03:15.48 | MonMotha | I asked |
| 03:15.49 | prpplague | MonMotha: we purchased it |
| 03:15.57 | MonMotha | they said they'd happily sell me a binary blob for a few thousand |
| 03:16.06 | prpplague | MonMotha: thats bogus |
| 03:16.13 | prpplague | MonMotha: you were speaking to marvell? |
| 03:16.13 | MonMotha | perhaps we just don't have the volume? |
| 03:16.22 | MonMotha | no, a sales rep, who may have been misinformed |
| 03:16.29 | prpplague | MonMotha: indeed |
| 03:16.45 | prpplague | MonMotha: i would assume he was since, i have access to the source |
| 03:18.34 | MonMotha | she, actually, but I guess I can check |
| 03:18.46 | MonMotha | what volume are you guys doing? |
| 03:19.41 | prpplague | we committed to 5k per year |
| 03:19.58 | MonMotha | hum |
| 03:20.05 | prpplague | we'll probably do 10k |
| 03:20.14 | prpplague | depending on how well they work in the field |
| 03:20.35 | MonMotha | we were told that Marvell would not deal direct on wifi stuff in low volume (which went undefined, but I felt <10k), but only through a partner |
| 03:20.41 | MonMotha | and that said partners didn't offer source on their drivers |
| 03:27.35 | prpplague | interesting |
| 03:27.35 | prpplague | RobotGuy: different product |
| 03:28.28 | MonMotha | prpplague: perhaps our rep is just unaware of the libertas stuff |
| 03:28.33 | MonMotha | she was mostly talking about their SoC devices |
| 03:28.52 | prpplague | MonMotha: thats what i would suspect |
| 03:29.08 | RobotGuy | prpplague: Sneaks a Sledge Hammer and takes it to a field to check how it works. ;) |
| 03:29.27 | RobotGuy | Durn, blew that one bigtime! |
| 03:33.08 | prpplague | hehe |
| 03:33.30 | prpplague | RobotGuy: no fun there, the sledge hammer was design and built over 20 months ago |
| 03:33.53 | RobotGuy | prpplague: Drats, well build me something brand NEW then, dagnambit! :P |
| 03:34.04 | MonMotha | heh, I was playing with some 802.15.4 today |
| 03:34.12 | prpplague | RobotGuy: hehe thats kinda the idea of tincantools, we use whats already built |
| 03:34.23 | MonMotha | I managed to get like 500 feet from our office outside before the link dropped |
| 03:34.29 | MonMotha | was pretty impressive |
| 03:34.37 | RobotGuy | prpplague: I want a shiny NEW tin can.. :D |
| 03:34.51 | prpplague | hehe |
| 03:34.57 | prpplague | MonMotha: nice |
| 03:41.23 | landley | And back. |
| 03:41.27 | landley | Feeling much suckage. |
| 03:50.52 | SpeedEvil | MonMotha: zigbee? |
| 03:51.12 | SpeedEvil | MonMotha: I'm interested in the insanely cheap zigbee-lite part from freescale. |
| 03:51.25 | MonMotha | SpeedEvil: we're using the Atmel part |
| 03:51.26 | SpeedEvil | $3 and has an open-source wifi stack. |
| 03:51.27 | SpeedEvil | Err |
| 03:51.36 | SpeedEvil | wireless |
| 03:51.43 | MonMotha | best specs I've seen and testing by other people as confirmed the specs |
| 03:51.46 | SpeedEvil | And lots of io. |
| 03:51.49 | MonMotha | that said, pain in the neck to talk to |
| 03:51.56 | MonMotha | it doesn't actually speak SPI, |
| 03:52.04 | MonMotha | despite the docs saying it does. It speaks something SPI-eqsue |
| 03:52.06 | SpeedEvil | I'm surprised zigbee hasn't broken out into the actual market. |
| 03:52.13 | SpeedEvil | Or has it, and it's not called that. |
| 03:52.20 | MonMotha | well, 802.15.4 is out there |
| 03:52.26 | MonMotha | "Zigbee" isn't so much |
| 03:52.30 | SpeedEvil | Searching ebay for zigbee typically reveals no hits on hardwaqre |
| 03:52.39 | SpeedEvil | Are they the same thing? |
| 03:52.49 | MonMotha | 802.15.4 is the physical/MAC side of zigbee |
| 03:52.51 | SpeedEvil | Unsure about the .4 |
| 03:52.52 | SpeedEvil | ah |
| 03:52.57 | MonMotha | zigbee adds mesh networking, routing, etc. |
| 03:53.06 | MonMotha | as far as actual PAN stuff, Bluetooth is much more common |
| 03:53.13 | RobotGuy | I like the many to many aspect of ZigBee. |
| 03:53.19 | SpeedEvil | And annoyingly more expensive. |
| 03:53.24 | SpeedEvil | Not to mention power consuming. |
| 03:53.51 | Mirell | Hmm... |
| 03:54.02 | MonMotha | RobotGuy: USB<=>IEEE 1394 as BlueTooth<=>IEEE 802.15.4, it seems |
| 03:54.06 | SpeedEvil | Talking of stuff available in ones that is. |
| 03:54.22 | MonMotha | you can sample 802.15.4 MAC/PHYs pretty cheap/free |
| 03:54.32 | MonMotha | but it's not used much in consumer products |
| 03:54.36 | RobotGuy | MonMotha: Really? |
| 03:54.51 | MonMotha | it has found its home mostly in control networks |
| 03:54.55 | SpeedEvil | RobotGuy: freescale will even sell you single parts or tens, or sample singles. |
| 03:55.04 | RobotGuy | So 1394 and 802.15.4 would be the better ones of the pairs? |
| 03:55.16 | MonMotha | RobotGuy: I wouldn't say better, but more capable |
| 03:55.18 | RobotGuy | SpeedEvil: I have samples of three Freescale i.MX processors. |
| 03:55.32 | MonMotha | and also probably a more solid design from a theoretical point of view |
| 03:55.35 | RobotGuy | MonMotha: More capable works. |
| 03:55.43 | MonMotha | USB and Bluetooth are both kinda hacky and rickety, but they do work |
| 03:55.56 | SpeedEvil | And are oh-so-widely depoloyed |
| 03:56.09 | SpeedEvil | Quantity has its own quality. |
| 03:56.09 | MonMotha | IEEE 1394 and 802.15.4 are both more complicated, but also more complete and clean |
| 03:56.16 | MonMotha | also tend to be a bit more expensive |
| 03:56.21 | MonMotha | see complexity |
| 03:57.10 | RobotGuy | I want my robots to have fast wireless. I'm planning to put W.A.L.T.E.R. right on the internet. |
| 03:57.21 | MonMotha | then I'd say wifi is the best optio |
| 03:57.25 | MonMotha | 802.15.4 is not fast |
| 03:57.34 | MonMotha | 802.11 is |
| 03:57.42 | RobotGuy | MonMotha: WiFi G is what I am thinking of. |
| 03:57.50 | MonMotha | 802.15.4 tends to work in harsher environments and over longer distances, though |
| 03:58.11 | MonMotha | and can't be hacked by your average script kiddie with a stock laptop :) |
| 03:58.23 | RobotGuy | MonMotha: Speaking of internet, I think I have the Gadget Ethernet setup now. |
| 03:58.42 | SpeedEvil | I'm planning on 802.11a for my UAV |
| 03:58.47 | SpeedEvil | for hte high bandwidth side. |
| 03:58.50 | SpeedEvil | +80cm dish |
| 03:58.56 | MonMotha | that should work |
| 03:59.04 | SpeedEvil | It gets a 400Mhz radio for low-speed |
| 03:59.11 | MonMotha | you could also run 802.15.4 on the 900MHz band, but gear for that is hard to find |
| 03:59.14 | SpeedEvil | And redundancy. |
| 03:59.29 | MonMotha | ah, see, if I'm going to ~400MHz, I'll crank the power as high as I need to :) |
| 03:59.30 | RobotGuy | I'll be happy enough with WiFi G and Bluetooth on W.A.L.T.E.R. |
| 03:59.37 | SpeedEvil | I've contemplated bluetooth too as backup. |
| 03:59.47 | SpeedEvil | But maybe three strings is a bit much. |
| 03:59.50 | MonMotha | I could theoretically do that with wifi, too, but it gets a bit more expensive since 2.4GHz amps aren't cheap |
| 04:00.02 | MonMotha | 440MHz also propagates better |
| 04:00.10 | MonMotha | at least with the kind of setups you have in those scenarios |
| 04:00.16 | SpeedEvil | In my primary design-case, not really. |
| 04:00.17 | MonMotha | but you aren't getting 11Mbps down there |
| 04:00.26 | SpeedEvil | It's fly straight up, fly straight down. |
| 04:00.34 | MonMotha | hum, so you have line of sight |
| 04:00.41 | MonMotha | yeah, then doesn't matter much |
| 04:00.47 | SpeedEvil | Unless something's already gone badly wrong. |
| 04:00.51 | MonMotha | 10GHz might actually be usable; very low noise |
| 04:00.53 | RobotGuy | I am also planning to get my Amateur Radio License this year. |
| 04:00.55 | *** join/#edev redblue (i=star@ppp176.109-253-207.mtl.mt.videotron.ca) |
| 04:01.11 | SpeedEvil | 5Ghz has however already constructed cheap hardware |
| 04:01.14 | MonMotha | true |
| 04:01.21 | SpeedEvil | not having to build special radios ++ |
| 04:01.34 | MonMotha | there is an amateur band near 5.8GHz, but I don't remember if it overlaps 802.11a |
| 04:01.45 | MonMotha | I know the amateur 2.4GHz band does overlap 802.11(b/g) |
| 04:01.55 | RobotGuy | MonMotha: I don't know either. |
| 04:01.56 | SpeedEvil | My primary motivation is that I may be operating where there is bad wifi contamination. |
| 04:02.04 | SpeedEvil | byt mostly that's only 2.4 |
| 04:02.08 | MonMotha | yes |
| 04:02.14 | MonMotha | you can also get around that by not using wifi |
| 04:02.23 | SpeedEvil | Of course. |
| 04:02.36 | SpeedEvil | But this is very mych a v0 project. |
| 04:02.37 | MonMotha | if you use something that spreads differently (especially orthogonally), then you get a lot of processing gain over the interference |
| 04:02.51 | SpeedEvil | Too many sensors, too much CPU |
| 04:03.08 | SpeedEvil | I'm figuring out what I can cut for V1, and keep reliable control. |
| 04:03.20 | MonMotha | I really wish I could muck around with some wifi chipsets more |
| 04:03.27 | MonMotha | being able to set the spreading codes would be awesome |
| 04:03.40 | SpeedEvil | Then V2 - if it happens - gets fully custom design with its own camera design, and halves the takeoff weight. |
| 04:04.05 | SpeedEvil | yeah. Hackable software-radio'd be nice. |
| 04:04.31 | MonMotha | there is the gnuradio stuff |
| 04:04.44 | MonMotha | but being able to just set basic stuff on COTS wifi gear would be nice |
| 04:04.45 | SpeedEvil | Indeed - but FPGAs suck for this. |
| 04:04.57 | SpeedEvil | Too many pins, too much power |
| 04:04.58 | MonMotha | of course, they really don't want you doing that since it's illegal unless you have some sort of license |
| 04:05.08 | MonMotha | eh, the too many pins isn't a big deal |
| 04:05.16 | SpeedEvil | Depends :) |
| 04:05.19 | MonMotha | and you can get some fairly dense FPGAs in small packages |
| 04:05.25 | MonMotha | the power can be a concern, though |
| 04:05.32 | MonMotha | but if you're not running portable, who cares |
| 04:05.43 | MonMotha | I don't care about pin count so much as package |
| 04:05.52 | MonMotha | and most are still available in 1mm BGA if not 0.5mm QFP |
| 04:06.06 | MonMotha | 1mm BGA at least leaves me on standard PCB design rules. Have to pay $$$ for assembly, but whatever |
| 04:06.07 | SpeedEvil | I'm still at the 'bake in oven' stage, if I can't make do with single layer boards. |
| 04:06.10 | RobotGuy | I'm building software onto my thumb drive now. |
| 04:06.39 | MonMotha | smaller BGAs without center cutouts cause you to need special design rules (blind/buried vias, microvias, and/or <5mil trace/space) |
| 04:06.47 | MonMotha | that sucks and costs lots of $$$ |
| 04:07.00 | MonMotha | you can get Spartan 3Es in QFP |
| 04:07.21 | SpeedEvil | For many designs, something like microchips latest DSpics would be interesting. |
| 04:07.23 | RobotGuy | MonMotha: The Freescale i.MX31 is a 400+ BGA. I wonder what its other specs are for assembly. |
| 04:07.38 | SpeedEvil | They have peripherals all connected to one big crossbar. (apart from the A/Ds) |
| 04:08.05 | RobotGuy | MonMotha: I would love to have Hammer connected to an FPGA, and I already have the Linux tools for Xilinx installed and working. |
| 04:08.46 | MonMotha | RobotGuy: you can get Spartan3Es in TQFP 144 |
| 04:08.56 | MonMotha | they are hand solderable, though it can take effort |
| 04:09.12 | RobotGuy | MonMotha: I know. :D |
| 04:09.32 | MonMotha | heck, the 250k gate version is available in VQ100 |
| 04:09.35 | RobotGuy | I have a Weller WES51 soldering station and have 0.4mm and 0.2mm tips on order. |
| 04:09.45 | SpeedEvil | 0.2mm tips isn't the point |
| 04:09.50 | MonMotha | you know, I have the same iron (just the digital temp controled version) |
| 04:09.51 | SpeedEvil | you tack each corner |
| 04:09.55 | MonMotha | I use the stock tip for everything |
| 04:10.02 | MonMotha | take corners, drag the rest, clean up with solderwick |
| 04:10.03 | SpeedEvil | then run a bead along the side |
| 04:10.06 | SpeedEvil | then wick it off |
| 04:10.11 | MonMotha | I can do 0.5mm QFPs given enough time |
| 04:10.13 | SpeedEvil | flux helps lots |
| 04:10.17 | MonMotha | and yes, it does |
| 04:10.23 | MonMotha | I flux the bejeezus out of the pads |
| 04:10.26 | RobotGuy | I have no clean flux and solder wick. |
| 04:10.35 | MonMotha | probably to the point of causing longevity problems since my flux is not no-clean |
| 04:10.36 | RobotGuy | Actually it is just less clean |
| 04:10.57 | MonMotha | I should point out that a soldermask is absolutely essential for that process |
| 04:11.09 | SpeedEvil | It's not |
| 04:11.11 | MonMotha | putting larger pads on the corners can help, too, at the expense of removing some room in the layout |
| 04:11.16 | SpeedEvil | It's practically essential |
| 04:11.26 | SpeedEvil | I've done it without, and it's not especially fun. |
| 04:11.30 | MonMotha | on larger QFPs, maybe. I've never managed a 0.5mm without a soldermask |
| 04:11.39 | SpeedEvil | oh - 0.5mm would be scary |
| 04:11.52 | MonMotha | I end up tearing traces off the board during the cleanup stage before I get it cleaned up |
| 04:11.54 | MonMotha | yeah |
| 04:12.02 | MonMotha | 0.8mm would be doable, and MAYBE 0.65 |
| 04:12.12 | MonMotha | 1mm is easy, of course |
| 04:12.25 | MonMotha | hell, 1mm is almost too big to drag |
| 04:13.02 | MonMotha | I should point out that I've had to give up and cut chips off the board to start over when using 0.5mm ones sometimes |
| 04:13.45 | RobotGuy | MonMotha: I'm just going to use Schmart Boards to put some chips on. |
| 04:16.39 | MonMotha | I do wish there was a way to get REALLY cheap PCBs |
| 04:16.46 | RobotGuy | I'm trying to get my courage up to operate on my Hammer Carrier Board. |
| 04:16.53 | SpeedEvil | Well - there is for single layer :) |
| 04:16.56 | MonMotha | the crazy spec ones (6+ layer, 6/6 trace/space, tiny vias, etc.) |
| 04:17.04 | SpeedEvil | ah. |
| 04:17.07 | MonMotha | I'll take "undefined" for the lead time |
| 04:17.14 | SpeedEvil | That is challengig to do at home. |
| 04:17.15 | MonMotha | I don't care if it takes 3 months |
| 04:17.31 | MonMotha | but I don't want to have to buy a full lot for $2000 |
| 04:17.37 | SpeedEvil | I haven't thought about pricing on that. |
| 04:17.44 | MonMotha | even 4 layers isn't cheap |
| 04:17.46 | SpeedEvil | I suspect it'd be scary even for tiny boards. |
| 04:17.58 | MonMotha | cheapest way I've found is the $66/ea deal from 4PCB, and that's minimum 3 pcs |
| 04:18.12 | MonMotha | even for 2 layer, it's tough to beat their $33/ea if you panelize |
| 04:18.33 | SpeedEvil | I've pondered something rather silly. A jumbo wirebonder. Able to solder 0.1-0.4mm wire betwen ICs on a ground-plane. |
| 04:19.16 | *** join/#edev huaynac (n=huaynac@61.6.65.198) |
| 04:20.46 | SpeedEvil | The horribly hard part is computing 'trace' inductances and cross-coupling |
| 04:24.06 | MonMotha | at speeds where that matters, I don't think it would be worth trying that kind of thing |
| 04:24.29 | SpeedEvil | Don't know. |
| 04:25.09 | SpeedEvil | In principle, if it could be made to work, It could be a hell of a lot faster prototyping. |
| 04:25.23 | SpeedEvil | In practice, I don't know what the limits'd be |
| 04:25.24 | MonMotha | there are already methods by which to literally print PCBs |
| 04:25.27 | SpeedEvil | yes |
| 04:25.32 | MonMotha | similar to SLA prototyping mechanical things |
| 04:25.38 | MonMotha | multi-layers can be done |
| 04:25.40 | SpeedEvil | not typically in ones, already connected to chips though |
| 04:25.43 | MonMotha | it's just expensive |
| 04:25.47 | SpeedEvil | in a desktop unit |
| 04:26.14 | MonMotha | if you just want 2 layer boards with no frills (like soldermasks or silkscreens), you can get like 20 boards made overnight for about $40 |
| 04:26.17 | MonMotha | it's really not that bad |
| 04:27.55 | ds2 | MonMotha: where? |
| 04:28.05 | MonMotha | ds2: 4pcb "barebones" deal |
| 04:28.25 | ds2 | MonMonth: step/repeat allowed? |
| 04:28.28 | MonMotha | for a pretty small board (like 1.5x1.5"), they end up being like $2/ea if you get several |
| 04:28.33 | MonMotha | no |
| 04:28.54 | MonMotha | but the price is volume based, so there's no reason to step/repeat it yourself |
| 04:29.02 | ds2 | Oh I see |
| 04:29.06 | ds2 | the size is not fix |
| 04:29.13 | ds2 | (waiting for page to load/looking for it) |
| 04:29.15 | MonMotha | there's a substantial setup fee rolled into the price, but after that you get hte boards CHEAP |
| 04:29.51 | MonMotha | if you want full spec 2 layers, the $33/ea deal lets you do step/repeat or multi-project panels for an extra $50 (for the whole order) |
| 04:30.00 | MonMotha | 70 sq in |
| 04:30.15 | SpeedEvil | MonMotha: yeah - the only way this thing'd have a point would be if it could also skip the assembly step. |
| 04:30.41 | MonMotha | assembly isn't that expensive, really |
| 04:30.50 | ds2 | okay, that company is completely blacklisted |
| 04:30.56 | MonMotha | ds2? |
| 04:31.07 | ds2 | peice of crap insist on me turning on everything and the kitchen sink to even get a description of the price |
| 04:31.12 | SpeedEvil | In ones, if you don't have contacts, it's moderately so. |
| 04:31.12 | ds2 | they can go eat dirt |
| 04:31.26 | MonMotha | oh yeah, their website sucks ass |
| 04:31.32 | MonMotha | if you call up a rep, they will quote |
| 04:31.38 | MonMotha | and their service is VERY good for the price |
| 04:31.44 | MonMotha | heck, it's good even at higher prices |
| 04:31.46 | ds2 | all i wanted was terms but instead they give me grief BAH |
| 04:31.59 | ds2 | I am willing to pay more for some reasonable web sites not this crap |
| 04:32.15 | MonMotha | it should give you specs (just not quote you the price, since that's partially javascript based and uses cookie sessions) |
| 04:32.25 | ds2 | it is bad enough they start insisting their links open new windows |
| 04:32.45 | MonMotha | they do? |
| 04:33.00 | MonMotha | where are you at? |
| 04:33.05 | ds2 | I don't see how that's useful... the only way that could work is I snail mail them floppies and I don't need that aggravation |
| 04:33.10 | ds2 | MonMontha: San Jose |
| 04:33.19 | MonMotha | no, URL |
| 04:33.26 | MonMotha | ds2: you can email them files |
| 04:33.30 | ds2 | http://www.4pcb.com/ |
| 04:33.32 | MonMotha | or FTP it to them, IIRC |
| 04:33.37 | MonMotha | go to www.barebonespcb.com |
| 04:33.39 | ds2 | then I clock in barebones |
| 04:33.40 | MonMotha | that'll quote it directly |
| 04:33.50 | ds2 | right and barebonespcb redirects me to a browserfeedback page |
| 04:34.20 | MonMotha | really... |
| 04:34.24 | MonMotha | I take it you have cookies disabled? |
| 04:34.38 | ds2 | cookies + JS disabled for sites that I do not know |
| 04:34.45 | MonMotha | you really can't expect a website like that to work without cookies. It's possible, yes, but not commonplace by any means |
| 04:34.54 | MonMotha | it's a session based website |
| 04:35.25 | ds2 | that's too bad they design it that way. All I want is basic info. No need to be interactive, i.e. a flier |
| 04:35.28 | MonMotha | I don't know of any PCB fabricators that do online quoting/ordering that don't require cookies |
| 04:35.35 | ds2 | but since they do that, I'll spread the word they are a bad company |
| 04:35.36 | MonMotha | the front page won't come up without cookies? |
| 04:35.45 | MonMotha | dude, I think you're going WAAAY overboard |
| 04:35.48 | ds2 | nope nothing other then their compliant about my browser |
| 04:35.59 | ds2 | <--- I don't tolerate bad vendors like that |
| 04:36.05 | MonMotha | I *abhor* cookies, and I'll turn them on when needed for legit reasons |
| 04:36.10 | MonMotha | ok, pay more |
| 04:36.31 | MonMotha | I have zero complaints about them |
| 04:36.33 | ds2 | that's fine. since I evaluate a lot of stuff from mobile devices anyways |
| 04:36.41 | ds2 | if I can't do it my mobile device, they are more of a hinderance |
| 04:37.17 | MonMotha | I do not have barebonespcb.com whitelisted for cookies, and it works for me |
| 04:37.29 | ds2 | maybe it is the JS... *shrug* |
| 04:37.29 | MonMotha | and you know that most mobile devices do support cookies if they support HTTP/HTML |
| 04:37.30 | SpeedEvil | seems to work dfor me |
| 04:37.38 | MonMotha | the WAP only stuff won't work, but you can't expect it to |
| 04:37.42 | MonMotha | WAP is so limited |
| 04:37.49 | ds2 | I disable cookies + JS as I am on the end of a slow GPRS link |
| 04:38.04 | ds2 | graphics are turned off at times but enough rant |
| 04:38.07 | MonMotha | but not images? |
| 04:38.17 | MonMotha | the javascript and cookies are usually tiny compared to the imaes |
| 04:38.31 | MonMotha | I agree that javascript and cookies get abused, but seriously |
| 04:38.32 | SpeedEvil | cookies add round ttrips |
| 04:38.33 | ds2 | problem with JS is they start using it to animate stuff, open new windows, etc |
| 04:38.41 | SpeedEvil | round trips can be 2s |
| 04:38.41 | MonMotha | so disable that crap |
| 04:38.42 | ds2 | so the answer there is no one allowed to use it |
| 04:38.50 | MonMotha | or pay more for your PCBs |
| 04:38.54 | MonMotha | I don't really care |
| 04:39.10 | MonMotha | but don't go saying they are a terrible company because their website uses javascript and cookies |
| 04:39.15 | MonMotha | that would descript 99% of websites these days |
| 04:39.20 | MonMotha | *describe |
| 04:39.30 | ds2 | it is all about choices =) |
| 04:40.05 | MonMotha | I also think you'll be hard pressed to find a PCB fabricator's website that does online quoting that doesn't require JS/cookies |
| 04:40.16 | MonMotha | you MIGHT get specs, but certainly not a quote |
| 04:40.29 | ds2 | when I am ready to submit an order, JS + cookies are fine |
| 04:40.36 | MonMotha | well, then there's pcbnet.com's website...giant. fucking. image. |
| 04:40.57 | ds2 | don't deal with that either |
| 04:41.10 | MonMotha | not even to name your own price on your boards? |
| 04:41.18 | ds2 | <-- very strict on vendors I will deal with |
| 04:41.30 | MonMotha | apparently |
| 04:41.45 | ds2 | nope. problem is I been burned by poor websites leaking CC info out |
| 04:41.53 | *** join/#edev DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1) |
| 04:41.54 | ds2 | if they don't look right, run away |
| 04:42.59 | MonMotha | that's what $0 fraud liability is for |
| 04:43.12 | ds2 | have you tried taking them up on that? =) |
| 04:43.13 | MonMotha | that and you can usually tell a downright shady website from a just stupid one |
| 04:43.17 | MonMotha | yes, actually |
| 04:43.24 | MonMotha | my CC company has been very helpful |
| 04:43.31 | ds2 | what bank, if I may ask? |
| 04:43.35 | MonMotha | National City |
| 04:43.40 | ds2 | hmmmm |
| 04:43.53 | MonMotha | I've heard the same from others who have filed fraud activity notices with them |
| 04:44.02 | davidc__ | same here - someone stole mail from us at put 10k worth of charges on our families cards |
| 04:44.04 | MonMotha | it's pretty much "give us a couple hours" and then you get your account credited |
| 04:44.07 | davidc__ | *and |
| 04:44.14 | davidc__ | and total cost to us was $0 |
| 04:44.19 | MonMotha | especially if it's obvious fraud |
| 04:44.25 | MonMotha | which leaked info from online usually is |
| 04:44.36 | MonMotha | heck, I know people who have been called by the bank to say "this doesn't look right" |
| 04:44.39 | ds2 | maybe it is just Chase then. they never credited me; made up stories on the phone; had to send them registered mail |
| 04:44.56 | ds2 | and this is after the vendor been called and agree it was fraud |
| 04:45.11 | ds2 | eventually, after a few months, I got credit back for it |
| 04:45.12 | MonMotha | well, Chase is a huge bank that gives out LOTS of cards to people with mediocre credit ratings |
| 04:45.26 | MonMotha | local or smaller banks and credit unions tend to do bette |
| 04:45.29 | MonMotha | or get an amex |
| 04:45.37 | MonMotha | I've heard VERY good things about amex |
| 04:45.41 | davidc__ | just a quick phonecall to visa |
| 04:45.56 | ds2 | hmmm |
| 04:46.12 | MonMotha | credit card disputes tend to actually be kinda one-sided in favor of the account holder |
| 04:46.20 | MonMotha | the merchant pretty much has to prove the charge was legit |
| 04:46.26 | ds2 | think part of it is their phone service is off shore |
| 04:46.43 | MonMotha | you, sir, need a better credit card |
| 04:46.45 | ds2 | things fell in place shortly after I sent them a register mail but they don't reimburse for postage |
| 04:47.07 | MonMotha | registered mail for a letter is usually only about $15 |
| 04:47.18 | MonMotha | an inconvenience, yes, but not th eend of the world |
| 04:47.34 | ds2 | *nod* but not having it happen is better |
| 04:47.46 | ds2 | it is more of having to constantly follow up that was the bigger hassle |
| 04:48.01 | MonMotha | true, but there's not a lot that they can do to proactively prevent it other than look for things out of pattern |
| 04:48.05 | MonMotha | which many banks do in fact do |
| 04:48.32 | MonMotha | and once fraud is suspected, they should cancel the old CCN and issue a new one to you |
| 04:49.09 | ds2 | oh that they did do; but reversing the charges... that's the painful part; the reason I am with chase has to do with rebate programs |
| 04:49.26 | MonMotha | I get 5% on gas, 2% on groceries, and 1% on everything else |
| 04:49.30 | MonMotha | seems decent to me |
| 04:49.47 | MonMotha | it does help to have other accounts with the bank |
| 04:49.53 | ds2 | I get 3% flat on everything with that crappy card |
| 04:50.06 | MonMotha | even if it just means you can hint at hte slightest possibility of taking your accounts elsewhere |
| 04:50.11 | MonMotha | ah yes, that card |
| 04:50.11 | davidc__ | yup - I've had calls from my card issuer saying "Hey, we think somethings up", is this you buying stuff? |
| 04:50.23 | Mirell | davidc__: I've had more annoying than that. |
| 04:50.33 | MonMotha | I don't consider calls like that annoying at all |
| 04:50.41 | Mirell | My check card was deactivated, and they never told me. |
| 04:50.45 | davidc__ | eh, I appreciate them calling me |
| 04:50.53 | ds2 | The sponsor of the card used to do it through citibank and citibank was great at dealing with this. One call, and things get suspended but not chase |
| 04:50.54 | MonMotha | I'd rather then try to proactively catch fraud than let it happen and make me deal with it |
| 04:50.57 | davidc__ | doesn't bother me in the slightest to know that their systems notice when I start buying stuff out of pattern |
| 04:51.08 | Mirell | Because obviously a $300 purchase at the University bookstore which I've purchased 50+ times in the past was fraudulent. |
| 04:51.19 | MonMotha | seems odd that they would think that |
| 04:51.34 | MonMotha | also seems odd that they would deactivate the card immediately without contacting you |
| 04:51.38 | MonMotha | was your contact info up to date? |
| 04:51.42 | Mirell | MonMotha: Which is why it was annoying |
| 04:51.49 | Mirell | MonMotha: Same cell phone number for 6+ years. |
| 04:51.55 | MonMotha | what bank? |
| 04:51.59 | Mirell | BoA |
| 04:52.03 | MonMotha | again, giant bank |
| 04:52.06 | Mirell | Ayup |
| 04:52.13 | MonMotha | the problems all seem ot be with people who are at giant, faceless banks |
| 04:52.33 | MonMotha | I've got branches all over the place with ATMs, and I get actual service |
| 04:52.43 | Mirell | Yah, I'll probably be switching to a local credit union here eventually |
| 04:52.46 | MonMotha | heck, before I moved, some of the tellers at my local branch knew me by name |
| 04:52.59 | Mirell | But don't want to do that right now, since it'll muck up my first direct deposit |
| 04:54.36 | ds2 | maybe i will give 4pcb a call when I can find time during business hours |
| 04:55.59 | MonMotha | ds2: every time I've talked with them, they've been pretty eager to help out |
| 04:56.04 | MonMotha | you can even ask their CAM guys questions |
| 04:56.20 | MonMotha | oh, the barebones boards are also free if you place a full spec prototype order later |
| 04:56.32 | MonMotha | you can get a discount equal to your barebones purchase |
| 04:56.40 | ds2 | sounds nice |
| 04:56.42 | MonMotha | it |
| 04:56.45 | MonMotha | it's a hell of a deal |
| 04:57.00 | MonMotha | their full spec protos arne't the cheapest of things, but they are at least competitive |
| 05:01.12 | ds2 | nice thing about working for a big company is they have groups to deal with vendors |
| 05:06.03 | Mirell | Hmm... |
| 05:06.04 | MonMotha | it can be handy to work directly, to |
| 05:06.12 | MonMotha | but yeah, sometimes I'd like to push that duty off |
| 05:06.14 | *** join/#edev roxfan2 (n=dunno@89.49-240-81.adsl-dyn.isp.belgacom.be) |
| 05:51.30 | *** join/#edev RobotGuy (n=RobotGuy@dsl093-038-120.pdx1.dsl.speakeasy.net) |
| 06:04.06 | *** join/#edev chouimat|away (n=dieu@kde/developer/chouinard) |
| 06:54.30 | *** join/#edev jbevren (n=jbevren@67.55.212.44) |
| 07:17.02 | *** join/#edev alnr (n=e4ward@e4ward.com) |
| 07:17.12 | *** part/#edev alnr (n=e4ward@e4ward.com) |
| 07:29.37 | *** join/#edev redblue (i=star@ppp193.108-253-207.mtl.mt.videotron.ca) |
| 08:04.31 | *** join/#edev hw_ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 08:57.31 | *** join/#edev lyakh (n=lyakh@p57BD292B.dip0.t-ipconnect.de) |
| 13:09.56 | *** join/#edev redblue (i=star@ppp022.108-253-207.mtl.mt.videotron.ca) |
| 13:34.16 | *** part/#edev T0mW (n=Tom@24.238.86.158.res-cmts.sth.ptd.net) |
| 13:55.52 | *** join/#edev RobotGuy (n=RobotGuy@dsl093-038-120.pdx1.dsl.speakeasy.net) |
| 14:10.22 | *** join/#edev hapchap (n=hapchap@mail.americanmicrosystems.com) |
| 14:24.21 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 14:29.22 | *** join/#edev GPSFan (n=kenm@12.10.255.246) |
| 14:29.22 | *** mode/#edev [+o GPSFan] by ChanServ |
| 14:39.47 | RobotGuy | Good morning. |
| 14:41.15 | RobotGuy | Is it still morning when one has not slept? |
| 14:44.11 | RobotGuy | I finally got a toolchain that will build libusb. |
| 14:49.32 | prpplague | who is "they" ? |
| 14:50.26 | RobotGuy | The Buildroot folks. |
| 14:50.42 | RobotGuy | I did what you suggested and tried a later buildroot. |
| 14:51.00 | prpplague | RobotGuy: and? |
| 14:51.13 | RobotGuy | Nothing of the Samsung stuff in this version. |
| 14:51.59 | prpplague | RobotGuy: yea, you have to copy that over, but you can use the generic build if you wanted to |
| 14:52.04 | RobotGuy | Are those changes you made to buildroot to support Hammer? |
| 14:52.24 | RobotGuy | prpplague: I am using the generic build now but can't get hammer to boot with the initrd it created. |
| 14:52.28 | prpplague | RobotGuy: yea |
| 14:52.43 | RobotGuy | Do you want me to upload the package of buildroot I have now? |
| 14:53.04 | RobotGuy | prpplague: I already patched in Squeak. |
| 14:54.47 | RobotGuy | prpplague: I just have to copy the Samsung directory to target/devices? |
| 14:54.49 | prpplague | RobotGuy: once the squeak stuff stabilizes i'll put into the main buildroot |
| 14:55.00 | RobotGuy | prpplague: Right |
| 14:55.09 | prpplague | RobotGuy: and add it the makefile and config includes |
| 14:55.52 | RobotGuy | prpplague: Makefiles scare me. I ususally don't go there unless I really have to. |
| 14:56.06 | prpplague | RobotGuy: its just a single line entry |
| 14:56.21 | RobotGuy | prpplague: I will look |
| 14:56.33 | prpplague | RobotGuy: its kinda moot anyway, as the samsung will be pushed to the main buildroot in a week or so |
| 14:59.32 | RobotGuy | prpplague: Cool, but we need it now. :) |
| 14:59.45 | prpplague | RobotGuy: why? |
| 14:59.46 | RobotGuy | prpplague: Robert Sheer ordered a Hammer |
| 14:59.59 | prpplague | RobotGuy: yea it was shipped on friday |
| 15:00.21 | RobotGuy | prpplague: Why would I not need it now? OK, well, I guess not. I am trying to build a new toolchain from your buildroot now. |
| 15:00.49 | prpplague | RobotGuy: the shipped tool chain works fine\ |
| 15:02.19 | RobotGuy | prpplague: Yeah, yeah, but I am determined to get this to work for me. I have to learn this eventually. |
| 15:02.30 | prpplague | RobotGuy: whats not working now? |
| 15:02.54 | RobotGuy | prpplague: I could not get a toolchain that builds libusb. |
| 15:03.11 | prpplague | RobotGuy: mine built libusb fine |
| 15:03.21 | prpplague | RobotGuy: sounds like you have something misconfigured |
| 15:03.30 | RobotGuy | prpplague: I know I have something wrong. I will get it. |
| 15:03.46 | RobotGuy | prpplague: I did get a good toolchain with the new version of buildroot. :D |
| 15:03.57 | RobotGuy | So there IS hope. :D |
| 15:05.29 | *** join/#edev bbradley (n=bbradley@82-32-35-42.cable.ubr05.azte.blueyonder.co.uk) |
| 15:06.02 | RobotGuy | I did build Squeak from Buildroot once already. |
| 15:06.16 | *** join/#edev chouimat|work (n=mchouina@209.217.106.98) |
| 15:06.34 | prpplague | RobotGuy: i get the feeling you are trying to do too many things at once |
| 15:06.55 | chouimat|work | morning |
| 15:06.59 | RobotGuy | prpplague: I am just trying to get a good toolchain right now. |
| 15:07.10 | RobotGuy | prpplague: I am learning a lot too. |
| 15:07.19 | prpplague | chouimat|work: greetings |
| 15:07.32 | prpplague | RobotGuy: you had one before, what was wrong with it? |
| 15:07.45 | chouimat|work | hi prpplague |
| 15:07.53 | RobotGuy | prpplague: It would not build libusb. |
| 15:08.41 | prpplague | RobotGuy: and you enabled cross-compile g++ as well as the build/install g++ options ? |
| 15:09.00 | prpplague | RobotGuy: all i had to do was turn those on, and run make |
| 15:09.07 | RobotGuy | prpplague: I thought I had, but I may still have had an option that conflicted. |
| 15:09.49 | *** join/#edev CosmicPenguin (n=nobody@163.181.251.103) |
| 15:09.49 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 15:10.28 | RobotGuy | G'Morning CosmicPenguin |
| 15:10.55 | RobotGuy | prpplague: What seems so easy for you, is not necessarily easy for a beginner though. |
| 15:15.18 | *** join/#edev samuelr (n=samuelr@dsl093-061-094.pit1.dsl.speakeasy.net) |
| 15:17.36 | CosmicPenguin | ugh |
| 15:17.51 | CosmicPenguin | I put my finger on my DOT heatsync and it was enough to turn on the fan |
| 15:17.56 | CosmicPenguin | that hurt |
| 15:35.28 | RobotGuy | Well, this toolchain build is looking good, so I am going to get some sleep. Back later. |
| 15:46.09 | Mirell | MacWorld today! |
| 15:46.53 | samuelr | Mirell: Awwww, YEAH! |
| 15:47.08 | samuelr | It's my absolute favorite reality TV show :-) |
| 15:47.55 | Mirell | Hehe |
| 15:58.32 | *** join/#edev landley (n=landley@cpe-72-177-114-214.austin.res.rr.com) |
| 15:59.45 | prpplague | samuelr: hey, whats cookin? |
| 15:59.58 | prpplague | samuelr: don't suppose you found any time to tinker with the hammer? |
| 16:03.02 | *** join/#edev sunrunner20 (n=jamesdoe@pool-71-252-240-120.dllstx.fios.verizon.net) |
| 16:05.21 | *** join/#edev g1powermac (n=g1powerm@unaffiliated/g1powermac) |
| 16:05.49 | Mirell | Only one more hour. |
| 16:28.41 | *** mode/#edev [+o prpplague] by ChanServ |
| 16:33.27 | *** join/#edev redblue (i=star@ppp103.109-253-207.mtl.mt.videotron.ca) |
| 16:37.46 | *** join/#edev ico2 (n=ico2@modem-3906.chimpanzee.dialup.pol.co.uk) |
| 16:41.04 | Mirell | 20 more minutes! |
| 16:57.31 | g1powermac | Mirell, hmm, waiting for the macworld keynote? |
| 16:58.23 | *** join/#edev samuel1 (n=samuelr@dsl093-061-094.pit1.dsl.speakeasy.net) |
| 16:59.20 | samuel1 | prpplague: Unfortunately, no. |
| 16:59.25 | Mirell | g1powermac: Ayup |
| 17:00.01 | g1powermac | ya, me too, bad thing is I half to leave my computer about 10 mins after the keynote starts |
| 17:00.08 | g1powermac | err |
| 17:00.08 | Mirell | Alas! |
| 17:00.12 | g1powermac | s/half/have |
| 17:00.21 | Mirell | I already got a bingo card on my desk |
| 17:00.29 | g1powermac | heh |
| 17:00.59 | CosmicPenguin | You are all silly |
| 17:01.27 | Mirell | CosmicPenguin: No tickets which are open which I can work on, due to lack of access still. |
| 17:01.31 | g1powermac | CosmicPenguin, no, we're just different ;-) |
| 17:02.36 | Mirell | THEY LETTING PEOPLE SIT DOWN OMG!!!!111!!!!11!!111!1111eleven |
| 17:02.56 | g1powermac | Mirell, umm, that is a bit much just for letting people sit down ;-) |
| 17:03.08 | Mirell | I know. |
| 17:03.19 | Mirell | I'm pretty bored at work. |
| 17:03.29 | g1powermac | unless it was a direct order from Steve ;-) |
| 17:04.21 | g1powermac | man, can't believe I'm gonna have to leave. . . |
| 17:04.54 | g1powermac | at least I got a free t-shirt coming from a friend of mine who is actually attending macworld :-) |
| 17:05.43 | Mirell | Hehe |
| 17:34.33 | *** join/#edev landley (n=landley@adsl-76-197-33-46.dsl.austtx.sbcglobal.net) |
| 17:42.00 | *** join/#edev DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1) |
| 17:42.35 | landley | Mirell: one more hour until what? |
| 17:42.51 | Mirell | landley: SteveNote |
| 17:42.57 | Mirell | Currently, it appears to be lame |
| 17:57.44 | landley | I thought he usually saves whatever he has to say to the very end. |
| 17:58.10 | landley | Is mac world annual, or twice yearly? |
| 17:58.18 | landley | I thought the last one was the iPhone, but it hasn't been out long enough... |
| 17:59.32 | Mirell | landley: Twice |
| 18:02.51 | landley | So what has he introduced so far? |
| 18:03.20 | Mirell | iTunes movie rentals, and HD video on iTunes |
| 18:06.16 | landley | The iTunes movie rentals I read about days ago (and netflix has already gone to the all you can eat online viewing option). |
| 18:07.32 | *** join/#edev ico2 (n=ico2@modem-873.cougar.dialup.pol.co.uk) |
| 18:09.56 | Mirell | Macbook Air now |
| 18:11.23 | landley | Macbook air? |
| 18:11.36 | landley | (Not related to those shoes you pump up?) |
| 18:12.03 | Mirell | landley: An ultra-portable laptop |
| 18:14.14 | landley | Like the Dell you're looking at getting? |
| 18:14.52 | Mirell | landley: Yah |
| 18:15.07 | landley | Thinking of changing your mind, or still going for the Dell? |
| 18:15.54 | Mirell | Doubtful I'll get the Apple one. |
| 18:16.02 | Mirell | Doesn't look to be as feature filled |
| 18:16.06 | Mirell | And more expensive |
| 18:16.11 | Mirell | For only slightly less weight/size |
| 18:16.46 | landley | And doesn't have a powerful enough 3D chip? |
| 18:17.19 | Mirell | Also only a 1.8" drive |
| 18:17.40 | SpeedEvil | I'm still surprised at how well my 7 year old (though re-celled) laptop holds up. |
| 18:17.53 | SpeedEvil | 1.1Kg - 8h battery life, 2.5" drive, 10" display |
| 18:18.07 | SpeedEvil | I added wifi and bluetooth though. |
| 18:18.21 | landley | Has the intel chipset with actual open source 3d drivers, dual processor 64-bit, 2 gigs ram, the works. |
| 18:18.34 | landley | It is also, alas, large enough to stop a charging rhino with a single blow to the snout. |
| 18:18.56 | Mirell | Damn. |
| 18:18.59 | Mirell | MacBook Air is tiny |
| 18:19.31 | landley | Depends whether or not the rhino is on lithium. |
| 18:19.40 | Mirell | http://s3.media.macrumorslive.com/p/f1200420839.jpg |
| 18:19.56 | Mirell | http://s3.media.macrumorslive.com/p/f1200420757.jpg |
| 18:25.12 | landley | Thin yes, small no. |
| 18:25.22 | landley | And if they didnt' make that sucker out of titanium or something, it looks _really_ fragile. |
| 18:25.41 | landley | Hit a rhino on the snout with that, the rhino wouldn't even notice. |
| 18:25.58 | landley | Might stop a charging squirrel, but only if it was already near its credit limit. |
| 18:27.02 | Mirell | They did make it out of titanium |
| 18:27.06 | CosmicPenguin | I wish there was more newbish name for 'sysfs' - its wierd trying to discuss the interface with such a technical name |
| 18:27.06 | Mirell | $1800 |
| 18:27.32 | Mirell | Er, aluminum, sorry. |
| 18:30.49 | Mirell | Welp, keynote is over. |
| 18:30.57 | *** part/#edev samue1 (n=samuelr@dsl093-061-094.pit1.dsl.speakeasy.net) |
| 18:31.17 | *** join/#edev samuelr (n=samuelr@dsl093-061-094.pit1.dsl.speakeasy.net) |
| 18:32.10 | *** join/#edev chouimat|busy (n=dieu@r2351064.cidc.net) |
| 18:34.08 | CosmicPenguin | I didn't feel the earth reverse on its axis |
| 18:38.34 | prpplague | CosmicPenguin: ? |
| 18:47.39 | landley | An aluminum laptop that think isn't going to last 5 minutes. |
| 18:48.04 | landley | CosmicPenguin: why is sysfs a hard name? |
| 18:49.00 | landley | There's even a proverb about how documenting that filesystem is like rolling boulders uphill and watching them roll back down again... |
| 18:49.18 | chouimat|work | Mirell: 3100 if you want the model with a 64GB solid state hd |
| 18:50.30 | landley | Not exactly compelling. |
| 18:51.51 | chouimat|work | landley: a lot of people will get one anyway :) |
| 18:52.21 | *** join/#edev RobotGuy (n=RobotGuy@dsl093-038-120.pdx1.dsl.speakeasy.net) |
| 18:53.58 | landley | Well of course. It's apple. |
| 18:55.52 | chouimat|work | landley: I know if my grandmother wanted a laptop she probably look into this one because :' it doesn't use much space when closed' |
| 18:58.03 | landley | They need to bring back the butterfly keyboard. |
| 18:58.28 | landley | Making a laptop paper thin doesn't make it any less wide. |
| 18:58.52 | landley | The screen you can up the resolution on, but the keyboard's limited by the size of your hands. |
| 18:59.07 | landley | In japan, they can make them smaller because the populace statistically has smaller hands... |
| 19:00.16 | RobotGuy | landley: Greetings! |
| 19:00.25 | landley | Hiya RobotGuy. |
| 19:05.20 | RobotGuy | How is everything? |
| 19:06.25 | RobotGuy | Hi SpeedEvil |
| 19:06.30 | landley | Still unemployed, still waffling what to do about it. The usual. |
| 19:06.32 | landley | And you? |
| 19:06.56 | RobotGuy | landley: I am doing better. :) I think I finally conquered the toolchains! |
| 19:07.04 | landley | I'll probably get a release of tinycc and another release of firmware out this month, then go get a job doing somethign else and ignore the projects for at least 3 months, probably more. |
| 19:07.19 | landley | Not what I _want_ to do, but money is apparently a requirement. |
| 19:07.21 | landley | Go figure. |
| 19:07.38 | RobotGuy | landley: Indeed! I wish I had more of that green stuff. |
| 19:08.02 | RobotGuy | Compiler? |
| 19:08.12 | landley | Yes, tinycc is a compiler. |
| 19:08.38 | RobotGuy | I respect those who can write compilers/ |
| 19:09.08 | RobotGuy | I've thought if writing my own robotics control language a few times. |
| 19:09.19 | SpeedEvil | logo! |
| 19:09.35 | RobotGuy | Isn't Logo mostly for Kids? |
| 19:09.49 | RobotGuy | There is Prolog also but I don't know anything about it. |
| 19:10.14 | RobotGuy | Oh yeah, and I have to get that Cobol port going.. Heh heh heh |
| 19:10.18 | landley | I didn't write this compiler, I inherited it. |
| 19:10.30 | landley | Wrote my own assembler back in 1990... |
| 19:10.42 | landley | (Sheesh, 1990? Counts...) |
| 19:10.44 | landley | I'm old. :P |
| 19:10.59 | RobotGuy | Who did you inherit it from, and were you better for the inheritance? |
| 19:11.10 | RobotGuy | I am old also. |
| 19:11.34 | RobotGuy | You must be object oriented to inherit. |
| 19:13.36 | Mirell | chouimat|work: ffs. |
| 19:13.45 | chouimat|work | Mirell: ? |
| 19:13.55 | Mirell | chouimat|work: The $3100 price mark |
| 19:13.56 | landley | Fabrice Bellard abandoned it to go work on qemu instead. |
| 19:14.08 | landley | I started poking at it in 2006 when its cvs hadn't had a checkin in about a year. |
| 19:14.19 | landley | I bundled up a couple dozen changes and sent them to Fabrice, who applied them. |
| 19:14.28 | landley | But it didn't spark any major new work on the thing. |
| 19:14.39 | chouimat|work | Mirell: it's an ok price mark for a computer ... but I find it a little underpowered |
| 19:14.39 | RobotGuy | landley: Is it useful for embedded stuff? |
| 19:14.52 | Mirell | chouimat|work: So do I. I'm getting the Dell XPS M1330 |
| 19:14.55 | landley | Well, it's a C99 compiler that fits in 100k. |
| 19:15.05 | landley | It's totally self-contained, no linker or anything needed. |
| 19:15.08 | RobotGuy | landley: How much of C does it have? |
| 19:15.17 | landley | It needs libc and the headers from there, but that's about it. |
| 19:15.20 | landley | RobotGuy: c99. |
| 19:15.21 | landley | All of it. |
| 19:15.27 | landley | Except for a couple of c99 corner cases. |
| 19:15.34 | landley | Variable arrays aren't implemented yet, nor are complex numbers. |
| 19:15.40 | RobotGuy | landley: Oh, cool. |
| 19:15.48 | landley | And it only has some of the gnu extensions. Enough to build a subset of the linux kernel, but not the whole unmodified thing yet. |
| 19:16.03 | landley | It has an assembler, but doesn't produce assembly output, which screws up the kernel build. |
| 19:16.15 | chouimat|work | Mirell: I'm think of getting a new motherboard cpu/ram and replace my P3/667 ... and reuse the p3/667 in a dying server (ok the server is a p3/1000 ... or maybe buy a new motherboad for the p3/1000 and a new case for the p3/667) |
| 19:16.18 | RobotGuy | landley: I wrote some complex number stuff around 1985 for electronics - impedance stuff |
| 19:16.18 | landley | The kernel wants assembly output to run sed against to generate one of its header files. |
| 19:16.36 | RobotGuy | landley: I did complex numbers in Pascal. |
| 19:16.38 | landley | I don't particularly care about complex numbers, but I _do_ want variable arrays. |
| 19:16.47 | landley | Faking it with alloca() doesn't quite cut it. That leaks in loops, it seems. |
| 19:16.48 | chouimat|work | Mirell: I know they're old but for build/test machine they are ok :) |
| 19:17.14 | RobotGuy | I had to use complex numbers for phase angle calculations. |
| 19:17.38 | landley | I think compex numbers should be a struct in a library, not a base language feature. |
| 19:17.45 | landley | But I wasn't on the c99 committee. :) |
| 19:20.27 | ds2 | btw, if anyone is interested... it _seems_ (haven't taken the plunge) that pcbpool's prices are pretty decent and from what they are saying, no restrictions on how the square inches can be partitioned |
| 19:21.04 | ds2 | and they have a claimed turn around time; not as cheap as SFE but less crap then other folks |
| 19:41.26 | RobotGuy | Is it possible to boot from a Thumb Drive? |
| 19:42.04 | RobotGuy | To boot Hammer, that is, with Linux. |
| 19:51.27 | landley | RobotGuy: depends on the bootloader. |
| 19:51.42 | landley | The SOC boots from the built-in flash, but that bootloader could in theory talk to the thumb drive and hand off. |
| 19:52.17 | RobotGuy | landley: Seems it should be possible. |
| 19:52.21 | landley | I keep meaning to attack the -tiny tree (now that tim bird's maintaining it again) and get a tiny linux kernel built. |
| 19:52.42 | landley | Easiest thing to do is have the bootloader load linux and have the initramfs in linux hand off to something else, ala that old "two kernel monte" thing. |
| 19:52.45 | RobotGuy | landley: I like having the full featured kernel. |
| 19:52.48 | landley | What do they call it these days, kexec? |
| 19:53.02 | landley | I want to build a minimal possible "hello world" kernel from a half-dozen C files. |
| 19:53.07 | landley | Just to see what's _involved_. |
| 19:53.12 | landley | Turns out to be nontrivial, though. :P |
| 19:53.19 | RobotGuy | landley: That's what regular booting does on a PC, if they use an initrd. |
| 19:53.22 | landley | (See "tinycc has no assembler", above.) |
| 19:53.43 | landley | initramfs, not initrd. |
| 19:53.49 | landley | initrd is obsolete and _way_ more complicated. |
| 19:54.00 | RobotGuy | landley: I'd like to get rid of the initrd, but that may not be doable for embedded systems like Hammer. |
| 19:54.17 | landley | You can always replace an initrd with an initramfs. |
| 19:54.30 | landley | You can even load an external initramfs. It's just in cpio format rather than ext2 or minix or squashfs. |
| 19:54.32 | landley | Not a block device. |
| 19:54.40 | RobotGuy | landley: I am still a little foggy on the difference between the two. |
| 19:56.58 | RobotGuy | landley: I have my script for building my rootfs and kernel setup to use a few environment variables now. |
| 19:57.02 | landley | initramfs is a kernel data section (always linked into the zImage) containing a cpio archive that gets extracted into rootfs, which is an instance of ramfs. |
| 19:57.54 | RobotGuy | landley: The initrd is a bootable system image, right? |
| 19:58.09 | adyer_ | RobotGuy you can boot with no init fs - we use u-boot, boot a kernel out of raw flash, and the kernel mounts a jffs2 root filesystem with no init stuff |
| 19:58.11 | landley | initrd is an external file, loaded separately by the bootloader and fed to the kernel via a command line argument with the physical address of where the bootloader decided to load it, containing a filesystem image. It's loaded into a ramdisk block device and interpreted through a filesystem driver such as ext2.ko. |
| 19:58.30 | landley | Initramfs is not a filesystem format. It's an archive. |
| 19:58.33 | landley | It's like a tarball. |
| 19:58.41 | RobotGuy | landley: I understand that. |
| 19:58.43 | landley | it gets _extracted_ into ramfs. |
| 19:58.50 | landley | So when it's running, there's no filesystem driver. |
| 19:58.56 | landley | ramfs basically mounts the disk cache as a filesystem. |
| 19:59.03 | landley | It uses dentries and page cache to store the filesystem contents. |
| 19:59.21 | landley | Meaning it dynamically sizes itself: writing more data allocates more memory, deleting or truncating files frees memory. |
| 19:59.36 | RobotGuy | landley: One thing I need to do is after I create my initrd, I need to copy the /usr stuff to my thumb drive and then recreate the initrd with just the /usr mount point. |
| 19:59.38 | landley | Plus there's less code involved, because you can never _not_ have the disk cache running. |
| 20:00.06 | landley | You always have rootfs mounted in a 2.6 kernel. |
| 20:00.15 | landley | They made it the head/tail of the doubly linked mounted filesystems list. |
| 20:00.30 | RobotGuy | landley: Do you understand what I want to do? |
| 20:00.31 | landley | They never have to check whether the list is empty, because that entry is always there. |
| 20:00.34 | landley | It's like PID 1. |
| 20:00.47 | landley | RobotGuy: I think so. |
| 20:00.55 | landley | You can create an initramfs with "cpio". |
| 20:01.05 | landley | I documented how to do this once upon a time... |
| 20:01.26 | RobotGuy | landley: I just want to create the initrd with a /usr mount point, after I copy /usr to my thumb drive. |
| 20:01.37 | adyer_ | RobotGuy you might also want to look into "pivot root" |
| 20:01.42 | RobotGuy | landley: OK, I need to read up on cpio then. |
| 20:01.57 | RobotGuy | adyer_: I want the easiest way to do what I want to do. |
| 20:02.04 | landley | (cd "$1"; find . | cpio -o -H newc | gzip) > "$2" |
| 20:02.25 | RobotGuy | landley: Huh? |
| 20:02.27 | landley | cd to the directory, do "find . | cpio -o -H newc | gzip > ../blah.img" |
| 20:02.52 | landley | Or just read Documentation/filesystems/ramfs-rootfs-initramfs |
| 20:02.54 | landley | .txt |
| 20:02.58 | RobotGuy | cd to /usr to do this. |
| 20:03.06 | RobotGuy | I will read. |
| 20:03.47 | landley | adyer_ pivot_root is for initrd only. For initramfs you need switch_root. |
| 20:04.11 | landley | (By the way, if you ever run out of syrup of ipcac, read up on the pivot_root implementation sometime.) |
| 20:04.16 | landley | It's amazingly disgusting. |
| 20:04.25 | landley | It's not a bad implementation, it turns out to be a _hard_problem_. |
| 20:04.54 | landley | In order to actually move the root mount point and free it so you can actually remove it, you've got to make sure nothing's using it. |
| 20:05.13 | landley | That means you have to iterate through every single process and move their "chroot" pointer, and their "cwd" pointer, and their ".." pointer. |
| 20:05.41 | landley | (Each process has three "current directory" indicator superblock references, any of which can point to the / superblock. |
| 20:05.43 | adyer_ | landley if I want to vomit I have a IBM PC-AT tech ref manual that does the trick |
| 20:05.57 | landley | In fact, the root one usually _does_. And for anything that did daemonize(), all three of 'em should. |
| 20:06.01 | landley | And that includes kernel threads. |
| 20:06.22 | landley | Yes, you have to go through and move kernel threads in order to pivot_root. |
| 20:06.36 | landley | Then do surgery on the mount tree in a way that has only vaguely defined semantics. |
| 20:06.57 | landley | essentially, you're moving a mount point under itself. |
| 20:07.03 | landley | neither absolute nor relative paths avoid a race condition there. |
| 20:09.49 | landley | According to Mirell's father, the development codename of Vista was "Hindenburg" |
| 20:14.27 | *** join/#edev ico2 (n=ico2@modem-873.cougar.dialup.pol.co.uk) |
| 20:20.37 | CosmicPenguin | somehow, I doubt that |
| 20:22.41 | landley | Yeah, but like calling Indiana "GNU/Solaris", it's fun to say. |
| 20:49.40 | Mirell | Hmm... |
| 20:49.43 | Mirell | Oh, thank god. |
| 20:49.59 | RobotGuy | Mirell: How is everything today? |
| 20:50.00 | Mirell | They finally turned on manually manage music on the iPhone with the latest iTunes |
| 20:50.05 | Mirell | RobotGuy: Fairly well |
| 20:50.20 | RobotGuy | Mirell: I have finally conquered the toolchains! :D |
| 20:59.04 | *** join/#edev hapchap (n=hapchap@mail.americanmicrosystems.com) |
| 21:12.25 | Mirell | Sweet |
| 21:12.30 | Mirell | Man, Apple's stock took a plunge |
| 21:13.50 | landley | RobotGuy: which toolchains? |
| 21:14.08 | landley | Mirell: well it's not surprising given that steve's speech at macworld seems to have been content-free. |
| 21:14.43 | RobotGuy | landley: The buildroot toolchain. |
| 21:15.05 | landley | Ah. |
| 21:24.37 | RobotGuy | So far, I have 20 MB RAM free on my Hammer. |
| 21:32.56 | Mirell | landley: Oh dear. MacBook Air has no user-replacable battery |
| 21:35.32 | landley | Mirell: yeah, that's Apple. |
| 21:35.55 | landley | Making it thin without making it less wide really doesn't strike me as a particularly useful approach. |
| 21:36.01 | landley | Especially in the wake of the Sony Vaio. |
| 21:36.15 | landley | That was all cool and shiny and none of them lasted a full year of actual use. |
| 21:39.53 | *** join/#edev samuelr (n=samuelr@dsl093-061-094.pit1.dsl.speakeasy.net) |
| 21:50.33 | *** join/#edev ico2 (n=ico2@modem-873.cougar.dialup.pol.co.uk) |
| 22:08.21 | Mirell | Still no real work done |
| 22:08.47 | landley | You're in your second week, right? |
| 22:09.16 | Mirell | 3rd |
| 22:18.40 | landley | Eh, you're still running about average. |
| 22:24.17 | Mirell | It's amusing how much businesses waste. |
| 22:51.03 | *** join/#edev redblue (i=star@ppp212.109-253-207.mtl.mt.videotron.ca) |
| 23:23.13 | landley | Mirell: big businesses, yes. |
| 23:23.20 | landley | Small businesses tend to be efficient, because they have to. |
| 23:26.49 | RobotGuy | I am still having problems getting mpfr to build. |
| 23:26.57 | *** join/#edev Redhatter (n=stuartl@gentoo/developer/redhatter) |
| 23:43.12 | landley | What's mpfr? |
| 23:43.19 | landley | (Buy a vowel?) |
| 23:43.36 | RobotGuy | landley: I don't really know. It's required to build a native toolchain. |
| 23:47.58 | *** join/#edev landley_ (n=landley@adsl-76-197-33-46.dsl.austtx.sbcglobal.net) |
| 23:48.39 | landley_ | Does anybody know about arm floating point? |
| 23:48.41 | landley_ | There's got to be a standard somewhere. |
| 23:48.45 | landley_ | I encountered some code that does an endianness reversal, except it's doing so on an array of _ints_ in a union with a double. |
| 23:48.48 | landley_ | That can't be right... |
