| 00:00:54 | Aelhaz | yup |
| 00:01:26 | Aelhaz | it's eensy |
| 00:01:34 | RedBeard | al: cool |
| 00:01:41 | Aelhaz | I can email you binaries & files |
| 00:01:42 | RedBeard | Aelhaz: how about ssh stuff? |
| 00:01:52 | RedBeard | al: I'm getting my xcompile set up |
| 00:02:08 | RedBeard | Aelhaz: sure, send 'em over |
| 00:02:11 | Aelhaz | I banged my head against a wall a few times with openssl and quit. I have a few ideas what the issue is, but I'm not certian |
| 00:02:28 | Aelhaz | i need a libz compiled against uclibc/arm first, I think |
| 00:03:14 | RedBeard | Aelhaz: okay |
| 00:04:09 | Aelhaz | the whole outlay of files is about 80k |
| 00:06:06 | RedBeard | Aelhaz: nice and small |
| 00:06:08 | Aelhaz | yeah |
| 00:06:34 | CosmicPenguin | Ok, so that experiment failed.... I am outta here |
| 00:07:31 | Aelhaz | if you want logging, thttpd might be a better server. boa logs directly to a file, didn't seem to be a great idea on a 4 meg flash filesystem |
| 00:08:26 | kergoth | boa doesnt support syslog? |
| 00:08:28 | Aelhaz | my httpd is gonna mostly serve 404's to people, so i don't care |
| 00:08:32 | Aelhaz | kergoth: didn't appear to |
| 00:08:37 | kergoth | ah, how annoying |
| 00:09:58 | RedBeard | dinner |
| 00:10:35 | Aelhaz | RedBeard: for ssh, there's either openssh or lsh. openssh might be a tad large, or not, dunno for sure. lsh is smaller, but i haven't figured out how much of it is supposed to work, what features are missing, etc |
| 00:12:08 | Aelhaz | on intel, glibc, openssh sshd is 240k, ssh client is 220k, keygen is 80k. I don't know how that'll relate to arm/uclibc |
| 00:12:30 | Aelhaz | only 3 megs of flash to work in, but it's compressed at least |
| 00:12:55 | RedBeard | Aelhaz: could that stuff be kept on a different box, accessed via nfs? |
| 00:13:07 | RedBeard | I'm being called to dinner |
| 00:14:17 | Aelhaz | RedBeard: I'd rather not, if i can avoid it |
| 00:14:17 | Aelhaz | cya |
| 00:32:27 | Russ|werk | Aelhaz: your main problem is openssl I think |
| 00:32:34 | Russ|werk | that always ends up being really big for me |
| 00:37:38 | Aelhaz | Russ|werk: Yeah, but it can be pared down |
| 00:37:53 | RedBeard | I seem to be having some trouble with embdebian.sourceforge.net |
| 00:38:08 | Aelhaz | you don't *have to build all those ciphers, let alone install them |
| 00:38:38 | RedBeard | apt-get is getting 404 errors on several of the packages |
| 00:38:42 | Aelhaz | could build openssh so the only cipher it supports is blowfish, and link openssl statically |
| 00:39:14 | Aelhaz | but i couldn't get openssl to build against uclibc here |
| 00:42:36 | Russ|werk | Aelhaz: I think openssl would be important enough that andersee would work with a bit if you filed specific bug reports |
| 00:42:44 | Russ|werk | (ie, *why* doesn't it compile) |
| 00:44:47 | Aelhaz | I'm not far enough into the process to give a meaningful bug report |
| 00:46:05 | Aelhaz | right now I'm trying to compile the stuff lsh depends on |
| 00:46:25 | Aelhaz | since there's a report that it works with uclibc |
| 00:47:01 | Aelhaz | liboop doesn't build. *shrug*. |
| 00:50:03 | Aelhaz | . . . don't even see a way to configure gmp for cross compilation . . . |
| 00:52:59 | Aelhaz | and hacking up gmp manually looks like a huge mess. |
| 01:02:20 | Aelhaz | lousy stinking configure script . . . |
| 01:02:23 | Aelhaz | goes back to openssl |
| 01:08:38 | Russ|werk | configure is much better than some of the stuff out there |
| 01:08:42 | Russ|werk | ever compiled ae_? |
| 01:08:45 | Russ|werk | er, ae? |
| 01:08:55 | Russ|werk | or berkley spice? |
| 01:11:51 | Aelhaz | i think i have cross compiled ae |
| 01:12:04 | Aelhaz | actually, I'm certian i did, for a glibc arm system |
| 01:13:48 | RedBeard | I don't remember if ae was in the linkup bsp or not |
| 01:14:08 | Aelhaz | i built it by hand for calypso |
| 01:14:49 | RedBeard | what about the ads graphics client? |
| 01:15:14 | RedBeard | nm, doesn't matter |
| 01:17:00 | Russ|werk | be nice if busybox had ae, the busybox vi is annoying (not a vi user...) |
| 01:17:20 | Aelhaz | bb-vi works basically ok for me |
| 01:17:26 | Aelhaz | I like nano-tiny better than ae |
| 01:17:40 | Aelhaz | but I'm not sure if it's harder to build or not |
| 01:19:04 | RedBeard | is a vi user and finds the bb-vi passable, but not robust enough for serious work |
| 01:20:05 | RedBeard | apt-get is looking for libc6-dev-arm_2.1.3-8e4.deb but the actuall name is libc6-dev-arm_2.1.3-8e4_all.deb |
| 01:20:24 | RedBeard | this is from emdebian.sourceforge.net |
| 01:21:39 | RedBeard | any ideas? |
| 01:22:25 | Aelhaz | I just downloaded debs and dpkg -i'd them |
| 01:23:20 | RedBeard | al: okay |
| 01:23:43 | RedBeard | was following the instructions... |
| 01:24:05 | RedBeard | I'll send frank a note |
| 01:25:33 | Aelhaz | iirc, openssh only needs rsa, blowfish, and IDEA ciphers. someone correct me here |
| 01:25:43 | Aelhaz | openssl includes like eight that openssh doesn't use |
| 01:26:57 | Russ|werk | Aelhaz: souds right |
| 01:27:08 | Russ|werk | if you are running it as sshd, you really only need one of them |
| 01:27:31 | Russ|werk | it might want tripledes too |
| 01:28:54 | Aelhaz | yeah |
| 01:31:12 | Aelhaz | 3des and blowfish for the datastream, IDEA and rsa may be needed for authentication |
| 01:31:18 | Aelhaz | looks like, at least |
| 01:31:44 | Aelhaz | wait, no, IDEA is out due to patent problems |
| 01:32:08 | Aelhaz | so, 3des, blowfish, and rsa |
| 01:33:25 | Aelhaz | and it looks like maybe it's possible to build it w/o ssh v1 compatibility, which would obviate the need for RSA |
| 01:33:42 | Aelhaz | ssh v1 is for suckers anyhow |
| 01:48:48 | Aelhaz | hmmm, openssl's make system bombs out if i disable ciphers. wtf is up with that . . . |
| 02:10:51 | Aelhaz | OK, openssl builds libcrypto.a and bombs out trying to make libssl.a, I don't think i need libssl.a for openssh |
| 02:11:59 | Aelhaz | openssh bombs out with some error about uclibc being built w/o large file support |
| 03:07:17 | | andersee was last seen on #tuxscreen 18 hours, 25 minutes and 14 seconds ago, saying: read the make info page on "The Two Flavors of Variables" [Fri Mar 22 08:42:03 2002] |
| 03:07:17 | jacques | ibot: seen andersee |
| 06:00:59 | foofoo | $99 is too expensive. I'll give you $25 for it. |
| 07:25:02 | dwart | lo all |
| 07:57:34 | | andersee was last seen on #tuxscreen 23 hours, 15 minutes and 31 seconds ago, saying: read the make info page on "The Two Flavors of Variables" [Fri Mar 22 08:42:03 2002] |
| 07:57:34 | jacques | ibot: seen andersee |
| 13:52:43 | Morn | is away: pretending to be a fish... |
| 14:46:48 | prpplague | morning all |
| 16:36:41 | prpplague | man its quite today |
| 17:06:07 | scanline | HELLO |
| 17:06:10 | scanline | :) |
| 17:41:33 | prpplague | scanline: howdy |
| 17:41:40 | scanline | hiya prpplague |
| 17:41:45 | scanline | 3 of you here today? |
| 17:41:58 | prpplague | scanline: know any good ntp servers? |
| 17:42:08 | scanline | time.nist.gov usually works |
| 17:43:20 | prpplague | scanline: bingo thanks |
| 18:04:09 | scanline | is away: Zzzz... |
| 18:14:30 | | prpplague was last seen on #tuxscreen 31 minutes and 10 seconds ago, saying: scanline: bingo thanks [Sat Mar 23 17:43:20 2002] |
| 18:14:30 | TomW | ibot: seen prpplague |
| 18:14:42 | TomW | prpplague: what's happening? |
| 18:18:30 | prpplague | TomW: not much |
| 18:18:42 | prpplague | TomW: got my maxim samples in this week |
| 18:18:52 | prpplague | TomW: but not had time to do anything with them |
| 18:19:01 | prpplague | TomW: i'm hoping for some time tomorow |
| 18:19:19 | prpplague | TomW: hows your stuff doing? |
| 18:19:33 | TomW | prpplague: I am watching my Tux board spitting out PCMCIA slot messages: "CS asked for 0v, still applying 3.3v"... endlessly. |
| 18:19:44 | prpplague | TomW: doh |
| 18:20:00 | TomW | prpplague: Maybe I should put a card in the slot? |
| 18:20:42 | prpplague | TomW: hmm i wonder why it still thinks its powered up? |
| 18:20:52 | TomW | prpplague: I have been busy with other stuff lately, but I am now coming around to getting some free time to mess with the Tux again. |
| 18:21:53 | TomW | prpplague: This is a "stripped" kernel. I removed a lot of driver code to hardware that is no longer connected to the CPU card. I guess it cannot contact some driver code, so it spits out errors. |
| 18:23:28 | prpplague | funn |
| 18:23:36 | TomW | prpplague: if I put the CF card in, then it sits there happy with itself. <sigh> |
| 18:23:48 | prpplague | TomW: lol |
| 18:24:21 | prpplague | TomW: i found some nice switched regulators from maxim as well as a nice lcd bias generator |
| 18:25:31 | TomW | prpplague: seen them. I stay away from maxim unless I absolutely have to use them. Look at Linear Technologies stuff, they have a lot of very good switchers and all their stuff seems to be production quantities. |
| 18:26:32 | TomW | prpplague: Maxim has a HUGE porfolio of products, but it seems that most are not full prodcution. E.G. the MAX727 & MAX728 -- 3.3 & 5 volt 1.5 amp switchers. |
| 18:27:39 | prpplague | TomW: ahh, i'm going to use the 739 and 740 |
| 18:27:48 | TomW | prpplague: If you look to my EZ328LCD design, I used the LT1372 to produce the +24v supply to bias the LCD. |
| 18:28:03 | prpplague | TomW: i'll check it out |
| 18:28:32 | prpplague | TomW: have do done much battery operated devices? |
| 18:28:44 | prpplague | s/do/you |
| 18:29:51 | TomW | prpplague: not handheld stuff, most of the work that I have done has been equipment powered by 7AH 12v sealed lead-acid cells. We have to make our stuff run for 24..48 hours on those cells. |
| 18:31:35 | prpplague | TomW: gotcha |
| 18:32:30 | prpplague | TomW: i also got some samples of the mx3232, but i guess i should have gotten the max3222 since it has the battery monitor |
| 18:32:46 | prpplague | TomW: the 739 and 740 have have batttery monitor circuits as well |
| 18:34:03 | TomW | prpplague: one trick you might try with Maxim, call them up and ask them for pricing, availability & lead times for the device that you want to use in 1000 lot quantities. |
| 18:35:36 | prpplague | TomW: well this is mainly a learning project so i'm not too concernced with quantities right now |
| 18:35:43 | TomW | prpplague: I orginally designed a power supply using the MAX738, got 3 samples, but when I called to get 20 peices to build some prototypes, they said they had no stock and no date for production. |
| 18:35:54 | prpplague | doh |
| 18:36:01 | prpplague | that is a problem |
| 18:36:32 | prpplague | lt have a sample program? |
| 18:37:01 | TomW | prpplague: They did at one time, I think that they went to a small lot sales program now. |
| 18:38:15 | TomW | prpplague: Another way I keep to easily available components is to rely upon a few large distributors: DigiKey & Arrow Electronics. If the product is not listed with either of these vendors, I don't use it. |
| 18:39:00 | prpplague | TomW: gotcha |
| 18:39:52 | TomW | prpplague: BTW, EVERY(!) manufacturer has a sample program of some kind. Remember, it is in their self-interest to ensure that design engineers have access to their products so that we can design them in. |
| 18:41:58 | TomW | prpplague: Some manufacturers don't deal directly with the "little guy", but of these, most have their components available through distributors. Sometimes, you have to build a relationship / track record with the distributor before they will do something like cut 100 peices off a 10K tape reel for you. |
| 18:42:11 | prpplague | TomW: you'd think that, but i've run into companys that don't care about prototyping |
| 18:43:23 | TomW | prpplague: For small run stuff, I have contacted, say, Molex and got 20 samples of a connector so I could satisfy my production run. |
| 18:44:22 | TomW | prpplague: some of the FPC connectors are only available via tape reels. |
| 18:45:29 | prpplague | TomW: my main problem is that i'm worried about getting board level integration of the processor,ram and flash |
| 18:45:46 | prpplague | TomW: i've got to find pcb company that can do it reasonably |
| 18:45:47 | TomW | prpplague: Ok, the Tux CPU board is now bitching about not having a PCMCIA card in the second slot... |
| 18:45:53 | prpplague | doh |
| 18:47:43 | TomW | prpplague: Look to Advanced Circuits (http://www.4pcb.com), they do a three day turn and their work is top-notch! Their pricing on prototype boards is very resonable. When you run their quote calculator, try different values for the proto quantities, sometimes you can get a LOT more boards for only a few dollars more. |
| 18:48:48 | prpplague | TomW: they do the soldering as well |
| 18:48:49 | prpplague | ? |
| 18:48:59 | TomW | prpplague: They panelize their boards, and I think that they use an 48x48 inch panel. |
| 19:40:18 | prpplague | man its quite today |
| 19:40:46 | TomW | prpplague: cul. |
| 19:41:08 | TomW | is away: to Mom's for dinner. :-) |
| 21:56:24 | prpplague | andersee: howdy |
| 21:59:58 | andersee | hey |
| 22:00:58 | andersee | welcome back |
| 22:01:30 | prpplague | thanks, man a c-n-p error |
| 22:01:36 | prpplague | s/man/made |
| 22:01:54 | andersee | oops |
| 22:03:54 | prpplague | andersee: how goes it? |
| 22:04:06 | prpplague | andersee: you planning on going to ottowa? |
| 22:04:16 | andersee | ottowa? |
| 22:04:33 | andersee | Any reason I should plan on it? |
| 22:05:08 | prpplague | andersee: it was just my understand that alot of the kernel gurus usually attend |
| 22:05:27 | andersee | attend what? |
| 22:06:14 | prpplague | ottowa linux symposium |
| 22:06:23 | andersee | Ahh |
| 22:07:01 | andersee | I've never been. Sounds like fun, but I expect it would cost like $1500 or so for me to attend... |
| 22:08:40 | prpplague | andersee: submit you a paper or something, maybe some one will pay for your expenses |
| 22:09:51 | andersee | Hmm, theres and idea |
| 22:11:07 | andersee | Submission of paper proposals is already closed... |
| 22:12:20 | prpplague | doh |
| 23:04:51 | scanline | is back (gone 05:00:42) |