| 00:11.50 | *** part/#uclibc nitinkg (n=ngupta@gateway-1237.mvista.com) |
| 00:36.15 | *** join/#uclibc Sgt-Donan (n=Loutre@feu30-1-82-242-59-35.fbx.proxad.net) |
| 00:56.00 | *** join/#uclibc angom_h (n=angom@red-corp-200.79.145.199.telnor.net) |
| 01:06.43 | *** join/#uclibc thraxisp (n=thraxisp@ottawa-hs-206-191-39-102.d-ip.magma.ca) |
| 01:28.33 | CIA-12 | 03gkajmowi * r14652 10uClibc++/ (ChangeLog Rules.mak include/limits include/map src/ios.cpp): (log message trimmed) |
| 01:28.33 | CIA-12 | - Clean up double-semicolons |
| 01:28.33 | CIA-12 | - Add parent class scoping as appropriate |
| 01:32.31 | *** join/#uclibc landley (n=landley@pool-141-151-176-146.pitt.east.verizon.net) |
| 01:33.21 | landley | So I'd decided that I'd just clean up udhcpc to be a good busybox app, actually using bb_getopt_ulflags() and so on, and just check in the changes. |
| 01:33.38 | landley | But A) my checking is still giving me an empty directory for networking/udhcp |
| 01:34.04 | landley | and B), I actually looked at the code, and there's rather significant cleanup opportunities to the point I'm ripping half of it out. |
| 01:34.31 | landley | Still working based on Russ Dill's version though, and he's welcome to maintain the result in busybox is he likes. |
| 01:37.07 | landley | Among other things, by all means call a script if it's there, but if it's not just call our built-in ifconfig functionality (which busybox _has_) to bring the interface up. |
| 01:37.18 | landley | We shouldn't _need_ a script, we have ifconfig built into busybox, and it can use it. |
| 01:42.49 | dalias | :) |
| 01:42.54 | dalias | btw |
| 01:43.14 | landley | ? |
| 01:43.35 | dalias | there's a philosophy of privilege separation |
| 01:44.03 | dalias | whereby the dhcp client should run as uid=dhcp or such and just communicate the address it gets to a separate process with root perms to run ifconfig :) |
| 01:44.31 | landley | It needs root permissions to put the interface into promiscuous mode. |
| 01:44.44 | dalias | yes but thats setup |
| 01:44.51 | dalias | it hasn't processed any input at that point |
| 01:45.07 | landley | So you need root permissions, then you need root permissions again, and you believe we should spend the bytes to fork and then call ourselves. |
| 01:45.08 | landley | I don't. |
| 01:45.11 | dalias | so it can drop all permissions before processing any input and thus be secure |
| 01:45.22 | landley | Or you can be secure by processing input intelligently and carefully. |
| 01:45.28 | dalias | i dont mean to modify udhcp that way |
| 01:45.33 | landley | You're saying the root process on the other end of the pipe _isn't_ processing input? |
| 01:45.34 | dalias | i mean you can do that with the script setup |
| 01:45.46 | landley | Calling the script should be optional. |
| 01:45.50 | dalias | that's the motivation for supporting using a script instead of directly configuring the interface |
| 01:45.52 | landley | As in CONFIG style optional. |
| 01:45.53 | dalias | yeah |
| 01:45.58 | dalias | optional is good |
| 01:46.22 | landley | No, the motivation for supporting a script instead of configuring the interface yourself is so that you don't need semi-portable interface code in something produced by a standards body. |
| 01:46.51 | landley | One of the biggest mistakes Berkeley ever made is the lack of /dev/eth0. |
| 01:47.01 | landley | It's a char device by nature. |
| 01:47.41 | landley | But busybox already contains ifconfig. |
| 01:48.08 | landley | We can split that code out into libbb (or networking.a) and re-use it. |
| 01:48.47 | landley | Heck, run_applet_by_name()... |
| 01:48.55 | landley | But reusing the code directly might be smaller. Dunno yet. |
| 01:48.56 | dalias | yes |
| 01:48.58 | landley | Leaving my options open still. |
| 01:49.01 | dalias | lack of /dev/eth0 is a huge mistake |
| 01:49.09 | dalias | why has linux never fucking fixed it? |
| 01:49.30 | landley | There's an existing standard. |
| 01:49.35 | landley | The existing standard is libc. |
| 01:49.39 | dalias | no there's not |
| 01:49.49 | landley | I didn't say it was a _complete_ standard. |
| 01:49.51 | dalias | configuring interfaces does not appear in susv3 at all |
| 01:49.56 | dalias | its totally omitted |
| 01:49.58 | landley | But for opening a socket on an already configured interface... |
| 01:50.08 | landley | Linux does have /dev/eth0, I think. It's optional. |
| 01:50.18 | landley | (It's one of those weird options like netlink.) |
| 01:50.24 | landley | (Nobody uses it because it's not always there.) |
| 01:50.28 | dalias | a socket isn't "on an interface" |
| 01:50.38 | dalias | opening sockets should require /dev/tcp, /dev/udp, etc |
| 01:50.45 | dalias | that would be a HUGE help for security |
| 01:50.53 | landley | *shrug* |
| 01:51.09 | dalias | because chroot environments could not create sockets unless you gave them the right device nodes |
| 01:51.33 | landley | Code up a kernel patch, and a patch to uclibc to use it. |
| 01:51.36 | dalias | thus preventing someone who exploits a chrooted program from being able to launch spam/attacks/etc. from it |
| 01:51.44 | landley | Yup. |
| 01:51.46 | landley | Agreed. |
| 01:51.49 | dalias | the same can be done with 'capabilities' |
| 01:51.53 | landley | The unix security model works pretty darn well. |
| 01:52.00 | landley | Nobody uses capabilities outside of the kernel itself. |
| 01:52.05 | dalias | but despite linux calling them "posix capabilities' to make them sound legit, they were rejected in committee |
| 01:52.16 | dalias | because they suck |
| 01:52.34 | landley | If you can't keep rwx straight, adding another 2^x potential combinations of priviledges isn't going to make things _more_ manageable. |
| 01:53.10 | landley | I've ignored most of the selinux stuff. |
| 01:53.18 | dalias | yep.. |
| 01:53.20 | dalias | me too |
| 01:53.51 | dalias | the only security feature i use is disallowing creation of links to files you don't own in world-writable dirs |
| 01:54.05 | dalias | because i can't be bothered to audit every single program for /tmp vulns |
| 01:55.56 | landley | Oh ick. |
| 01:55.58 | landley | Just ick. |
| 01:56.07 | landley | static int listen_mode; is a global. |
| 01:56.24 | landley | static void change_mode(int new_mode) |
| 01:56.24 | landley | { |
| 01:56.24 | landley | <PROTECTED> |
| 01:56.24 | landley | <PROTECTED> |
| 01:56.24 | landley | <PROTECTED> |
| 01:56.39 | landley | <PROTECTED> |
| 01:56.39 | landley | <PROTECTED> |
| 01:56.39 | landley | } |
| 01:57.04 | landley | This is from dhcpc.c. |
| 01:57.07 | landley | I presume fd is another global. |
| 01:57.28 | landley | I'm sorry, this applet _dies_. |
| 01:57.47 | dalias | ... |
| 01:57.53 | dalias | :) |
| 01:58.03 | dalias | all the bb code is like that |
| 01:58.09 | dalias | globals all over the place |
| 01:58.11 | landley | Uh, no. |
| 01:58.17 | landley | Globals yes. |
| 01:58.21 | landley | Globals named "fd"? |
| 01:58.27 | dalias | if its static, no problem |
| 01:58.42 | landley | And a wrapper to set the value of an "int"? |
| 01:58.50 | dalias | :) |
| 01:59.04 | dalias | it also logs |
| 01:59.04 | landley | yes it's a problem, it's 4 bytes allocated in every app and prezeroed, which means that on nommu platforms it uses real memory in every busybox instance. |
| 01:59.11 | landley | Gratuitously. |
| 01:59.17 | dalias | yep |
| 01:59.23 | landley | It's not logging that it received or sent a packet, which is the interesting level of action. |
| 01:59.36 | landley | It's not logging that it _did_ anything, it's logging that it toggled an internal variable's state. |
| 01:59.42 | dalias | :) |
| 02:00.19 | dalias | working on libc has taught me a lot about memory usage, the evil of relocations and global vars, etc. |
| 02:00.59 | dalias | and the evils of heap allocation, unbounded memory use, ... |
| 02:26.01 | psm | dalias: do some apps work w/ your getopt/getopt_long combo? I am unable to replace the gnu one on my uClibc box |
| 02:47.19 | dalias | yes, all the busybox applets work |
| 02:48.07 | psm | have you tried anything else using getopt_long? |
| 02:48.09 | dalias | i first tested it with busybox actually. that was my motivation for writing it |
| 02:48.18 | dalias | any apps other than bb you mean? |
| 02:48.24 | psm | yes |
| 02:48.57 | dalias | no. until i have a more complete system to test on, there aren't too many apps i feel like building |
| 02:49.23 | dalias | since i'd need lots of libs compiled against my libc |
| 02:49.31 | CIA-12 | 03landley * r14653 10busybox/modutils/insmod.c: Stephane Billiart found an endianness bug in insmod. |
| 02:49.33 | dalias | so far i just have ncurses :) |
| 02:53.22 | psm | try nano/texinfo, those use getopt_long and ncurses |
| 02:54.04 | dalias | i could try nano.. |
| 02:56.53 | dalias | wtf |
| 02:56.56 | dalias | nano uses glib?? |
| 03:02.01 | CIA-12 | 03landley * r14654 10busybox/include/platform.h: Define both big endian and little endian macros. |
| 03:05.01 | dalias | psm, seems to work fine |
| 03:06.08 | dalias | and.. very cool, my utf8 support works in nano just great! |
| 03:06.27 | dalias | let's see if nano correctly handles combining chars.. |
| 03:09.24 | dalias | wow it does! |
| 03:11.27 | dalias | i'm impressed that such a crappy editor actually handles it right |
| 03:15.45 | CIA-12 | 03psm * r14655 10buildroot/toolchain/gcc/ (4.0.3/304-index_macro.patch 4.1.0/304-index_macro.patch): Compile even if index is a macro |
| 03:15.53 | *** join/#uclibc thraxisp (n=thraxisp@ottawa-hs-206-191-39-102.d-ip.magma.ca) |
| 03:26.48 | *** join/#uclibc sjhill (n=sjhill@eth13.com-link.com) |
| 03:53.10 | ds | sjhill: can you convey any secrets on how to build a mips-uclibc-nptl toolchain? |
| 03:59.07 | sjhill | not without you paying me |
| 03:59.14 | sjhill | mhuwhahahaha |
| 03:59.21 | dalias | :) |
| 03:59.32 | dalias | btw |
| 03:59.39 | dalias | how much of nptl is kernelspace, and how much is userspace? |
| 03:59.43 | sjhill | ds: what are you having problems with? |
| 04:00.17 | sjhill | dalias: dunno' ... maybe 25K/75U |
| 04:01.11 | dalias | i'm just curious if it would be possible to make a non-2.6-dependent pthread lib with nptl's approach plus a few hacks |
| 04:01.33 | ds | sjhill: I was getting weird compile errors like uclibc was just missing source files and such |
| 04:02.01 | sjhill | ds: my trunk is not buildable currently |
| 04:02.02 | ds | having hacked past that, any programs compiled with it segfault |
| 04:02.11 | sjhill | correct |
| 04:02.12 | ds | ok, that makes sense |
| 04:02.22 | sjhill | in about 2 months, everything will be in there |
| 04:02.39 | ds | so I'm currently using a snapshot from 2/18 |
| 04:03.13 | sjhill | doesn't matter what you use |
| 04:03.23 | sjhill | i still have all of the code needed for it to work |
| 04:03.26 | ds | ah, ok |
| 04:03.38 | ds | er |
| 04:03.44 | ds | "have" or "don't have"? |
| 04:04.04 | sjhill | everything works for NPTL and uClibc |
| 04:04.07 | sjhill | for MIPS |
| 04:04.25 | sjhill | i'm writing my paper for OLS and getting ready for a big merge of code |
| 04:04.25 | ds | ok, that's what I thought from your emails |
| 04:04.58 | ds | I guess the best question is: what versions of tools should I ideally be using? |
| 04:05.17 | sjhill | binutils-2.16.1 or newer |
| 04:05.25 | sjhill | gcc-4.1.0 released or newer |
| 04:05.38 | sjhill | kernel headers 2.6.13 or newer |
| 04:06.15 | ds | what about a mips-linux snapshot from novemberish? does that matter? |
| 04:07.20 | dalias | bleh never mind, i dont think i want to read nptl, much less use anything from it... |
| 04:07.27 | dalias | its full of nasty gcc/elf/etc hacks |
| 04:08.58 | sjhill | ds: you mean mips-linux kernel? |
| 04:09.22 | ds | sjhill: yes |
| 04:09.40 | sjhill | might be okay |
| 04:09.51 | sjhill | i'm using 2.6.15 tarball from www.linux-mips.org |
| 04:10.55 | ds | ok, that's what I'm using |
| 04:11.09 | ds | so I'll just slog through and debug it |
| 04:11.16 | sjhill | debug what? |
| 04:11.24 | ds | why I'm getting segfaults |
| 04:11.29 | sjhill | *sigh* |
| 04:11.34 | sjhill | you're wasting your time |
| 04:11.47 | sjhill | i have the 2000+ lines of patches needed to make the dynamic loader work |
| 04:12.05 | sjhill | or are you talking just normal binaries? |
| 04:12.18 | ds | both, actually |
| 04:12.22 | sjhill | forget NPTL |
| 04:12.33 | sjhill | the normal stuff, that is a problem |
| 04:13.07 | ds | are you saying that you have stuff that's not checked into the nptl branch? |
| 04:13.14 | sjhill | yes. |
| 04:13.17 | ds | ah, ok |
| 04:13.21 | ds | i missed that part |
| 04:13.24 | ds | sorry |
| 04:13.37 | sjhill | np |
| 04:13.44 | sjhill | i thought i was being clear, but guess not |
| 04:14.57 | ds | I can't say that I tested static binaries too thoroughly |
| 04:15.19 | ds | so I may have just done something blatantly stupid |
| 04:15.28 | sjhill | static NPTL binaries will absolutely not work |
| 04:15.34 | sjhill | i haven't even debugged them |
| 04:15.40 | sjhill | they segfault all over the place |
| 04:16.10 | dalias | it looks like i'm going to have to write my own pthread implementation.. |
| 04:16.14 | dalias | maybe uclibc can use it once i'm done |
| 04:16.22 | dalias | imo nptl is hopelessly broken |
| 04:16.27 | sjhill | you are an idiot |
| 04:16.42 | dalias | um, why do you say that? |
| 04:16.54 | sjhill | the reason the code is not checked in is because i did it under contract |
| 04:16.59 | sjhill | once i get paid, the code gets released |
| 04:17.02 | sjhill | which is in two months |
| 04:17.12 | sjhill | impatient aren't we? |
| 04:17.18 | dalias | no |
| 04:17.22 | dalias | i just doubt it will work |
| 04:17.35 | dalias | since it's originally written by drepper :) |
| 04:17.48 | sjhill | it already works and passes over 5000 tests |
| 04:17.55 | sjhill | get bent |
| 04:17.58 | dalias | sjhill, thats not what i mean by work |
| 04:18.12 | dalias | by work i mean independent of gcc extensions and elf dynamic linking semantics |
| 04:18.33 | dalias | ds, nope |
| 04:19.25 | dalias | i'm ircing from a computer using glibc 2.1 which will be using my implementation in a few months once i finish it. the libc is basically done but i want dynamic linking and so i have to write a linker.. :( |
| 04:19.43 | sjhill | don't plan on your code getting into uclibc |
| 04:19.50 | dalias | i'm not trying to |
| 04:19.54 | sjhill | good |
| 04:20.02 | dalias | but psm seems to like doing so |
| 04:20.27 | sjhill | doing what? |
| 04:20.33 | dalias | putting my code in uclibc |
| 04:20.58 | *** join/#uclibc kergoth_ (n=kergoth@c-24-118-213-150.hsd1.mn.comcast.net) |
| 04:21.59 | sjhill | hey kergoth_ |
| 04:22.16 | sjhill | as long as it doesn't break my nptl, i don't care |
| 04:22.25 | sjhill | psm is good at breaking my builds all the time |
| 04:22.33 | sjhill | i want him moved into a branch and not on the trunk |
| 04:22.51 | dalias | heh |
| 04:22.54 | dalias | all of it's optional |
| 04:22.59 | sjhill | i don't care |
| 04:23.03 | dalias | small non-bloated functions replacing the imported gnu code |
| 04:23.16 | sjhill | he's optional stuff breaks things because he doesn't test thoroughly and it pisses me off |
| 04:23.17 | mjn3 | sjhill: i personally consider current svn to _be_ psm's branch. which is why i've never checked anything into it |
| 04:23.31 | sjhill | mjn3: oh Lord |
| 04:23.40 | dalias | hehe |
| 04:23.42 | dalias | hi mjn3 |
| 04:23.46 | mjn3 | hey |
| 04:24.11 | sjhill | mjn3: and why is andersee letting him do that? |
| 04:24.38 | mjn3 | sjhill: fyi... i'll likely fix the locale patches for gcc 4.x in buildroot this weekend |
| 04:25.16 | sjhill | mjn3: thank you thank you |
| 04:25.30 | mjn3 | i guess andersee has given out lots (in my mind) of commit access recently because he doesn't have time to work on things |
| 04:27.19 | mjn3 | not like i have much time for things either lately. although i do have some fixes that will be made available whenever we release code linked with them. someone else would need to port them to svn since i'm basing things off of .28 |
| 04:27.46 | mjn3 | neither erik or i are using svn for any "work" stuff. to unstable |
| 04:28.16 | dalias | hmmm... very interesting |
| 04:28.36 | dalias | according to the docs i've found, mozilla does not even need pthread. :) |
| 04:28.48 | dalias | so maybe i can just skip thread support entirely :) |
| 04:29.28 | ds | sjhill: btw thanks. I owe you at least a beer at OLS |
| 04:30.39 | sjhill | ds: much appreciated |
| 04:32.42 | mjn3 | sjhill: which gcc version is your preference? i'll make sure i catch that one |
| 04:33.16 | sjhill | gcc-4.1.0 released |
| 04:33.23 | mjn3 | ok |
| 04:33.29 | sjhill | excellent |
| 04:34.06 | dalias | anyway since you know it well, i guess i can ask: |
| 04:34.24 | dalias | is there any chance that nptl can be made portable to non-gcc and non-elf environments? |
| 04:34.47 | dalias | knowing drepper my guess is no :( |
| 04:35.38 | mjn3 | while i've had a run-in or two with drepper, you can't blame him for coding to the platform he's paid to support |
| 04:36.08 | dalias | well, imo you kinda can |
| 04:36.20 | sjhill | not you can't |
| 04:36.41 | sjhill | dalias: have you ever maintained anything like glibc or uclibc? |
| 04:36.55 | mjn3 | there are lots of things i'd do differently myself if i didn't have commitments, deadlines, etc |
| 04:37.11 | dalias | i blame him for making linux systems slow and bloated and for making software at large less portable because of all the glibc and gcc extensions available |
| 04:37.27 | sjhill | that's why uclibc exists |
| 04:37.32 | sjhill | we get rid of the bloat |
| 04:37.49 | dalias | and yes, i have |
| 04:38.15 | dalias | i'm one of the main mplayer maintainers, and i also have my own libc implementation that's almost done |
| 04:38.18 | sjhill | mjn3: remind me to discourage andersee from giving dalias commit access |
| 04:38.24 | dalias | of course in a way writing from scratch is easier than maintaining |
| 04:38.33 | dalias | i don't want commit access |
| 04:41.14 | dalias | i have commit access to too much already :) |
| 05:09.17 | ashes | i just received my new work boots. they're dupont nomex over kevlar over gore-tex over kevlar over cotton, with stainless steel sole, toe, and metatarsal guard. im wondering if a silicone waterproofer is good to put on the nomex, or if it will damage it |
| 05:09.49 | sjhill | ashes: i don't know, but it has nothing to do with #uclibc, so go away |
| 05:10.03 | dalias | haha |
| 05:10.05 | ashes | everything has to do with uclibc |
| 05:10.18 | dalias | sjhill, he's going to use uclibc on the embedded system inside his new boot design |
| 05:10.31 | ashes | yes |
| 05:10.56 | ashes | uclibc in the operating system for reflex ballistic plastic |
| 05:11.32 | dalias | ;) |
| 05:11.34 | ashes | so there will be no risk of shooting myself in the foot |
| 05:12.16 | ashes | google isn't telling me if silicone waterproofer damages nomex |
| 05:13.42 | dalias | well you'd better learn if you want to design a better boot :) |
| 05:14.07 | ashes | i dont like dupont. dupont wont pull a product for public health reasons unless the gov't orders them to, even when it is known to kill people (teflon), however they do make some of the best stuff |
| 05:14.19 | dalias | yep.. |
| 05:14.37 | dalias | do you know the history of freon and the 'ozone hole'? |
| 05:15.03 | ashes | i dont think so |
| 05:16.16 | dalias | well the ozone issue surfaced right around the time that the patent on freon was about to expire |
| 05:16.40 | dalias | so conveniently freon got banned in new cooling systems right when it would have become free to everyone else |
| 05:17.00 | dalias | and dupont released new patented, extremely-toxic chemicals for coolants |
| 05:17.01 | ashes | drug companies do the same thing |
| 05:17.22 | ashes | they're playing the rules |
| 05:17.32 | dalias | the most amusing part is... |
| 05:17.47 | dalias | the best known coolant is simple propane or butane |
| 05:18.02 | ashes | the best frying pan is cast iron |
| 05:18.25 | ashes | cast iron is more stick resistant than teflon |
| 05:18.29 | dalias | but due to not being able to patent that, they got it banned by publishing propaganda that it was an explosion hazard |
| 05:18.33 | dalias | yep |
| 05:18.42 | dalias | cast iron is great if you take care of it correctly |
| 05:19.13 | ashes | it takes a while to learn how to care for cast iron, but after learning its easy |
| 05:20.00 | ashes | i wash my pan with a tiny bit of soap, and just use oil next time, and its always slick |
| 05:21.32 | ashes | teflon releases nerve gas after something like 280C. put the burner on max for 3 minutes with an empty teflon pan, and the pan will reach 280C |
| 05:21.48 | dalias | :) |
| 05:21.51 | ashes | also, most teflon is not genuine teflon |
| 05:22.24 | ashes | genuine dupont teflon is the safest, but imitation teflon is must crappier |
| 05:22.54 | dalias | yes thats the problem with big corrupt corporations. when the small ones come to challenge them, the small ones are even more corrupt |
| 05:23.12 | dalias | because they have to resort to doing bad things to be competitive with the existing monopoly |
| 05:23.47 | ashes | all synthetic non-stick coatings are toxic at high heat |
| 05:24.04 | ashes | even porcelain often contains lead |
| 05:24.10 | ashes | for some reason |
| 05:24.32 | dalias | i think we pissed sjhill off ;) |
| 05:25.56 | ashes | for pots, stainless steel is best. stainless steel is high in nickel, and nickel can cause health problems, but fatalities are very rare. as long as its not used for a frying pan which gets scratched a lot, its fairly safe |
| 05:26.28 | ashes | porcelain made in the usa or canada won't contain lead |
| 05:26.34 | dalias | hey |
| 05:26.42 | ashes | but its usually not made in the usa or canada |
| 05:26.54 | dalias | if i have a nonstick coated wok and the surface is chipping off |
| 05:27.06 | dalias | do you know a good way to remove it entirely and make it into a plain steel one? |
| 05:27.19 | ashes | i have thrown out all my non-stick pans |
| 05:27.36 | ashes | sandblast it |
| 05:27.44 | ashes | but its better to buy a new pan |
| 05:27.52 | ashes | copper can be toxic too |
| 05:27.58 | ashes | more than nickel |
| 05:29.03 | ashes | stainless steel pots, cast iron frying pans, and cast iron bakeware or made-in-use porcelain bakeware, is the best you can do |
| 05:29.06 | ashes | or |
| 05:29.32 | ashes | titanium. titanium has no known toxicity, however its not non-stick |
| 05:29.56 | ashes | made-in-usa, even |
| 05:30.34 | ashes | titanium cookware is fairly expensive too. it costs more than copper |
| 05:31.17 | dalias | i can't just kill the surface with high temperature or chemicals? :) |
| 05:31.29 | dalias | (outdoors obviously) |
| 05:32.07 | ashes | remove the plastic handle and put it in a self cleaning oven, however the gasses coming out of the oven during the self clean (with a teflon pan inside) will be very toxic |
| 05:32.25 | dalias | the handle is wood, not plastic. :) |
| 05:32.56 | dalias | will that burn off the teflon entirely? |
| 05:33.31 | ashes | i dont suggest putting a teflon pan in a self cleaning oven, unless you leave all the windows open and leave the house until 6 hours after the self-cleaning has finished |
| 05:33.55 | dalias | ashes, a modern one has an exhaust fan |
| 05:33.55 | ashes | it should. teflon is not made for the heat from self cleaning |
| 05:34.09 | ashes | but i dont suggest it |
| 05:34.18 | dalias | anyway i dont have a self-cleaning oven |
| 05:34.22 | ashes | cast iron pans are not expensive |
| 05:34.32 | dalias | its not that its expensive |
| 05:34.34 | ashes | used ones work just as well too |
| 05:34.37 | dalias | its that i have something nonusable right now |
| 05:34.42 | dalias | and i'd like to make it usable |
| 05:34.46 | ashes | toss it out |
| 05:34.54 | ashes | its contaminated |
| 05:34.55 | dalias | i have another wok i can use too |
| 05:35.02 | dalias | but it can be detoxed ;) |
| 05:35.05 | ashes | they make cast iron woks |
| 05:35.16 | ashes | and stainless |
| 05:35.19 | dalias | yes |
| 05:35.21 | dalias | i have a stainless one |
| 05:35.27 | dalias | its very nice |
| 05:35.40 | ashes | but stainless is not pourus. food will always stick to it |
| 05:36.03 | dalias | not if you season it well |
| 05:37.34 | ashes | i dont think stainless steel can be seasoned |
| 05:37.51 | ashes | also |
| 05:38.04 | ashes | use peanut oil whenever possible |
| 05:38.15 | ashes | it has the highest smoking tempeture |
| 05:38.52 | ashes | i would love to open a safety clothing company |
| 05:38.52 | dalias | well i'm not positive its stainless, but thats my guess |
| 05:40.16 | ashes | berylium nickel alloy toe capped boots, moulded in one peice with the sole plate.. 80,000 pount capacity |
| 05:40.53 | ashes | spectra stitches |
| 05:41.15 | ashes | folded seams |
| 05:42.05 | ashes | it violated gov't regulations to advertize the tested capacity of steel toed boots |
| 05:42.46 | dalias | haha |
| 05:43.06 | ashes | they can only say 'our boots meet or exceed ansi' |
| 05:43.43 | ashes | im not sure why |
| 05:43.55 | ashes | i suspect its to keep a level playing field |
| 05:44.34 | ashes | and also, if someone advertized their boots are twice as strong as gov't regulations, it would undermine the govt regulations |
| 05:45.21 | ashes | every year thousands of people have steel toes fail |
| 05:45.24 | ashes | world wine |
| 05:46.17 | ashes | the govt regulations aren't good enough |
| 05:48.06 | ashes | the best steel toed boots in the world are rated for about 4990 pounds. in independent tests, most boots start to fail around 6000 pounds |
| 05:49.12 | dalias | btw do you work on uclibc? |
| 05:49.19 | dalias | or just talk about boots? |
| 05:49.19 | ashes | i use it |
| 05:49.22 | dalias | ah |
| 05:49.34 | ashes | i have submitted patches |
| 05:50.01 | ashes | such as the ar4random patch. i have a couple more patched to make to get mktemp to use arc4random |
| 05:50.11 | ashes | and propolice |
| 05:50.57 | dalias | ah my mktemp is really lame :) |
| 05:51.01 | dalias | it just uses sequential numbers |
| 05:51.52 | ashes | the arc4random function will produce a cryptographicall random sequence even if /dev/urandom is no where to be found |
| 05:52.09 | ashes | using the pid and date |
| 05:53.51 | ashes | its a library for any non-cryto program that needs random numbers |
| 05:54.18 | dalias | well pid and date are known, so it's not cryptographically secure |
| 05:54.27 | ashes | crypto applications need real entropy whether its available or not |
| 05:54.33 | dalias | even if its a secure hash, if the input is known it can't be secure |
| 05:55.07 | ashes | the output is cryptographically random, but not the source entropy |
| 05:55.26 | ashes | it only uses the pid+date if urandom can't be used/found |
| 05:58.16 | ashes | from one milisecond to the next, the output will be completely different. the output can't be reverse engineered to figure out the source entropy, beit the date or urandom |
| 05:59.38 | dalias | however if you know the first random number in the sequence, you can brute force all dates within 1 second very quickly and find the seed that gives the right number |
| 06:01.40 | ashes | i need sleep |
| 06:01.42 | ashes | gnight |
| 06:02.23 | dalias | goodnight |
| 06:11.22 | *** join/#uclibc sjhill (n=sjhill@eth13.com-link.com) |
| 07:31.36 | *** join/#uclibc blindvt (n=bf@M922P016.adsl.highway.telekom.at) |
| 11:57.39 | *** join/#uclibc jacques (n=username@nslu2-linux/jacques) |
| 12:04.09 | *** join/#uclibc FireEgl (i=Atlantic@Atlantica.DollarDNS.Net) |
| 12:13.03 | *** join/#uclibc FireEgl (i=Atlantic@Atlantica.DollarDNS.Net) |
| 12:52.20 | *** join/#uclibc Sgt-Donan (n=Loutre@feu30-1-82-242-59-35.fbx.proxad.net) |
| 15:03.43 | *** join/#uclibc Sgt-Donan (n=Loutre@feu30-1-82-242-59-35.fbx.proxad.net) |
| 15:03.44 | *** join/#uclibc JockeHome (n=JockeHom@84-217-6-182.tn.glocalnet.net) |
| 15:10.58 | *** join/#uclibc blindvt (n=bf@M832P003.adsl.highway.telekom.at) |
| 15:42.16 | *** join/#uclibc sjhill (n=sjhill@eth13.com-link.com) |
| 16:25.05 | *** join/#uclibc prpplague (n=billybob@72.22.159.157) |
| 16:36.24 | *** join/#uclibc prpplague (n=billybob@72.22.159.157) |
| 17:39.16 | *** join/#uclibc andersee (n=andersee@codepoet.org) |
| 17:43.41 | *** join/#uclibc angom_h (n=angom@red-corp-200.79.145.199.telnor.net) |
| 19:08.44 | *** join/#uclibc andersee (n=andersee@codepoet.org) |
| 19:09.12 | CIA-12 | 03andersen * r14656 10buildroot/target/device/AMD/DBAu1500/ (39 files in 2 dirs): update to linux 2.4.16 and current uClibc. |
| 19:30.46 | andersee | SpanKY: around? |
| 19:31.06 | solar | he not around much on the weekends I've noticed |
| 19:32.02 | andersee | looking at current uClibc svn -- looks as if getdents* is broken |
| 19:32.10 | andersee | i.e. make segfaults |
| 19:32.27 | andersee | mjn3 mentioned there might be an existing gentoo patch for such? |
| 19:34.05 | solar | hrmm I do seem to recall there was something up with the readdir() that spanky noticed which probably ment getdents() but he should be fixing stuff in svn and backporting to our tree |
| 19:34.17 | solar | let me check however what we have in the tarball of patches |
| 19:34.32 | andersee | solar: thx |
| 19:35.25 | sjhill | andersee: ah, so you did see the same thing as me |
| 19:35.30 | sjhill | good good |
| 19:35.50 | andersee | sjhill: yep, 'make' segfaulted right off |
| 19:35.54 | sjhill | s/not/now/ |
| 19:36.02 | sjhill | excellent |
| 19:36.17 | andersee | strace suggests getdents is busted |
| 19:36.21 | sjhill | yep |
| 19:36.35 | sjhill | it's what i used through all of my development |
| 19:36.50 | sjhill | you have chosen....wisely |
| 19:37.09 | andersee | esp since sjhill has not committed the rest of his pile of fixes |
| 19:37.16 | andersee | ;-) |
| 19:38.09 | mjn3 | depends on when he gets paid i'd expect |
| 19:39.02 | solar | not seeing anything http://rafb.net/paste/results/ITIxIa49.html |
| 19:40.34 | andersee | hmm |
| 19:40.37 | andersee | solar: thx |
| 19:41.15 | sjhill | mjn3: correct |
| 19:41.27 | mjn3 | andersee, check the mailing list maybe? thought i saw some mention of the problem recently |
| 19:41.29 | sjhill | prpplague: depends on if Broadcom gets their butt in gear |
| 19:41.40 | *** join/#uclibc tty56 (n=johannes@2001:6f8:1331:3:bd1a:494:9473:4e69) |
| 19:42.12 | mjn3 | sjhill, regretably the gear they put it in is usually LOW |
| 19:42.15 | solar | andersee: http://www.gentoo.org/cgi-bin/viewcvs.cgi/src/patchsets/uclibc/0.9.28/48_all_uClibc-svn-getdents.patch?root=gentoo&rev=1.1&view=markup |
| 19:42.23 | solar | but thats a backport from svn |
| 19:44.20 | prpplague | sjhill: hehe |
| 19:44.38 | andersee | solar: thx, will check it out |
| 19:45.35 | solar | that patch has not made it onto any gentoo systems yet. so I cant say if it works or not. |
| 19:46.13 | solar | I do recall that the mdev from busybox is what triggered the bug with getdents not returning proper results in the first place |
| 19:46.48 | sjhill | mjn3: :) |
| 19:49.16 | prpplague | later folks |
| 20:15.53 | sjhill | mjn3, andersee: can you guys read Abiword files? |
| 20:16.12 | mjn3 | yes |
| 20:17.10 | sjhill | mjn3, andersee: please read and review, preferably today |
| 20:17.34 | sjhill | i still have a lot to write, but i want to make sure i've given good foundation and lead in to the really technical discussion |
| 21:55.23 | *** join/#uclibc TMM1 (n=aman@residents-liberty-potomac-129-174-185-149.residents.gmu.edu) |
| 22:22.23 | *** join/#uclibc tty56 (n=johannes@2001:6f8:1331:3:bd1a:494:9473:4e69) |
| 23:10.59 | *** join/#uclibc blindvt_ (n=bf@M871P021.adsl.highway.telekom.at) |
| 23:17.52 | *** join/#uclibc blindvt__ (n=bf@M909P000.adsl.highway.telekom.at) |
| 23:28.07 | *** join/#uclibc TheMasterMind1 (n=aman@residents-liberty-potomac-129-174-185-149.residents.gmu.edu) |