irclog2html for uclibc on 2002.09.17

04:26:16anderseemorning Glenn
04:57:34bug1hi Erik
05:00:27bug1im going to have a quick look into #1200, an fuser applet
05:00:44bug1dont know much about it though.
05:07:31anderseeDid you see vodz' comments?  I didn't understand what he was suggesting.  Did you understand what he was saying?
05:09:51bug1vodz comment on fuser... no ?
05:10:11bug1in the bts ?
05:23:58bug1on second thoughts fuser is going into the too hard box for the moment
05:25:12anderseenono.  vodz' comments on the mailing list today re the copy_file change
05:27:32bug1something about user interaction on special files
05:27:55anderseeright.  I didn't quite follow him.
05:28:52bug1force a mode == cp -f
05:30:18bug1"It is possible to leave your duplicated code for unlink() in force a mode," i think he means you dont need your patch if compiling busybox without user interactin enabled
05:30:42bug1im not sure iether
12:39:34sjhillg'morning mjn3
12:39:42mjn3morning.  how goes it?
16:34:03mjn3andersee: hey.  if you haven't started on get_header_zip.c, i'll take work on that now if you like
16:48:24anderseemjn3: cool, go for it.
16:48:36anderseemjn3: BTW, where did you find out about the problem?
16:49:38mjn3andersee: someone on the zaurus list mentioned that unzip wasn't working.  i confirmed it failed on arm and investigated
16:51:29anderseeAhh
16:51:52anderseeThis morning I rebuilt latest busybox stable for my tuxscreen, and it didn't boot.  :-(
16:52:09anderseeSo I'll need to figure out what I broke....
16:53:12mjn3ohoh
18:27:41mjn3andersee: hey.  just doing some testing on the unzip stuff
18:32:13anderseecool.  I'm just starting to try and find what got broken last night.
18:36:59anderseemjn3: I'm testing in buildroot (with UML).  Reverting ash makes it boot....
18:41:14mjn3andersee: regarding unzip... it should be more robust but it no longer distinguishes between invalid archives, read errors, etc. on failure.  that a problem?
18:42:25anderseeI suppose we can live with that...
18:42:44anderseeAt the end of the day, they all cause the same problem.
18:43:20mjn3true
18:43:36mjn3anybody who doesn't like it can fix it themselves..  ;-)
18:43:46mjn3off to get a haircut
18:45:06anderseeDid you commit that fix then?
18:45:58mjn3-awaynot yet, but i need to get to the barbers in 5 min.  i'll do it when i get back
18:46:12mjn3-awayalso, watch needs to be fixed... the clear screen thing
18:47:55anderseek
19:39:30mjn3andersee: did you bix bb.stable ash?
19:40:04anderseeyup.  Just checked in the fix...
19:40:07anderseehehe
19:40:28mjn3what's funny?
19:40:29anderseeWhile adding uClinux patches I accidentaly made ash use vfork!
19:40:49anderseegets a boot to the head
19:41:59mjn3andersee: well, you did say you were having trouble "boot"ing
19:42:06anderseehehe
19:42:28anderseeso I think once your gzip fix is in we should be all set for a release.
19:43:12mjn3how do you want to handle the clearscreen thing in watch.c?
19:44:11anderseeprintf("\033[H\033[J");
19:45:06anderseemjn3: but watch is not in the stable release...
19:45:31mjn3ahh..
19:47:45anderseemjn3: k, I checked in the fix for watch screen clearing
19:49:25BZFlagandersee: would be nice if "reset" restored the cursor. course it would be nicer if qpe didn't try to turn off the cursor on terminals it would not be running on.
19:50:06anderseeBZFlag: Hmm.
19:50:48anderseeBZFlag: Is there some utility that _does_ restore the cursor?
19:51:16mjn3andersee: perhaps we need a standard clear function in libbb.  i suppose for dumb terminals would could dump $LINES newlines.  not too good for teletypes though.  ;-)
19:51:16BZFlagexiting and restarting minicom? =(
19:51:36anderseeBZFlag: or starting vim...
19:52:37anderseeBZFlag: looks like 'setterm -cursor [on|off]' will do the job
19:53:43BZFlaghmm. no setterm at the moment. =(
19:54:24BZFlag^[[?25l ?
19:54:57anderseeBZFlag: setterm is part of util-linux
19:57:02BZFlagand 25h turns it on.
19:57:31BZFlagprintf "\033[?25h" in busybox. ;-)
19:58:53BZFlagprintf("\033c"); -> printf("\033[?25h\033c"); in reset.c ?
19:59:26anderseeBZFlag: you have my blessing to commit that.  Its getting a bit garbled via irc...
19:59:50mjn3andersee: the unzip patch is committed to unstable
20:00:15anderseemjn3: k.  I'll take a look.  Want to backport or shall I?
20:00:40mjn3andersee: better you.  i don't have the stable tree checked out
20:01:43anderseemjn3: k
20:02:38anderseeBZFlag: does it have the '?' or is that some esc char thing
20:03:56BZFlagnope it's a '?'
20:04:22anderseeBZFlag: oh.  ok.
20:04:28BZFlaglooks like ?25h is "display caret" in vt100 speak.
20:04:51anderseemjn3: hehe.  Guess we don't have unzip in stable, so no problem there...
20:05:40BZFlaghttp://216.239.35.100/search?q=cache:ECEl7VCPGCAC:www.catalyst.com/support/help/cstools3/library/terminal/sequences.html+ansi+%22display+caret%22&hl=en&ie=UTF-8
20:06:03BZFlagand should reset also do a clear?
20:06:17anderseeBZFlag: cool.  Tested (using setterm -cursor off then busybox reset) and it works.
20:07:29BZFlagsweet.
20:09:48BZFlagso I guess I'd prefer printf("\033[?25h\033c\033[H\033[J");
20:10:42BZFlagthough I suppose the \033[H is redundant (move cursor home)
20:11:32BZFlagBB already assumes vt100+ terminal yes?
20:11:40BZFlagassumes so.
20:11:42anderseebasically
20:12:08BZFlagso "\033[?25h\033c\033[J" should suffice.
20:12:21anderseeBZFlag: What does that do?
20:12:37BZFlagrestore cursor, reset terminal, clear screen.
20:13:26BZFlagcurrent bb reset restores the terminal in minicom which moves the cursor home, but does not clear the screen.
20:13:37anderseeBZFlag: I think I'm going to rebuild the tuxscreen, and assuming it looks fine I'm going to release busybox 0.60.4
20:13:46BZFlagsweet.
20:14:01BZFlagshould get a new tux image up now that modprobe et all is fixed.
20:14:17anderseeWant me to change to that final reset string?
20:16:19anderseeOh.  I have one more applet to review...
20:20:01BZFlagsure. might as well. I don't know of any terminals that would not handle that sequence. Some would ignore it, but they should all work.
20:58:32sandmanandersee: any plans for a bb 0.61 release ?
21:04:05anderseesandman: yup
21:04:36anderseesandman: but I'd like to get the 0.60.4 release out before even thinking about that....
21:04:58anderseesandman: looks like 0.60.4 will go out the door today.
21:05:18anderseesandman: And I'll probably take a day or two off from busybox work
21:05:40anderseesandman: So perhaps late this week or next week.
21:09:51sandmanandersee: Anything in unstable that needs more testing ?
21:10:11sandman(Besides the missing SUID documentation ;)
21:11:29anderseesandman: well, right now source dependancies are complete crap...
21:12:45anderseesandman: So thats one thing I'd like to look at
21:14:25sandmanandersee: All the applets seem to work perfectly so far -- Everyone running familiar linux on his iPAQ is beta-testing it for 2 months now ;)1~
21:14:45BZFlagautogen source deps?
21:16:41mjn3sandman: the unzip problem i fixed today was only reported yesterday and it has been there since unzip was added 8 months ago
21:17:24sandmanmjn3: unzip is one of the few applets that isn't included in the familar bb build ..
21:21:42anderseemanually updates the busybox daily snapshots
21:22:59anderseeUnless someone finds a but in busybox stable, its going to be released as-is (with minor doc and changelog updates)
21:33:56andersees/but/bug/
22:00:15mjn3andersee: the new top fails on the zaurus.  initial guess... 2.4.6 doesn't have the Tgid entry.  also, top.o is a major pig sizewise.  :-(
22:19:56anderseesigh.   Oh well.
22:22:32mjn3i'll take a look at it this evening...
22:22:45mjn3just about to go do supper

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with infobot logs, split per channel and by date, etc.