Re: stty erase setting
On Sun, May 29, 2005 at 16:12:01 +0300, Elan Ruusamäe wrote: > setting stty erase from /etc/bashrc fcks up my backspace, at least on So you have something fucked up. I suggest reviewing your inputrc and removing all hardcoded sequences. What's your term? Is it local or remote? > and in any case, such things should be in eighter > 1. terminfo > 2. /etc/profile It cannot be in profile, because it's terminal dependand and you can change it without relogin. And how do you think you set it from terminfo? How can _remote_ terminal guess what stty to set? xterm: ~: stty speed 38400 baud; line = 0; erase = ^H; ^^^ -brkint -imaxbel ~: tput kbs | hexdump -C 08|.| 0001 ~: ssh $ stty speed 38400 baud; line = 0; [oh, there's NOTHING here] -brkint -imaxbel $ tput kbs | hexdump -C 08|.| 0001 $ less /etc/passwd /akhdkasjdhas^H^H^H^H^H^H^H ^^ these are literally inserted! linux console: ~: stty speed 38400 baud; line = 0; -brkint -imaxbel ~: tput kbs | hexdump -C 7f|.| 0001 ~: ssh $ stty speed 38400 baud; line = 0; -brkint -imaxbel $ tput kbs | hexdump -C 7f|.| 0001 $ less /etc/passwd /akhdkasjdhas^H^H^H^H^H^H^H ^^ they do what they should So I assume that no 'erase' in stty means use default '^?' (linux-console backspace). Anyway resetting ANYTHING to tput-returned values MUST NOT break anything. If it's not the case - there must be some misconfiguration elsewhere. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
STBR arpwatch
asd -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: lstat.spec - make %{_wwwrootdir}/lstat/doc symlink to %{_do...
On Fri, Jun 03, 2005 at 12:41:48 +0300, Elan Ruusamäe wrote: > well. imho this is wrong. This is how it's done for many packages. > documentation that application uses should be still bundled with package main > files, as if i install package --exclude docs, i get partial package. Yes. That's what exclude docs mean. > as this is against for point one from devel-hints-en.txt: Could you emphase where exactly? > Every package should, whenever possible, contain documentation files. It does. > Documentation files should be marked with %doc, which makes it possible That's right. > to refrain from installing them by passing the --exclude-docs rpm > option, if that's what the user desires. You shouldn't: Yeah. > - mark files crucial to the program's functionality as %doc It's not crucial to use lstat. > and perhaps consider creating -doc subpackage, for folks having --exclude > docs > and still get those documentation files used by app? Feel free. But package these files only once, as it's 600KB. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
postfix upgrade failure
poldek> upgrade postfix-2.2.3-1 Processing dependencies... postfix-2.1.5-5 obsoleted by postfix-2.2.3-1 There are 1 package to install, 1 to uninstall: I postfix-2.2.3-1 R postfix-2.1.5-5 Need to download about 1MB of archives. After unpacking about 4MB will be used. Retrieving ftp://ftp2.pld-linux.org/dists/ac/PLD/athlon/PLD/RPMS/postfix-2.2.3-1.athlon.rpm... . 100.0% [1.3M (233.0K/s)] Executing rpm --upgrade -vh --root / --noorder... Preparing...### [100%] 1:postfixwarning: /etc/mail/access created as /etc/mail/access.rpmnew warning: /etc/mail/main.cf created as /etc/mail/main.cf.rpmnew warning: /etc/mail/master.cf created as /etc/mail/master.cf.rpmnew warning: /etc/mail/virtual created as /etc/mail/virtual.rpmnew ### [100%] Editing /etc/mail/master.cf, adding missing entry for anvil service Editing /etc/mail/master.cf, adding missing entry for scache service Editing /etc/mail/master.cf, adding missing entry for discard service Editing /etc/mail/master.cf, adding missing entry for tlsmgr service Stopping Postfix service...[ DONE ] Starting Postfix service...[ FAIL ] error: %post(postfix-2.2.3-1) scriptlet failed, exit status 1 There were errors maillog says: Jun 17 03:15:37 pepin postfix/smtpd[17237]: warning: connect to private/tlsmgr: No such file or dir ectory Jun 17 03:15:38 pepin postfix/postfix-script: stopping the Postfix mail system Jun 17 03:15:38 pepin postfix/postfix-script: fatal: the Postfix mail system is already running I had to manually uninstall postfix-2.1.5 and restart postix. It works. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Thu, Jun 30, 2005 at 11:01:36 +0300, Elan Ruusamäe wrote: > jeee!!! jipp!!! partyy Yeah, brilliant idea: $ infocmp | grep kbs is2=\E7\E[r\E[m\E[?7h\E[?1;3;4;6l\E[4l\E8\E>, kbs=^H, $ stty -a | grep erase intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = ; eol2 = ; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; Here you got geniuses: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=77833#c13 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=77833#c14 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=11294#c13 http://howtos.linux.com/howtos/BackspaceDelete/system.shtml http://lists.gnu.org/archive/html/screen-users/2004-07/msg3.html -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Thu, Jun 30, 2005 at 17:09:38 +0300, Elan Ruusamäe wrote: > so if you're so smart, perhaps you implement it in a way it works everywhere > (for me) not partially (for you)? Just tell me what doesn't work PARTIALLY for you (as it works everywhere for me). -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Thu, Jun 30, 2005 at 16:12:37 +0200, Jan Rekorajski wrote: > And because RH is just srewed we have to be more? You could just read this. > Come on, I have _never_ had any problems with it _until_ you "fixed" it. You've got my bugreport with testcase, #5671 AFAIR. And what are your problems? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Thu, Jun 30, 2005 at 18:26:14 +0200, GoTaR wrote: > > And because RH is just srewed we have to be more? > > You could just read this. Just to be precise: we are screwed, RH has fixed everything (including broken screen). -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Thu, Jun 30, 2005 at 18:27:49 +0200, Tomasz Pala wrote: > > You could just read this. > > Just to be precise: we are screwed, RH has fixed everything (including > broken screen). Debian is the same. So we've got ridiculous situation: RH/FC/Debian works everywhere, PLD works fine locally and remotely with RH/FC/Debian, but it's not working remotely with _another PLD_. And you say, that RH is screwed up, right? Do you mean that RH in some magic way has influenced PLD-PLD interoperability, mhmm? Interesting. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Thu, Jun 30, 2005 at 18:37:25 +0200, Jan Rekorajski wrote: > Randomly non-working backspace in every xterm[1]/screen/whatever combination. Great, now I can reproduce and trace it. Thanks a lot. > It just works without your "fix". It works just because readline understand hardcoded ^? sequence. ncurses has not. EOT -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Thu, Jun 30, 2005 at 22:19:50 +0200, Jan Rekorajski wrote: > Please show me _where_ exactly stty erase is set to tput kbs in Debian > and FC. That's the question I was trying to answer for a few days. Without success. They used to do this even a year ago, after that sth else was changed... I cannot find any trace. > I'm looking all over /etc in those two distributions and just > can not find it. That's right. > And, yes, so happens that I have both installed and > both are quite fresh. > > It seems to me they have fixed your problems some other, proper, way. > You just didn't look deep enough. For now I only suspect one thing: they use termcap, we use terminfo. I've already smashed one of my systems - bash, pam, sshd, readline, terminfo databases taken from Debian, and no change. There's other workaround possible - changing behavior of xterm to produce ^? instead of ^H: ~MetaBackSpace: string(\177)\n\ Is hardcoded and recognized by readline, and it would be the same as stty setting for ncurses, but it's not the right way. The same as using [ "$TERM" = "xterm ] && tset -e Really there's noone here who knows, where the hell is source of stty settings after remote login? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Thu, Jun 30, 2005 at 23:00:44 +0200, GoTaR wrote: > There's other workaround possible - changing behavior of xterm to > produce ^? instead of ^H: http://groups.google.com/groups?hl=pl&lr=&biw=787&threadm=linux.kernel.E15lLJL-0003lF-00%40bowl.fysh.org&rnum=34&prev=/groups%3Fq%3Dtermcap%2Bterminfo%2Berase%26hl%3Dpl%26lr%3D%26biw%3D787%26start%3D30%26sa%3DN Hmmm... if I understand Alan Cox correctly, we shall do this. So now there are two questions: 1. any objections? 2. what about that bloody stty propagation? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Fri, Jul 01, 2005 at 00:02:10 +0200, GoTaR wrote: > Hmmm... if I understand Alan Cox correctly, we shall do this. So now > there are two questions: > 1. any objections? Let's go further - Debian has it's policy, but _they didn't change [*] xterm-color terminfo_, so remotely it's broken (blues, what were you talking about problems with Debian/Ubuntu? ;>). Instead they created xterm-debian, but it's not known on other systems (hurray for idiots). http://log.antiflux.org/grant/2004/03/29/backspace-delete http://www.qsl.net/n0nb/linux/debiantips.html http://www.ibb.net/~anne/keyboard.html They all (just like Debian, Cox, Torvalds) suggest using '^?'. It's good idea but only if we change kbs in terminfo (well, kdch1 is changed anyway, khome and kend added, what a fucking mess...), otherwise any terminfo bindings (ncurses) will fail. Unfortunatelly we cannot do this. > 2. what about that bloody stty propagation? I'll stick with that anyway. It guarantees, that every application using termios will work. So here comes the most important question: we adapt to Debian (supported by authorities) or xterm's specification? In case of khome/kend we had chosen (well, kde-konsole has forced us) to follow xterm against Debian. I'm still waiting for test cases what was wrong with my fix! [*] let me guess, for interoperability with other systems... s p o i l e r PS. someone has said, that with my fix screen doesn't work properly, yeah? $ tail -n3 /etc/screenrc # On Debian GNU/Linux, `<--' (Backspace key) should send char `\177': bindkey -k kb stuff "\177" So if there won't be any serious discussion and detailed scenarios I'm going to restore my fix and you will have to fix YOUR BROKEN programs. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Fri, Jul 01, 2005 at 13:05:36 +0200, Jan Rekorajski wrote: > > I'm still waiting for test cases what was wrong with my fix! > > I'm getting ^? instead of erased character. Scenario: > > a = host without stty erase set OK, my machine. > b = host with stty erase in bashrc > > start X on a > run xterm Wait a minute - your xterm doesn't set stty erase to '^H' !? > ssh to b > run vim, write something, try backspace The same scenario works for me. It didn't without stty erase on the remote side. I'm using both xterms - from XFree86 and X11.org. > Don't do it, you'll end up fixing whole world. I'm wondering if Debian is not trying to fix the world. For now we know: PLD Debian RH/FC others... ideal xterm ^H ^? ^H or ^? generates kbs for ^H ^H ^H ^H or ^? xterm-color stty after not propagated propagated remote loginset In Debian backspace bindings doesn't work unless terminal is specified as xterm-debian. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Fri, Jul 01, 2005 at 13:38:49 +0200, Jan Rekorajski wrote: > I got somewhere, first what it looks like on RH 7.3: > > [EMAIL PROTECTED] baggins]$ echo $TERM > xterm H... why have you xterm not xterm-color? They differ... > [EMAIL PROTECTED] baggins]$ infocmp | grep kbs > kbeg=\EOE, kbs=\177, kc1=\EOq, kc3=\EOs, kcub1=\EOD, xterm-color at FC2: is2=\E7\E[r\E[m\E[?7h\E[?1;3;4;6l\E[4l\E8\E>, kbs=^H, > [EMAIL PROTECTED] baggins]$ grep -i backarrowkey > /usr/X11R6/lib/X11/app-defaults/XTerm > *VT100*backarrowKey: false This is default setting (according to xterm man page). > [EMAIL PROTECTED] baggins]$ stty -a | grep erase > intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = ; erase setting on RH is propagated from your local terminal. Try: 1. 'stty erase ^H' on your local terminal 2. ssh to RH 3. check stty settings. > So what we need to do: > > a) kill X11-XTerm.ad.patch or s/0x08/\177/ and/or add *VT100*backarrowKey: > false The problem is XTerm-color includes XTerm, so the line: ~MetaBackSpace: string(\177)\n\ would apply for both of them. > b) the same with XTerm.ad-pl > c) s/^H/^?/ in terminfo entries for xterm -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Fri, Jul 01, 2005 at 14:26:16 +0200, GoTaR wrote:
> PLD Debian RH/FC others... ideal
>
> xterm ^H ^? ^H or ^?
> generates
>
> kbs for ^H ^H ^H ^H or ^?
> xterm-color
>
> stty afternot propagated propagated
> remote login set
strace -e ioctl -v -f on sshd master process:
ioctl(6, SNDCTL_TMR_START or TCGETS, {c_iflags=0x500, c_oflags=0x5,
c_cflags=0xbf, c_lflags=0x8a3b, c_line=0,
c_cc="\x03\x1c\x7f\x15\x04\x00\x01\x00\x11\x13\x1a\x00\x12\x0f\x17\x16\x00\x00\x00"})
= 0 ^1^ ^2^
1 - erase
2 - kill (kill, quit and others aren't propagated to PLD either!)
etc.
~: stty kill ^B
~: ssh pld-machine
$ stty
speed 38400 baud; line = 0;
-brkint -imaxbel
$ logout
~: ssh debian-machine
$ stty
speed 38400 baud; line = 0;
kill = ^B;
-brkint -imaxbel
--
GoTaR gotar>http://vfmg.sourceforge.net/
http://tccs.sourceforge.net/
___
pld-devel-en mailing list
pld-devel-en@lists.pld-linux.org
http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
UseLogin yes fuck -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Fri, Jul 01, 2005 at 13:38:49 +0200, Jan Rekorajski wrote: > a) kill X11-XTerm.ad.patch or s/0x08/\177/ and/or add *VT100*backarrowKey: > false > b) the same with XTerm.ad-pl > c) s/^H/^?/ in terminfo entries for xterm You didn't read me carefully. This is Debian: $ echo $TERM; infocmp | grep kbs xterm-color is2=\E7\E[r\E[m\E[?7h\E[?1;3;4;6l\E[4l\E8\E>, kbs=^H, I'm testing your change for one day and it's to be reverted. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Sat, Jul 02, 2005 at 19:45:03 +0200, Jan Rekorajski wrote: > > You didn't read me carefully. This is Debian: > > > > $ echo $TERM; infocmp | grep kbs > > xterm-color > > is2=\E7\E[r\E[m\E[?7h\E[?1;3;4;6l\E[4l\E8\E>, kbs=^H, > > And do they acually USE xterm-color? They don't (at least through xterm binary) - maybe that's why they forgot (!?) to change it too. But out xterm sets TERM=xterm-color, so when I login to Debian/RH all the backspace keybindings not work (read below). > Show me `TERM=xterm infocmp | grep kbs` $ TERM=xterm infocmp | grep kbs kb2=\EOE, kbs=\177, kcbt=\E[Z, kcub1=\EOD, kcud1=\EOB, $ TERM=xterm-color infocmp | grep kbs is2=\E7\E[r\E[m\E[?7h\E[?1;3;4;6l\E[4l\E8\E>, kbs=^H, The latter one is problem at Debian and RH, because currently it's different that ours. The former one was not problem without yout changes, because our xterm doesn't set TERM=xterm. > > I'm testing your change for one day and it's to be reverted. > > And what's wrong with it? I didn't see _any_ side-effects, and I tested > it with a few combinations of new/old terminfo/xterm. With TERM=xterm it's ok (better than before). With xterm-color (our default) there's difference in generated sequence and terminfo, thus key bindings (NOT backspace as erase!) in ncurses don't work. E.g. mutt in pager mode: previous-line scroll up one line Either we readapt xterm-color to Debian and RH (kbs=^H) or we convince them to change it too - I have no idea if it is intentional difference between xterm and xterm-color or their both mistake. One more thing - what stty erase sets xterm after your changes? You should have added "*ttyModes: erase ^?". -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Sat, Jul 02, 2005 at 23:10:27 +0200, Jan Rekorajski wrote: > My proposal is for us to stop being so uber-smart and just revert to > defaults. Kill all those s/xterm/xterm-color/ and s/xterm/we-know-better/ > patches and just stick to TERM=xterm like the rest of the world. I agree, we shall do it the _common_ way. However Debian/RH/FC has ^H for xterm-color, so following above rule you should revert ncurses-xterm-color-new.patch, don't you? ;) > And we should also go with debian-guildines here (like eg. FC) and set > kbs=\177, then we will have consistent envinronment on console and in X. The question is: are we going to fulfill this policy wider than it's creator? If yes - ncurses-xterm-color-new.patch stays. But I prefer to keep in sync with our big brothers. > > One more thing - what stty erase sets xterm after your changes? You > > should have added "*ttyModes: erase ^?". > > Maybe I should add "*backarrowKey: false". No, it's false by default. There must be line I gave, because IN FACT we're changing xterm's default and it must be noticed about it. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Sat, Jul 02, 2005 at 19:11:34 +0200, GoTaR wrote: > UseLogin yes > > fuck No, one must set: UsePrivilegeSeparation yes in order to propagate stty settings. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Sun, Jul 03, 2005 at 12:33:38 +0200, Jan Rekorajski wrote: > On Sun, 03 Jul 2005, Michal Kochanowicz wrote: > > > Third question: > > 3. WTF backspace doesnt work... "" ;( > > I guess you upgraded terminfo but not X11. Either upgrade terminfo or add ttymodes/ptyInitialErase for a time. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Sun, Jul 03, 2005 at 11:59:28 +0200, Jan Rekorajski wrote: > > I agree, we shall do it the _common_ way. > > > > However Debian/RH/FC has ^H for xterm-color, so following above rule you > > should revert ncurses-xterm-color-new.patch, don't you? ;) > > That's right, especially when no sane program is using it in linux-world. OK, I've changed SPECS/ncurses.spec and SOURCES/X11-PLD.patch (finding no other reference to xterm-color, hope all other patches would match). > > No, it's false by default. > > man xterm: >backarrowKey (class BackarrowKey) Erghhh... I've read about backarrowKeyIsErase: backarrowKeyIsErase (class BackarrowKeyIsErase) Tie the VTxxx backarrowKey and ptyInitialErase resources together by setting the DECBKM state according to whether the initial value of stty erase is a backspace (8) or delete (127) character. The default is ``false'', which disables this feature. Indeed, this would be useful if we haven't had vt100.translations: #override, but in our case it won't change anything, as well as the backarrowKeyIsErase resource (as far as I can see, anyway this resouce works backwards - looks for stty erase). Changing these two has no effect. > > There must be line I gave, because IN FACT > > we're changing xterm's default and it must be noticed about it. > > You're right here, the change must be complete. There are two options which can set this, both having the same effect: ttyModes (class TtyModes) [...] This is very useful for overriding the default terminal settings without having to do an stty every time an xterm is started. [...] ptyInitialErase (class PtyInitialErase) If ``true'', xterm will use the pseudo-terminal's sense of the stty erase value. If ``false'', xterm will set the stty erase value to match its own configuration, using the kb string from the termcap entry as a reference, if available. [...] The default is ``false''. That's why after changing ncurses it worked for you:) I've got old terminfo and thus needed ttyModes or ptyInitialErase: true. My suggestion is temporary solution for people not having current both xterm and terminfo (or maybe add some R:?). In my opinion it's all finished now. Any comments? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Sun, Jul 03, 2005 at 21:23:10 +0200, Jan Rekorajski wrote: > konsole, gnome-terminal and a lot more ;) > If there won't be any objections I will remove all those hacks tomorrow. kdebase-konsole-TERM.patch - TERM kdebase-konsole-default-keytab.patch - backspace kdebase-konsole_all.patch - TERM multi-gnome-terminal-xterm-color.patch - TERM Eterm-xterm-color-fixes.patchc - ? gnome-terminal-TERM.patch - there's nothing to do -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (AC-branch): bashrc - removed stty erase setting to `tput ...
On Sun, Jul 03, 2005 at 13:53:18 +0200, Tomasz Pala wrote: > In my opinion it's all finished now. Any comments? It's not over yet On PLD, Debian 3.1, FC2, FC3: TERM=xterm infocmp | grep kf[1234]= TERM=xterm-color infocmp | grep kf\[1234\]= gives: xterm: kf1=\EOPkf2=\EOQkf3=\EORkf4=\EOS xterm-color:kf1=\E[11~ kf2=\E[12~ kf3=\E[13 kf4=\E[14~ fixed in X11-XTerm.ad.patch and XTerm.ad-pl. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: php-apache.ini doesn't work
On Sun, Jul 24, 2005 at 23:53:41 +0300, Elan Ruusamäe wrote: > > session.save_path = /var/run/php > > > > but until I've changed it in php.ini all my sessions were saved into > > /tmp. > fixed in HEAD for php4 some time ago > > the ini file under apache2 is php-apache2handler.ini, not > php-apache.ini Indeed. Hopefully fixed on HEAD. It was blind commit so there's .1 release, someone could check/rebuild it please? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
zebra/quagga vs bind init priority
I've just fixed rc-scripts.db.txt. There's a problem with bind and current sequence: 14 named bind 15 zebra quagga/zebra As named binds to IPs it finds at start up and there's no interfaces before zebra is run it listens nowhere. How about moving zebra to 13? I'm not sure about: 14 stund stun 14 stunnel stunnel 14 svscan daemontools 14 svscan freedt but: 15 rdate rdate(-bsd) would be even happier with: 16 bgpdquagga/zebra 16 ospf6d quagga/zebra 16 ospfd quagga/zebra 16 ripdquagga/zebra 16 ripngd quagga/zebra moved to the front. So my proposal: 13 zebra q/z 15 bgp*/ospf*/rip* q/z 16 rdate rdate(-bsd) -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ,,Forwardowanie IP'' za NAT
EN! On Mon, Jul 25, 2005 at 07:38:38 +0200, Andrzej Krzysztofowicz wrote: > A nie prosciej zrobic to na poziomie routingu? > Tzn. nie NAT-owac polaczen z/do okreslonych IP. It's easier, but you waste adresses for gateway, network and broadcast. It's not problem when you're LIR, but most people in Poland have no more than /27. Soon I'll face dividing /20 for customers and I'm thinking about using /32 masks - anyone has some experience? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ,,Forwardowanie IP'' za NAT
On Mon, Jul 25, 2005 at 14:42:40 +0200, Andrzej Krzysztofowicz wrote: > > It's easier, but you waste adresses for gateway, network and broadcast. > > No, I don't. > Hint: use routing to a single host, not full network. In case you've got one range and must assign one of it's public addresses to the router anyway. But it's common to get one link /30 and the rest in /27. How to set gateway on client with address from /27 not having this range on any interface? E.g. router: 1.1.1.2/30 - world 10.0.0.0/24 - local 2.2.2.0/27 - 32 clients > > about using /32 masks - anyone has some experience? The problem is it'll be dozen intefaces on 20 machines and hundreds of clients, too much to set all the routes by hand (or wasting ~50 ifaces for 3 IPs=150, 3.5% of available address space) and struggling with tc queues, so: > And either: > - configure routing to X.Y.Z.10 via X.Y.Z.1 on any host in X.Y.Z.0/24 on eth0 > that needs the routing (especially the gw, X.Y.Z.254), or Can OSPF or iBGP set it NOT runing on client machine, only in backbone? > - configure proxy arp on X.Y.Z.1 machine (to respond on arp requests for > X.Y.Z.10 on eth0) Is proxy arp safe in large ethernet segments? I'm not quite sure if NAT isn't simplier method... 1. simple routing: I can divide IPs on per-router basic and don't care if there's only 1 IP on interface or 32 and how many interfaces there is - all the interfaces consume common allocation, 2. client gets his local IP from DHCP, don't have to manually set second one (many of them wants to use neighborhood and other network trash), has direct (switch) access to other machines in LAN, so doesn't bother our router to carry it's traffic, 3. I can forget about public IPs in QoS. Disadvantages: 1. it must be explained to clients, that they do have public IP, 2. above is not true if he wants to use AH. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: openssl-0.9.7g-2 openssh AMD64
On Wed, Jul 27, 2005 at 17:42:56 +0200, Andrzej Zawadzki wrote: > >>and in AMD64 syslog I have: > >>sshd[pid]: segfault at "nr" rip "nr" rsp "nr" error 4 > >> > >>That's only happens on AMD64. > > > > > > We had the problem on mandriva after changing of openssl, try to rebuild > > openssh. > > Thank You! I did like You said and now works :-) Oh... so that's why I have no access to one of my machines... > ps. cvs: rel. up and send to builders? ...and fix my server;> -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
jabber c2s dead loop
After last week updates jabber doesn't work. Any try to connect ends
with endless:
select(0, NULL, NULL, NULL, {0, 1}) = 0 (Timeout)
Only kill -9 helps.
--
GoTaR gotar>http://vfmg.sourceforge.net/
http://tccs.sourceforge.net/
___
pld-devel-en mailing list
pld-devel-en@lists.pld-linux.org
http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
http://ftp.pld-linux.org/ via proxy
Test Page for Apache Installation If you can see this, it means that the installation of the Apache web server software on this system was successful. You may now add content to this directory and replace this page. [...] Any ideas what's messed up? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: http://ftp.pld-linux.org/ via proxy
On Fri, Aug 26, 2005 at 16:32:40 +0200, Adam Gołębiowski wrote: > > Any ideas what's messed up? > > Yep, in your network, In my 3 independent networks;) And I really have no idea how it is possible that direct access works, but proxy fails to get it (in 2 cases), one machine cannot with nor without proxy and one works... > ftp.pld-linux.org points to something else than > ep09.pld-linux.org. I just checked and it's probably ftp mirror at > icpnet. If you'd like to use http based access to ftp, use > ftp1.pld-linux.org which allways points to ep09. Hey, I don't care what points where, anyway it changes too often (maybe some ftp-current-dns-settings.rpm? ;P) and that's why DNS has been invented. I-the-user just want to have http access (as poldek doesn't work with proxy on ftp://) and it WAS working, now something was broken. > Or bug ftp admins to configure their httpd ... Now it's even better: ~: wget -O - http://ftp.pld-linux.org --17:05:26-- http://ftp.pld-linux.org/ => `-' Resolving ftp.pld-linux.org... 149.156.143.253 Connecting to ftp.pld-linux.org|149.156.143.253|:80... failed: No route to host. ~: ping ftp.pld-linux.org PING ftp.pld-linux.org (149.156.143.253) 56(84) bytes of data. >From 149.156.6.33 icmp_seq=2 Packet filtered ; -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: http://ftp.pld-linux.org/ via proxy
On Fri, Aug 26, 2005 at 17:17:50 +0200, Adam Gołębiowski wrote: > > so, adamg, please, fix it 8-] > > -ENOTENOUGHMANA :) It worked for me without proxy because I'd got it in hosts file. Now, is there any chance someone would fix it or sth? Meanwhile I'll set static hosts for some of my proxy servers as it's much easier to add 3 lines than changing 2*12 and refetching indexes or renaming index directories... Polpak: ~: host ftp.pld-linux.org ftp.pld-linux.org is an alias for ftp1.pld-linux.org. ftp1.pld-linux.org is an alias for ftp.sk.pld-linux.org. ftp.sk.pld-linux.org has address 217.73.17.117 ~: host 217.73.17.117 117.17.73.217.in-addr.arpa domain name pointer ep09.pld-linux.org. GTS: ~: host ftp.pld-linux.org ftp.pld-linux.org is an alias for ftp5.pld-linux.org. ftp5.pld-linux.org has address 62.21.96.15 ftp.pld-linux.org is an alias for ftp5.pld-linux.org. ftp.pld-linux.org is an alias for ftp5.pld-linux.org. ~: host 62.21.96.15 15.96.21.62.in-addr.arpa domain name pointer ontario.icpnet.pl. Futuro/Crowley/ceti: ~: host ftp.pld-linux.org ftp.pld-linux.org is an alias for ftp5.pld-linux.org. ftp5.pld-linux.org has address 62.21.96.15 -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: http://ftp.pld-linux.org/ via proxy
On Mon, Aug 29, 2005 at 09:16:28 +0200, Adam Gołębiowski wrote: > But DNS changes too, and the idea behind recent changes was to bring us > to the situation where querying for ftp.pld-linux.org gives us a server > that is theoretically best for us. I understand and support this idea. > > I-the-user just want to have http access (as poldek doesn't > > work with proxy on ftp://) and it WAS working, now something was broken. > > fill a bug report :) AFAIR I've written about it to mis. By the way: http://www.polanet.pl/~gotar/flyspry.png What are these fields? ;] > Anyway http://ftp5.pld-linux.org/ is now working. Thanks:) -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
http://ftp5 again...
error: vfff: http://ftp.pld-linux.org/dists/ac/ready/athlon/packages.i/packages.dir.diff.toc.gz: no such file -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn... (Re: SOURCES: ghostscript-afpl-am.patch (NEW), ghostscript-afpl-ijs_pkg...)
On Wed, Sep 07, 2005 at 10:20:54AM +0100, wrobell wrote: > > svn gives us some advantages. disadvantages? any real, which makes life > really painful? How to resolve conflict? I've got a situation: ~: vi blabla ~: svn ci blabla snv reports conflict here [fixing it manually] ~: svn ci blabla svn reports there's unresolved conflict -- Tom Pala <[EMAIL PROTECTED]> http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn... (Re: SOURCES: ghostscript-afpl-am.patch (NEW), ghostscript-afpl-ijs_pkg...)
On Fri, Sep 09, 2005 at 00:27:58 +0200, Arkadiusz Miskiewicz wrote: > Yeah, then show me how to fetch single app with spec + patches using command > line - cvs and sh only. cvs up builder sh builder -g foo.spec (there IS working builder). -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn... (Re: SOURCES: ghostscript-afpl-am.patch (NEW), ghostscript-afpl-ijs_pkg...)
On Fri, Sep 09, 2005 at 17:32:08 +0100, wrobell wrote:
> > cvs up builder
> > sh builder -g foo.spec
> >
> > (there IS working builder).
>
> and when we are talking about skipping rpm/{SPECS,SOURCES}
> structure for svn repo, then we are forbidden to use scripts :)
The difference is this scripts gets a dozen of files in a dozen
requests and one has to do 2 cvs ups.
A future script will get 10k specs with 10k svn requests, and how
many updates would require? (see p. 3)
Let's make a list of COMMON tasks (excluding trivial ones), appropriate
command sets and compare it.
1. building a package from given branch:
cvs:./builder -bb -R foo bar or ./builder -g and rpmbuild -bb
svn:!? rpm uses SPECS/SOURCES subdirs, how the hell this will work
with subdir-per-packet? Another messing script?
2. updating all resources I've checked out (having CVS/Entries.Static):
cvs:cd SPECS, cvs up; cd ../SOURCES; cvs up
svn:!?
3. fetching all specs:
cvs:cvs up
svn:
for pkg in `svn ls http://.../packages/`; do
do whatever you need
done
ohhh, it really sucks...
4. commiting changes:
cvs:cvs ci foo.spec; cd ../SOURCES; cvs ci *
svn:svn ci foo
hmmm, not so big difference, don't talk about atomic commits anymore.
Anyway - if we want this functionality, we could move SOURCES and SPECS
into higher level repo directory, so that they would be subdirs.
And now NOT COMMON ones:
5. reverting a commit:
cvs:cvs up -r X foo; mv foo{,.tmp}; cvs up -A foo; mv -f foo{.tmp,}; cvs ci
foo
svn:svn revert foo
no difference for me. Not an argument.
5. renaming
cvs:manually
svn:svn rename
well, I don't care about names, I asked to do this once or twice.
6. access to deleted TAGS
cvs:none
svn:is
who needs it anyway (OK, you need this in DEVEL example, I don't)
--
GoTaR gotar>http://vfmg.sourceforge.net/
http://tccs.sourceforge.net/
___
pld-devel-en mailing list
pld-devel-en@lists.pld-linux.org
http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn... (Re: SOURCES: ghostscript-afpl-am.patch (NEW), ghostscript-afpl-ijs_pkg...)
On Sat, Sep 10, 2005 at 00:02:45 +0200, Paweł Sakowski wrote:
> Untrue. rpm uses macros. The way the macros are currently defined uses
> SPECS/SOURCES subdirs. A quick-and-unperfect %(pwd) in the appropriate
> place would make it:
>
> svn co .../foo;cd foo;rpmbuild foo.spec
OK, easily aliased.
> > 5. reverting a commit:
> > cvs:cvs up -r X foo; mv foo{,.tmp}; cvs up -A foo; mv -f
> > foo{.tmp,}; cvs ci foo
>
> Easier:
> cvs up -j X -j HEAD foo;cvs ci foo
> (maybe the other way round... I don't remember the exact -j syntax)
I don't remember too, so mv trick is actually easier;)
--
GoTaR gotar>http://vfmg.sourceforge.net/
http://tccs.sourceforge.net/
___
pld-devel-en mailing list
pld-devel-en@lists.pld-linux.org
http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn... (Re: SOURCES: ghostscript-afpl-am.patch (NEW), ghostscript-afpl-ijs_pkg...)
On Fri, Sep 09, 2005 at 23:28:32 +0200, Arkadiusz Miskiewicz wrote:
> Do you know that here you are not talking about cvs vs svn but about how we
> want to organise repository layout?
We don't. rpmbuild uses flat structure and so I want the repos.
> rpmbuild -bb --define _topdir somewhere
>
> because you would have /packages/gcc/trunk/{SPECS,SOURCES}
You are kidding, aren't you? I'm using rpmbuild a hundred times often
than fetching sources and milion times than reverting commit and you
want me to type these defines?
> Note that you are comparing quite ugly, huge and hardly maintainable builder
> script (1787 lines) with svn. Now please do that with cvs client only without
> additional stuff.
This script EXISTS. Give me a handy tool for svn.
> SPECS/SOURCES to nicer scheme and we should talk about that scheme. With new,
IMHO current is nice.
> > 3. fetching all specs:
> > cvs:cvs up
> > svn:
> > for pkg in `svn ls http://.../packages/`; do
> > do whatever you need
> > done
> >
> > ohhh, it really sucks...
> Again, you seem to think svn == /packages/something layout. These are two
> separate things.
So what? I want to checkout all the specs, I'm not interested in sources
or anything else.
> With the /packages you get some things and also loose some
> things.
That's obvious. Particulary I don't get anything from what you're saying
for now.
> > 4. commiting changes:
> > cvs:cvs ci foo.spec; cd ../SOURCES; cvs ci *
> > svn:svn ci foo
> >
> > hmmm, not so big difference, don't talk about atomic commits anymore.
> > Anyway - if we want this functionality, we could move SOURCES and SPECS
> > into higher level repo directory, so that they would be subdirs.
> Atomic is the key here. You are trying to ignore that which is weird.
I don't get it. Why is:
cvs ci SPECS/foo.spec SOURCES/foo.patch
worse than the same in svn?
> > And now NOT COMMON ones:
> >
> > 5. reverting a commit:
> > cvs:cvs up -r X foo; mv foo{,.tmp}; cvs up -A foo; mv -f
> > foo{.tmp,}; cvs
> > ci foo svn: svn revert foo
> >
> > no difference for me. Not an argument.
> You are talking about _single_ file. Try to revert whioe logical commit of
> for
> example update of gcc from 4.0 to 4.0.1.
Do you want me to write you 4-line script?
> It's single command with svn and tens with cvs.
And we all are doing it every day... just use TAGging when working on
100 files.
> > well, I don't care about names, I asked to do this once or twice.
> You seem to do not care about a lot of important things.
Yeah, I'm evil.
> I remember complaints
> from developers that files where copied by cvsadmin not as fast as developers
> wanted.
Oh my god! Really!?
> > 6. access to deleted TAGS
> > cvs:none
> > svn:is
> >
> > who needs it anyway (OK, you need this in DEVEL example, I don't)
> There is HEAD and AC-branch on some package (multiple files). You do merge to
> head and then delete ac-branch.
We're deleting such branches?
> Few days later you want to revert that and...
...and I have never had such situation.
> 7. work on branches
> cvs: problematic if you want to work on several branches of the same files
> (quite common for me, head + ac-branch); can be somehow workarounded by using
> builder-like script
I'm doing it in separate directories.
echo NAC-branch > CVS/Tag
> svn: easy, separate dirs,
Yeah, separate - SPECS, SPECS-devel, SPECS-foo.
> 8. writting scripts for scm like builder
> cvs: problematic, parsing everything as strings
> svn: great python bindings (for example I have repo where there is
And all clear. I don't like python at all;)
> 9. doing diffs
> cvs: problematic if you want to compare whole packages, easy for individual
Hey, we're not working on sources, but on specs and patches.
> Please think about what could be done better.
If you give me tool for things I'm doing right now (as written), I won't
care about background.
--
GoTaR gotar>http://vfmg.sourceforge.net/
http://tccs.sourceforge.net/
___
pld-devel-en mailing list
pld-devel-en@lists.pld-linux.org
http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: jabber c2s dead loop
On Fri, Jul 29, 2005 at 18:53:40 +0200, Tomasz Pala wrote:
> select(0, NULL, NULL, NULL, {0, 1}) = 0 (Timeout)
>
> Only kill -9 helps.
Broken db files. Is there any trustworthy jabberd implementation for
Linux?
--
GoTaR gotar>http://vfmg.sourceforge.net/
http://tccs.sourceforge.net/
___
pld-devel-en mailing list
pld-devel-en@lists.pld-linux.org
http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn...
On Sat, Sep 10, 2005 at 14:48:17 +0200, Paweł Sakowski wrote: > > It can work, but you'd have to checkout rpm module, which contains > > SPECS/ and SOURCES/ > > Yes, but then you have to use the ugly trick of "press Ctrl+C after a > few lines". No. cvs co rpm/SPECS/foo and you have CVS/Entries.Static (or manually touch it, no difference). > It's much cleaner to cvs co SPECS/builder > SOURCES/rpm.macros. It's only one-time cosmetics. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn...
On Sat, Sep 10, 2005 at 14:27:38 +0200, Paweł Sakowski wrote: > > cvs ci SPECS/foo.spec SOURCES/foo.patch > > > > worse than the same in svn? > > Because it won't work with cvs unless you add -d$(cat SPECS/CVS/Root) :P It will. It would require putting SOURCES and SPECS into higher level repo. Just like in: ~/cvs/mplayer: cvs log libdha/config.h mp3lib/dct64.c | wc -l 69 > BTW, that's another reason for directory-per-package. With SPECS/SOURCES > there's no straightforward way to make sure that you're committing > everything you've changed. Indeed. But my zsh tells me which files has been modified after i press tab having 'cvs ci' on command line. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn... (Re: SOURCES: ghostscript-afpl-am.patch (NEW), ghostscript-afpl-ijs_pkg...)
On Sat, Sep 10, 2005 at 11:31:35 +0200, Arkadiusz Miskiewicz wrote: > I dont care what you want. Your help in developing PLD is marginal. Indeed, thanks to people like you. I'm tired of 'hey, who has broken mysql today?' > No, read again. Scripts needs to be written the same way as for cvs. svn it's [...] > As I wrote already - it has to be written _before_ anything can happen. [...] > Use ./builder -f SPECS? Where builder has to be written. > > > Few days later you want to revert that and... > > ...and I have never had such situation. > Because you do (almost) nothing? Yes. And I'm trying not to fuck up things (and leave for somebody else to care like you've done many times, no offence). > > And all clear. I don't like python at all;) > There are other bindings. perl, java. Ok, perl is good. > Ehh, so this whole talk is pointless since such tool NEEDS to be written > first. Here's my proposal: insert a few packages into SVN, start writing some tools and SHOW us (me) how it works, we'll test it and make decision basing on experience not hipothetical situations. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn...
On Sun, Sep 11, 2005 at 12:33:55 +0200, Jakub Piotr Cłapa wrote: > You can use zsh clever autocompletion in our SPECS/? Yes. ~/rpm/SPECS: cvs ci [tab] modified file aptitude.spec lefthand-platform.spec mc-mp.spec sbcl.spec w32codec.spec dpkg.spec man.spec mutt.spec vice.spec wv2.spec giFT.spec maxima.specrpm.spec vim.spec -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn...
On Sun, Sep 11, 2005 at 12:45:02 +0200, Jakub Piotr Cłapa wrote: > IMHO developers need this info mostly on the pld-cvs list and not in the > spec file. (and certainly not in the binary RPM) svn log -r1:HEAD gives In the past rpm has the ability to put into database only part of the changelog (but in other format, not ours) during install. Today: ~: rpm -qa --changelog | wc -c 4519475 And no way to remove it. It's a problem on almost-diskless stations. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn...
On Sun, Sep 11, 2005 at 12:53:52 +0200, Jakub Piotr Cłapa wrote: > > ~/rpm/SPECS: cvs ci [tab] > > It was way too slow on my machine (maybe I'll give it another try now). I don't use it in directory with SPECS/* ;) Only in this, containing CVS/Entries.Static. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: cvs vs svn...
On Sun, Sep 11, 2005 at 15:04:15 +0300, Elan Ruusamäe wrote:
> i guess it's because originally the changelog contained entries started with
> timestamp, but pld has only one such entry (stated with %{date} line).
Yes.
> change %changelog parser to recognize cvs $log timestamps?
It won't help now. This functionality was removed from rpm a year or two
ago.
--
GoTaR gotar>http://vfmg.sourceforge.net/
http://tccs.sourceforge.net/
___
pld-devel-en mailing list
pld-devel-en@lists.pld-linux.org
http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
ftp5.pld-linux.org
It's desynchronized again. No updates for a few days and: error: vfff: http://ftp.pld-linux.org/dists/ac/test/athlon/packages.i/packages.dir.diff.toc.gz: no such file -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ftp5.pld-linux.org
On Fri, Sep 16, 2005 at 10:25:43 +0200, Tomasz Pala wrote: > It's desynchronized again. No updates for a few days and: > > error: vfff: > http://ftp.pld-linux.org/dists/ac/test/athlon/packages.i/packages.dir.diff.toc.gz: > no such file Anyone!? Knock knock! -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
ettercap vs ettercap-ng
ettercap and ettercap-NG crap are not the same projects. The person who replaced the tool with this shitty something (luzik) is requested to revert all the changes in ettercap.spec till rev. 1.47. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ettercap vs ettercap-ng
On Fri, Sep 30, 2005 at 08:57:52 +0200, Marcin Król wrote: > Maybe its not the same project, but ettercap.sourceforge.net says > something different: "Please note ettercap 0.6.x is deprecated; please > upgrade to 0.7.". There's many obsoleted or abandoned projects we use, like gtk+1 or syslog. > Anyway, ettercap-NG may be moved to ettercap-NG.spec, and ettercap.spec > may be reverted to version 0.6.x (which is dead). That's the way it should be done. The person who did this upgrade has probably never used ettercap... -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES: klibc-kill_bashism.patch (REMOVED) - last change reverted...
On Fri, Sep 30, 2005 at 15:40:12 +0200, Arkadiusz Miskiewicz wrote: > [ is builting feature in shells so man test isn't adequate here. test is defined in SUSv3 and should work as said there: http://www.opengroup.org/onlinepubs/009695399/utilities/test.html -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
openoffice-libs-{kde,gtk}
Should't they provide something that's required by openoffice? Now after installing openoffice without them nothing can run. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: openoffice-libs-{kde,gtk}
On Fri, Sep 30, 2005 at 18:25:19 +0200, Arkadiusz Miskiewicz wrote: > > Should't they provide something that's required by openoffice? Now after > > installing openoffice without them nothing can run. > Not in the current form. There is need for -cli subpackage it seems. Explain. I've heard there are native widgets but without any of gtk/kde libs I can't run anything. Maybe only some wrapper should be fixed? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ettercap vs ettercap-ng
On Sat, Oct 01, 2005 at 00:44:56 +0200, Andrzej 'The Undefined' Dopierała wrote: > > That's the way it should be done. The person who did this upgrade has > > probably never used ettercap... > you mean luzik or paszczus? ;) The former one. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Where has quagga gone!?
I can't see it AC. STBR -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
ftp5 is down again
If it's PLD quality, I have no more comments. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: Where has quagga gone!?
poldek --up install net-snmp-5.1.2-6 error: vfff: /dists/ac/PLD/i686/PLD/RPMS/net-snmp-5.1.2-6.i686.rpm: no such file -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: dspam.spec - use debug instead of verbose, - updated switches.
On Tue, Oct 11, 2005 at 06:38:15PM +0200, Arkadiusz Miskiewicz wrote: > > - use debug instead of verbose, > debug is different than verbose here! verbose is fine for running package on > production - it just gives more logs. Hmmm... --enable-debug Enable debugging support for DSPAM. Don't enable this unless something needs testing! --enable-verbose-debug Cause DSPAM produce verbose debug output and write them into LOGDIR/dspam.debug file. Implies '--enable-debug'. Never enable this for production builds! I'm not quite sure. -- Tom Pala <[EMAIL PROTECTED]> http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
i686 poldek pndir indexes
athlon and amd64 are ok, while i686: Retrieving ac::packages.ndir.md... Retrieving ac::packages.ndir.gz... .. 100.0% [9.2M (363.2K/s)] error: http://ftp.pld-linux.org/dists/ac/[...]/packages.ndir.gz: broken file Retrieving ac-ready::packages.ndir.md... Retrieving ac-ready::packages.ndir.gz... .. 100.0% [327.8K (93.1K/s)] error: http://ftp.pld-linux.org/dists/ac/ready/i686/packages.ndir.gz: broken file Retrieving ac-test::packages.ndir.md... Retrieving ac-test::packages.ndir.gz... .. 100.0% [69.6K (18.4K/s)] error: http://ftp.pld-linux.org/dists/ac/test/i686/packages.ndir.gz: broken file -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: i686 poldek pndir indexes
Is it possible to fix it in this year!? On Sat, Oct 15, 2005 at 12:09:33 +0200, Tomasz Pala wrote: > athlon and amd64 are ok, while i686: > > Retrieving ac::packages.ndir.md... > Retrieving ac::packages.ndir.gz... > .. 100.0% [9.2M (363.2K/s)] > error: http://ftp.pld-linux.org/dists/ac/[...]/packages.ndir.gz: broken file > > Retrieving ac-ready::packages.ndir.md... > Retrieving ac-ready::packages.ndir.gz... > .. 100.0% [327.8K (93.1K/s)] > error: http://ftp.pld-linux.org/dists/ac/ready/i686/packages.ndir.gz: broken > file > > Retrieving ac-test::packages.ndir.md... > Retrieving ac-test::packages.ndir.gz... > .. 100.0% [69.6K (18.4K/s)] > error: http://ftp.pld-linux.org/dists/ac/test/i686/packages.ndir.gz: broken > file -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: i686 poldek pndir indexes
error: http://ftp.pld-linux.org/dists/ac/[...]/packages.ndir.gz: broken file warn: http://ftp.pld-linux.org/dists/ac/PLD/i686/PLD/RPMS/: load failed, skipped error: http://ftp.pld-linux.org/dists/ac/ready/i686/packages.ndir.gz: broken file warn: http://ftp.pld-linux.org/dists/ac/ready/i686/: load failed, skipped -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: i686 poldek pndir indexes
On Wed, Oct 19, 2005 at 01:56:26 +0200, Pawel A. Gajda wrote: > $ rpm -q poldek > poldek-0.19.0-1.20051010.17.1 I had got poldek-0.19.0-0.20050613.22.1 and this was a problem, thanks. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
smokeping
Revision 1.26 2005/10/26 13:56:02 blues - started 2.0.x series - this is needed for current rrdtool It's not. 1.43 I've upgraded to works as well. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: smokeping
On Sun, Oct 30, 2005 at 23:57:00 +0100, Paweł Gołaszewski wrote: > > Revision 1.26 2005/10/26 13:56:02 blues > > - started 2.0.x series - this is needed for current rrdtool > > > > It's not. 1.43 I've upgraded to works as well. > > It is. It's not;) > Try to run smokeping command having rrdtool 1.2 I've got two running. > I don't remember exactly but it was written that rrdtool is not supported > and if I want to use rrdtool-1.2.x you have to use 2.x series. 2005/5/31 -- 1.42 * make it work with the --statis option and PROBE_CONF -- Niko * make it work with rrdtool 1.2.x (technically) ! -- Niko > It was impossible to even start daemon. You are talking about <=1.41 (the point of forking 2.0). -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: smokeping
On Mon, Oct 31, 2005 at 00:21:22 +0100, Paweł Gołaszewski wrote: > I've installed from ftp and after that I've made changes. Indeed, the one from ftp didn't work. Thus I've made an upgrade;) > OTOH - what is the point to keep smokeping 1.x ? Dunno. BTW do you know how to disable antialiasing in rrdtool 1.2? I don't like my stats anymore with this crap enabled:/ -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: smokeping
On Mon, Oct 31, 2005 at 10:51:25 +0100, Paweł Gołaszewski wrote: > > BTW do you know how to disable antialiasing in rrdtool 1.2? I don't like > > my stats anymore with this crap enabled:/ > > I don't know. I like it anyway :) Hmmm... rrd_graph.c: case 'R': if ( strcmp( optarg, "normal" ) == 0 ) im->canvas->aa_type = AA_NORMAL; else if ( strcmp( optarg, "light" ) == 0 ) im->canvas->aa_type = AA_LIGHT; else if ( strcmp( optarg, "mono" ) == 0 ) im->canvas->aa_type = AA_NONE; :) -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: New task #5722 in pdksh: build-in echo broken
On Wed, Nov 02, 2005 at 06:16:34PM +0100, Jacek Konieczny wrote: > > Any idea how to fix the script in a portable way, so it outputs > a defined number of backslashes? Always use -e. > And bash and our /bin/echo are broken too? It's considered to be feature:/ -- Tom Pala <[EMAIL PROTECTED]> http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
logrotate maillog
I've got:
/var/log/maillog {
delaycompress
create 640 root logs
postrotate
/etc/rc.d/init.d/syslog-ng reload >/dev/null 2>&1
endscript
}
in /etc/logrotate.d/syslog-ng, but after every rotation maillog is 600
root:root. Any ideas why?
--
GoTaR gotar>http://vfmg.sourceforge.net/
http://tccs.sourceforge.net/
___
pld-devel-en mailing list
pld-devel-en@lists.pld-linux.org
http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: logrotate maillog
On Sun, Nov 06, 2005 at 17:56:30 +0200, Elan Ruusamäe wrote: > maybe new log is created by syslog-ng, not logrotate? syslog-ng shouldn't create file before SIGHUP. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: logrotate maillog
On Sun, Nov 06, 2005 at 18:53:39 +0100, Jacek Konieczny wrote:
> > maybe new log is created by syslog-ng, not logrotate?
>
> Yes, it is.
When?
1. mv maillog archive/maillog.1 (delaycompress)
2. touch maillog; chown root:logs maillog; chmod 640 maillog (create)
('Immediately after rotation (before the postrotate script is run)')
3. SIGHUP syslog-ng (postrotate)
> And syslog-ng should also be configured to set right
> permissions.
I'll try:
/var/log/maillog {
delaycompress
create 640 root logs
postrotate
/etc/rc.d/init.d/syslog-ng reload >/dev/null 2>&1
chown 640 /var/log/maillog
chgrp logs /var/log/maillog
endscript
}
--
GoTaR gotar>http://vfmg.sourceforge.net/
http://tccs.sourceforge.net/
___
pld-devel-en mailing list
pld-devel-en@lists.pld-linux.org
http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
w32codec: get_version
It finds version as ppc-20051120 instead of 20051120. Ankry? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: w32codec: get_version
On Sat, Nov 26, 2005 at 14:08:11 +0100, Andrzej Krzysztofowicz wrote: > > It finds version as ppc-20051120 instead of 20051120. Ankry? ...instead of 20050412. Now there are ppc codecs with their own dates. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: w32codec: get_version
On Sun, Nov 27, 2005 at 15:55:16 +0200, Elan Ruusamäe wrote: > > ...instead of 20050412. Now there are ppc codecs with their own dates. > what are these? (i have no access to any ppc) ftp://ftp1.mplayerhq.hu/MPlayer/releases/codecs/ChangeLog 20051120 * essential-ppc - all codecs that don't have a free replacement on PPC Linux * all-ppc - all codecs that work on PPC Linux > those should be perhaps included for %ifarch ppc package? (instead of current > ones, or additionally?) Instead of. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
fly.mif.pg.gda.pl ftp desynchronized
ftp.pld-linux.org at 153.19.42.122 has no updates since 23.11. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: fly.mif.pg.gda.pl ftp desynchronized
On Sat, Dec 03, 2005 at 13:21:05 +0100, Andrzej Krzysztofowicz wrote: > > ftp.pld-linux.org at 153.19.42.122 has no updates since 23.11. > > Strange. As rsync runs from cron every hour and noone is "running" now. ~: wget -q -O - http://ftp1.pld-linux.org/dists/2.0/ready/athlon/packages.i/ | grep "packages.ndir.2005.11.2" [...] packages.ndir.2005.1..> 29-Nov-2005 16:51 1.7K packages.ndir.2005.1..> 29-Nov-2005 17:51 3.6K packages.ndir.2005.1..> 29-Nov-2005 18:21 15K ftp2, ftp3: packages.ndir.2005.1..> 23-Nov-2005 22:21 6.5K GZIP compressed docume> packages.ndir.2005.1..> 23-Nov-2005 23:51 27K GZIP compressed docume> packages.ndir.2005.1..> 24-Nov-2005 01:21 235K GZIP compressed docume> > Access problems? ftp4 and ftp5 gives me 'connection refused'. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: fly.mif.pg.gda.pl ftp desynchronized
On Sat, Dec 03, 2005 at 14:07:54 +0100, Andrzej Krzysztofowicz wrote: > > > href="packages.ndir.2005.11.23-19.21.09.gz">packages.ndir.2005.1..> > > 23-Nov-2005 22:21 6.5K GZIP compressed docume> > > ftp2 does not support native http access. > It redirects to another one. > > Check more carefully. Nevertheless it redirects to broken site. So is ftp3 going to be fixed soon, or do I have to manually switch DNS once again to get some fresh stuff? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Sat, Dec 10, 2005 at 23:53:04 +0100, Arkadiusz Miskiewicz wrote: > rc-scripts and replacing it with something that could supervise services, > start things based on dependencies (not priorites) but at this moment no one Small chances for such think to work, as many of our services shall return HGW instead of DONE. Moreover - how to set dependencies when packager don't know e.g. network configuration? Does 'bind' require running 'zebra*', or is 'network' sufficient? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Sun, Dec 11, 2005 at 01:23:12 +0100, Arkadiusz Miskiewicz wrote: > > Small chances for such think to work, as many of our services shall > > return HGW instead of DONE. > Huh? You seem to have no idea how supervising is usually done. The processes Indeed. How long does it wait until assumes that service is running (it should wait to eliminate risk of races - init part of one program can take longer than starting another)? What does it do when required service dies after some other (which depends on it) has already started (or is starting)? - I'm particullary interested in service restart/reload, as they're triggered automatically from logrotate. > are run in foreground as childs of supervising process. If it dies then > process knows about that. Mmhhhmm... 'background'? Hm, why don't we use such a 'supervising process' now? Date: Thu, 2 Sep 2004 21:11:32 +0200 Subject: service start Message-ID: <[EMAIL PROTECTED]> This problem occours not only with bind, AFAIR I had the same with quagga (when logging via files not syslog - they still have wrong permissions in AC), ircd, jabber (broken database), innd (indexes from other arch, innwatch counting free i-nodes on reiserfs), mailgraph (no read access to logs), mysql (many many times, I've seen DIED while server was started, and DONE when died...). In short: some services can start just-to-die-SOON giving DONE. > No idea. Maybe bind requires network while zebra becomes part of network when > it's installed? bind requires interfaces (to bind to) - that's for sure. But how about routing? Dependency means 'can stand up and WILL work if...' (bind to interfaces, wait for routing and network access) or 'can do it's job NOW' (needs routing for serving requests)? > No one started the project means almost everything is > unknown. And I doubt someone will. I won't put it into production machine, because don't want to end up with everything shut down because of log rotation while temporary network shutdown (by the way - I've got quagga-ospf dying a moment after configuration changes...). After all - what are the advantages? Faster startup (due to backgrounding jobs)? Sounds coming from my disks say no (maybe if someone has broken DNS, didn't initialize postgresql database or has some other misconfiguration). Dependencies? Well, in fact we can imagine a situation when virtually every service works for it's own without anything else. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Sun, Dec 11, 2005 at 14:06:39 +0100, Arkadiusz Miskiewicz wrote: > background for user but foreground for supervising process. We don't use it I see. However I'm afraid there will be problems with services starting a few different processes, like jabber (router, resolver, sn, c2s, s2s). > I don't care about faster startup. I care only about one thing - start > service > again if it died. Such watchdog can be simply created by croning `service [/etc/init.d/*] status`. Paralell startup can be nice feature on desktops - starting [kg]dm and allowlogin before network/MTA/donkey/apache/*sql -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Sun, Dec 11, 2005 at 08:52:36 -0500, Andrew A. Gill wrote: > That's a very ugly solution. I haven't been following this > init.d discussion too closely, but what you're proposing is > ugly. Daemons shouldn't die and they don't without any reason. I've got only one broken - ospf, if you have more you should revise system configuration and hardware. Generally speaking there's no common need for watching over services, thus I don't see much interest in switching startup scripts. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Sun, Dec 11, 2005 at 16:10:19 +0100, Jakub Piotr Cłapa wrote: > The startup time with init-ng is proved to be less than a half compared > to plain SysVinit. With 3-6 months uptimes it's irrevelant if restart due to new kernel takes 5 or 10 minutes. My conclusion coming from this discussion is that's desktop solution, not server one. Am I right? -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Sun, Dec 11, 2005 at 18:56:28 +0100, Jacek Konieczny wrote: > Irrevelant? My customers _do_ care if a maintenance break (e.g. reboot > during kernel upgrate) lasts 5 or 10 minutes. If so, why don't you use some HA? Anyway I simply do not believe, that it's 'proven' to be two times faster, especially when speaking of machines which have some real load. > the machine boots. It is much harder to explain thousands of customers, > calling during machine boot, why their services don't work. $ wc -l /etc/passwd 1634 /etc/passwd I don't have this problem fortunatelly;) -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Sun, Dec 11, 2005 at 19:01:55 +0100, Jacek Konieczny wrote: > That would be in an ideal world. Daemons have resource leaks and will be To my information: which one? > killed when reach the resource limit (with process supervision ulimit > becomes very usefull). Hmmm... what's with ulimit without supervision? > They may crash on some unexpected event. That may > be after months of uninterrupted work. It is usually better if they come > back again if possible, then if they are down until the administrator > manually start them again. In ideal world;) I'd be afraid of possible bugs in supervisor, but will be really happy if people like you will test it sometime;> -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Sun, Dec 11, 2005 at 21:29:59 +0100, [EMAIL PROTECTED] wrote: > These two alternatives are where we should search for a way out. Or we > hold an enviroment to build patches and so on (the question is how long > and how many steps back). > > The '???' alternative sound great - but we have to invet it :> It already is - distro lines. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Sun, Dec 11, 2005 at 19:55:39 +0100, [EMAIL PROTECTED] wrote: > The equasion is simple - I see new functionality added, few new fatures, > faster boot and nothing taken - so, where is the problem? You get the same > thing, that earlier, but faster and better - where is the problem? So please answer to all my issues from other mails (network/quagga/routing concern, service reload from logrotate...). The problem is: - unpredictable behaviour upon system failure (I'd prefer not to restart *sql after some recoverable cluster error, because it may hide problem and all I get are problems with synchronizing timelines), - more failure points (now service means service, not also some deps), - more complicated management scheme - how to TEST my own services on production machine? Playing doom-admin? Check all the scenarios? It's nice solution for fast system startup (but I doubt it's signifficant change - do me a favour and add '&' at the end of /etc/rc.d/rc line 213 to check it please). Process supervising if configurable can be nice feature too, but for me it's not worth risking mass murder by supervisor bug. I do trust init and that's all. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Mon, Dec 12, 2005 at 02:06:39 +0100, Andrzej Krzysztofowicz wrote: > > mass murder by supervisor bug. I do trust init and that's all. > > Maybe you need just to have a choice ? Exactly. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Mon, Dec 12, 2005 at 22:48:45 +0100, Jakub Piotr Cłapa wrote: > http://initng.thinktux.net/index.php/Boot_charts_Official > > First check, than argue. With pleasure. Just tell me what to install and I will do it on one of my backup systems. > remain simillar. The simplest reason I can think of (there are other > too) is that there are deamons whose startup is CPU bound (Apache? I'm aware of the reason and don't think I'll profit much. Well, unless we're talking about 486 and floppy... -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Mon, Dec 12, 2005 at 23:10:40 +0100, Jakub Piotr Cłapa wrote: > We can have ready and test. Ready could be then used for bug-fixes and > critical updates before the whole new environment is complete. Just ONE thing: where would you inject gcc4? Second? I've got BIG problem with mysql in AC, because I can't upgrade 4.0.x to anything above on production machine. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Tue, Dec 13, 2005 at 00:32:23 +0100, Arkadiusz Miskiewicz wrote: > You are free to create mysql40.spec in the way that it can coexist with > current mysql. I'm waiting for glen with his mysql4 ;> And as it's to be done in my job I'm waiting for agreement from my employer to license/sth. BTW someone knows what should be in there? > ps. I also still use 4.0 with hundreds of databases without any problem on > current AC. Just updates are done manually. ~: rpm -qa \*mysql\* | sort mysql-4.0.23-3 mysql-client-4.0.23-3 mysql-devel-4.1.14-2 mysql-libs-4.0.23-3 mysql-libs-4.1.14-2 mysql-libs-5.0.15-5 mysqlstat-0.0.0.4-2 mysqlstat-cgi-0.0.0.4-2 perl-DBD-mysql-3.0002-1 php-mysql-5.0.5-17 I hope it won't stop working someday. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Tue, Dec 13, 2005 at 02:55:35PM +0100, [EMAIL PROTECTED] wrote: > > I'm testing it right now on my laptop( yea - it doesn't have this much > load, but it runnes few server-like daemons, like apache and mysql + pgsql > ). The boot speeds up *really*! X starts in like < 10sec from powering on > the laptop! I had already written that it can be nice feature on desktops. But when your system becomes operational? Means working services - that's what customers may be interested with. > trinary > more. Why? Example: I've forgot to connect my pen and mount > outputed an error, which resulted in failure of running the mounting init > script, which stopped > 50% of the system from running, which is insane! Hiehheiehe, I knew it;) -- Tom Pala <[EMAIL PROTECTED]> http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Tue, Dec 13, 2005 at 09:10:33PM +0100, [EMAIL PROTECTED] wrote: > > > I had already written that it can be nice feature on desktops. But when > > your system becomes operational? Means working services - that's what > > customers may be interested with. > > Are You sure You've read previous posts? Are you sure you've read mine? Don't you understand what means RUNNING service, SERVING requests? It's not the same as 'I can login and write blablabla in shell'. Parallel running via backgrounding jobs gives better usage of resources when you've got different resource-consuming services. And only then. A ratio depends on what you're running and 'proven 2 times faster' can be ability to LOGIN to machine, not becoming fully OPERATIONAL. All customers don't give a shit if you can login or not. > Give me a break dude - It works, it'll work better and faster, But sometimes something get killed or won't start at all. Yeah! > it is more configurable, we It IS? > just need to work on it and switch old init scripts to new ones - is that > what You're reffering to? So what are you waiting for? Don't waste time talking with me, just do it, show me. For now all what you're doing is 'yeah, yeah, it is cool, fast, better, and you are wrong having doubts'. > You won't be moving to th after dec.20, when AC will be frozen - that's > obvoius, so I don't understand, what are You complaining bout? It's pretty Nothing. I'm just against gentooism like `it's two times faster'. > much obvious, that th will be unstable and not all functionality will be > done just after freezing AC, but that isn't the point here - or is it? The > point is to use the oportunity of new line to implement some new, better If it is better, answer one of my question: how would you start bind in relation with quagga? > officialy a part of PLD? Is it used by default? So what are You > complaining bout? That beta software doesn't work 100% correct? There are How can it work, when NOONE can tell me it's rules. How to start bind? > even no nativ initng scripts in PLD yet and You jump to conclusion, that > they are bad. If something works 'free style', and my services can randomly become unavailable, it IS bad. > Have You even bothered to look up what initng is, or to read > throughly theese posts? You didn't bother to read mine. Or you don't understand them. If you want parallel services running just add '&' in proper place (as you read my mails you know exactly where). All I want to know is: how dependencies work (will my services be killed by logrotate service reload after fatal error in quagga?). -- Tom Pala <[EMAIL PROTECTED]> http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Tue, Dec 13, 2005 at 01:09:11 +0100, Jakub Piotr Cłapa wrote: > You could give it a try: bootchart.org Nice, but I don't have a place to run it;/ -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Tue, Dec 13, 2005 at 18:14:48 +0100, Jakub Piotr Cłapa wrote: > boost from concurrency since most programs/daemons don't depend on each > other and don't need to wait for others to start first. If they don't depend, why do we have priorities in current init scripts? If we abandon priorities, we must introduce dependencies. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Wed, Dec 14, 2005 at 16:53:13 +0100, [EMAIL PROTECTED] wrote: > that, because it doesn't work yet and that's why the whole idea is wrong No, I'm not saying. I'm STILL asking some 'HOW' questions and don't get answers. > to start with". There are possibilites which we could utilize and give me > a good reason, good retorical reason, why shouldn't we use that. So - what I'll repeat one more time, slowly: b e c a u s e n o o n e c a n t e l l m e H O W t o s t a r t n e t w o r k s e r v i c e s . > wan't work when we're done scripting, what is impossible because of the > nature/architecture of initng and so on. So far You've just claimed the How do you want to script, when you don't know WHAT is to be written? > obvoius - it isn't ready and productional yet, but the question is, should It's not even ready for testing, because apparently there's no fundamentals (or any person here who knows something more than 'parallel'). > > Nothing. I'm just against gentooism like `it's two times faster'. > > Do I look to You like a Gentoo user? Yes. You believe that something you don't know is better, because you've read it somewhere. > I've just written down my personal experience with only slightly tweaked > initng with PLD and the result is as follows - boot time, including *all* > services and daemons, speed up dramatically, probably more than 2 times! I don't care about your system. I'm worried about MINE and you can't tell me how to configure it to work. > > If it is better, answer one of my question: how would you start bind in > > relation with quagga? > > I don't know what quogga is, Sad... > but I don't see any service would be > unrunable in the new initng. Without any non-zero interface-interval bind will run, work, but won't listen on zebra set-up interfaces, so won't serve request. > I'm not ignorant so much, to say it doesn't > matter. That seems to be the first con not to make it in PLD, so please > tell me, what is so speciffic bout it and we'll think is it really such an > obstacle. Just tell me how dependencies work and I will know everything. I'm trying to get this information for 4 days. > > How can it work, when NOONE can tell me it's rules. How to start bind? > > service named start :D Manually after every system restart? ;P > > If something works 'free style', and my services can randomly become > > unavailable, it IS bad. > > Jumping to conclussions again? Why randomly become down? Assume that my OSPF dies (it happens with PLD quagga). It's not restarted by supervisor because I've disabled this function (Q1: can in be disabled? Q2: can in be disabled selectively?). Now it's time to rotate logs - postrotate script does service reload and all my services are shut down, because there's no required OSPF daemon running (Q0: how dependencies work, what's configurable?). > Why free style? No rules==free style (I'm asking for these rules since monday, if YOU are so interested in merging this project into PLD why don't you know them?). > Has anyone here written bout random restarts, random services going down. Yes, me. I've written the scenario a few times and got no answer. > It's up to how we write the scripts. That's a large field to discuss. Also > the trinary or larger logic - is it neceserry, how to ensure proper boot > even after errors, what gets restarted and when. Well, if everything is TODO, what can I test or comment? EOT until we'll have something to discuss about. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Thu, Dec 15, 2005 at 18:40:24 +0100, Tomasz Grobelny wrote: > Why don't you guys try to write down _requirements_ ie. list of properties > that the new system must have before introducing it to PLD? Because some people tried to convince me, that it is done and we must only adapt some scripting to PLD. So if it exists I wanted to know how it works. It turned out that there's noone here who knows anything more than 'parallel' and 'dependencies'. So how do you people want to put it in into PLD, when you don't have any idea how to maintain the system? > Once more: the idea is to write down WHAT IS NEEDED and then think if it's > doable. Or maybe just someone interested will check initNG or other really existing projects? ;> -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Fri, Dec 16, 2005 at 16:14:34 -0700, Aredridel wrote: > > Have you seen a distro that supports full machine upgrade (incl. > > configuration fixes/rewrites) in less than one night ? > > Let's be the first. Take a look at amavisd-new config file - it's de facto a piece of perl code you must rewrite. Have you got any idea how to automate process of upgrading it? It takes me about an hour. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ANN: Closing AC
On Sat, Dec 17, 2005 at 20:27:11 +0100, Marcin Król wrote: > No, you don't need to rewrite it with each upgrade. I'm using Unless you don't want to have troubles someday, when you'll have to merge dozens of changes;) > my fault, I've just deleted wrong file :)). It still works as it should > with ~2 year old config. Do you have some advanced configuration? You depend on configuration compatibility, which can end up anytime. BTW 2 years ago things like: @score_sender_maps = ( [...] new_RE( [ qr'[EMAIL PROTECTED]'i => [ new_RE( [ '.' => -10.0 ] ) ], ] ) , [...] ) $per_recip_whitelist_sender_lookup_tables = ( new_RE( [ qr'[EMAIL PROTECTED]'i => [qw( customer.com )] ] ) ); didn't work I suppose. > Of course minor changes were required but those > usually take ~15 minutes. Mostly banned files updates... It takes me 15 minutes (per system) too, thanks to diff; $ ls -1 /etc/amavis* /etc/amavisd.conf /etc/amavisd.conf-20030616-9 /etc/amavisd.conf-20030616-9.rpm /etc/amavisd.conf-20040806-2 /etc/amavisd.conf-20040806-2.rpm /etc/amavisd.conf-20040906-1 /etc/amavisd.conf-20040906-1.rpm /etc/amavisd.conf-2.2.0-2 /etc/amavisd.conf-2.2.0-2.1 /etc/amavisd.conf-2.2.0-2.rpm /etc/amavisd.conf-2.2.1-1 /etc/amavisd.conf-2.2.1-1.rpm /etc/amavisd.conf-2.2.1-3 /etc/amavisd.conf-2.2.1-3.rpm /etc/amavisd.conf-2.3.3-5 /etc/amavisd.conf-2.3.3-5.rpm /etc/amavisd.conf.rpmnew but I maintain several of them. -- GoTaR gotar>http://vfmg.sourceforge.net/ http://tccs.sourceforge.net/ ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
