Re: RHEL 3.0 or Fedora for a 'production environment' ?
by Mike Lurk
From: Alan Cox <alan(a)redhat.com>
Subject: Re: RHEL 3.0 or Fedora for a 'production environment' ?
To: fedora-test-list(a)redhat.com
Date: Mon, 3 Nov 2003 13:45:49 -0500 (EST)
Reply-To: fedora-test-list(a)redhat.com
> And this is why I know of a few data centers already considering the
move
> to slackware for all new boxes and gradual replacements over time, I
> think RH have lost a lot of friends with this move, the days of RH
When you try maintaining software for years to the standards Red Hat
does it turns out to be quite costly. Some people will choose to take
that cost themselves, others may choose to buy RH or other business
oriented
products.
> populating data centers are gone, with a support of only a few months
> fedora will never see its way in there, can you imagine a center with
> hundreds and hundreds of boxes runing around replaceing it every 3
> months?
Fedora is aimed at developers - people who contribute and want cool
stuff
rather than people who simply want the cheapest possible business deal
without regard to quality. There is always a market it isnt worth
addressing
- the people who won't even pay the cost of the service they provide.
> I cant, some of our servers are still running 7.3, cause we knew RH
will
> put out errata if its seriou enough, general updates who cares, get
the
> tarball, looks like thats how we'll be updateing untill we decide
which
> way to go, update by tarballs in future, or move onto slackware or
suse.
Your right, and if the business economics work for you then nobody would
expect you to do anything different.
Alan
I am looking another OS besides Windows to use. Microsoft licensing
issues. Fedora fits the bill.
Mike
20 years, 5 months
Broken Fedora mozilla ppc packages (and the fix)
by Joel Barrios
Yes, I know it's really crazy, but I wanted to test the fedora PPC
stuff.
After installing PPC packages from Fedora in an iMac DV (I just avoided
initscripts, devlabel, gpm and mkinitrd), I was wondering why everything
worked fine except for mozilla/epiphany. I even rebuilt from SRPMs
several 1.4.1 SRPMs releases with 100% of negative results, until I
examined the mozilla spec file and found out why Fedora's mozilla RPMs
for PPC are broken:
# build mozilla - XCFLAGS is for NSS which can't use CFLAGS for
# various reasons
%ifarch ia64
OPTFLAGS=-O0
%else
OPTFLAGS=-O2
%endif
XCFLAGS=-g
CFLAGS=-g
%ifarch ia64
CXXFLAGS="-fno-inline -g"
%else
CXXFLAGS=-g
%endif
Guess the PPC mozilla packages were built just doing rpmbuild --rebuild.
So, <http://ftp.redhat.com/pub/redhat/linux/rawhide/ppc/Fedora/RPMS/>
has mozilla _PPC_ binary pacakges built with OPTFLAGS=-O2 and
CXXFLAGS=-g (!?!).
After removing these lines from the spec file, that really should
consider other architectures, and doing rpmbuild -bb, I could
successfully build mozilla packages that actually _work_ on PPC.
Joel Barrios D.
Linux Para Todos
http://www.linuxparatodos.net/
20 years, 5 months
Status of bug (rfe?) 107834
by Peter Boy
There are still some probs due to the system wide usage of uft8. One of
them is addressed in bug 107834 regarding characters above ascii 128 in
Filenames (e.g. samba shares with windows clients).
Similar problems arise, if you use evolution to exchange mail with a
windows user and even with a linux guy not using Red Hat. Characters
above ascii 128 look "silly" and are unreadable.
Of course, you can change the system default in /etc/sysconfig/i18n to
use another locale. This solves some problems but rises others.
Localised output of some local programs get unreadable and/or look quit
ugly.
@Redhat: are there concrete planings to address these issues which
existent since introduction of utf8 as a system default?
20 years, 5 months
Fedora and RHSAs?
by The Matt
Does anyone know if the cups* and coreutils RHSAs that are appearing
today affect the current Fedora RPMs?
up2date/yum don't show any updates for them, but with the freeze that
seems to be on I don't know how to take that sign.
If they are vulnerable, is there a way to shore up my box until the
updates appear?
Matt
--
I am a theoretical chemist. Fear me! Please.
Matt Thompson -- http://ucsub.colorado.edu/~thompsma/
440 UCB, Boulder, CO 80309-0440
JILA A510, 303-492-4662
20 years, 5 months
Re: Release on monday?
by Patrick
Good point!
--- On Mon 11/03, seth vidal < skvidal(a)phy.duke.edu > wrote:
From: seth vidal [mailto: skvidal(a)phy.duke.edu]
To: fedora-test-list(a)redhat.com
Date: 03 Nov 2003 10:46:29 -0500
Subject: Re: Release on monday?
releasing things on fridays suck. You get little publicity and media coverage.
_______________________________________________
No banners. No pop-ups. No kidding.
Introducing My Way - http://www.myway.com
20 years, 5 months
TUX on test 3
by Dima Gutzeit
Hi everyone ,
I've been trying to run the TUX and that what I got :
Starting tux: /etc/init.d/tux: line 69: /proc/sys/net/tux/threads: No such file or directory
/etc/init.d/tux: line 70: /proc/sys/net/tux/documentroot: No such file or directory
/etc/init.d/tux: line 71: /proc/sys/net/tux/cgi_uid: No such file or directory
/etc/init.d/tux: line 72: /proc/sys/net/tux/cgi_gid: No such file or directory
/etc/init.d/tux: line 73: /proc/sys/net/tux/cgiroot: No such file or directory
/etc/init.d/tux: line 75: /proc/sys/net/tux/keepalive_timeout: No such file or directory
/etc/init.d/tux: line 77: /proc/sys/net/tux/logging: No such file or directory
/etc/init.d/tux: line 78: /proc/sys/net/tux/logfile: No such file or directory
Tux version: 2/0/0
TUX protocol version mismatch, kernel 2 != daemon 3
[FAILED]
Does anyone has ideas ?
Thanks
20 years, 5 months
Is anyone using postfix and pine??
by Tom Diehl
Hi,
I am trying to get postfix and pine working together using the postfix
rpm that comes with severn and Mike Harris's pine rpm. I can send a message
from the command line just fine but if I try to use pine it will not send.
I get the following in the logs:
Nov 3 00:33:53 tigger postfix/smtpd[2143]: connect from localhost[127.0.0.1]
Nov 3 00:33:53 tigger postfix/smtpd[2143]: 761723CC2: client=localhost[127.0.0.1]
Nov 3 00:33:53 tigger postfix/postdrop[2144]: fatal: uid=508: unexpected record type: 84
Nov 3 00:33:54 tigger postfix/smtpd[2143]: warning: premature end-of-input on /usr/sbin/postdrop while reading input attribute name
Nov 3 00:35:13 tigger postfix/smtpd[2042]: > localhost[127.0.0.1]: 421 Error: timeout exceeded
Nov 3 00:35:14 tigger postfix/smtpd[2042]: vstream_fflush_some: fd 1 flush 29
Nov 3 00:35:14 tigger postfix/smtpd[2042]: timeout after MAIL from localhost[127.0.0.1]
Nov 3 00:35:14 tigger postfix/smtpd[2042]: vstream_fflush_some: fd 1 flush 0
Nov 3 00:35:14 tigger postfix/smtpd[2042]: vstream_fflush_some: fd 5 flush 33
Nov 3 00:35:14 tigger postfix/postdrop[2043]: fatal: uid=508: unexpected record type: 84
Nov 3 00:35:14 tigger postfix/postdrop[2043]: remove maildrop/4E3623CC0
Nov 3 00:35:15 tigger postfix/smtpd[2042]: disconnect from localhost[127.0.0.1]
Google says that sendmail and postdrop binaries are from different postfix versions
but I rm'd the postfix rpm and reinstalled it. My pine config file is the same one
I have been using for a couple of years. It works fine on 8.0.
Anyone else tried this??
--
......Tom Registered Linux User #14522 http://counter.li.org
tdiehl(a)rogueind.com My current SpamTrap -------> mtd123(a)rogueind.com
20 years, 5 months
Re: Bluetooth and 2.4.22-1.2115.nptl problems?
by Daniel Durgin
On Sun, 2003-11-02 at 08:01, Daniel Durgin wrote:
> At boot up I get a messege about lsmod not being
able
> to load 2.4.22-1.2115.nptl bluetooth modules,
because
> they already exist.
>I get the messages too, but they don't appear to
cause >a problem with
>bluetooth -- that I've come across yet, anyway.
False alarm. I need to change Iptables a little all
set now.
Thanks
Dan
__________________________________
Do you Yahoo!?
Exclusive Video Premiere - Britney Spears
http://launch.yahoo.com/promos/britneyspears/
20 years, 5 months
Fedora compiler question
by Peter Boy
I remember having read in previous messages, that the Fedora 2.4 kernel
is compiled using a different gcc version as the default. So, if I would
like to recompile the kernel or to compile additional modules for the
kernel I have to take provisions to use that older version.
Unfortunately I can't remember the details and could find detailed
descriptions in the archiv.
Is this still true and could someone give me the necessary information?
Thanks
Peter
20 years, 5 months
Re: How to Turn Off ECN [was: internet not working with arjan's 2.6 testkernels]
by Paul Morgan
Tim Kossack asked:
>hm, i followed michael's advice (thank you very much) and at least my
>laptop is working now when i turn off ECN.
>however, it seems that i have to newly turn it off every time i
>reboot-is this normal?
>is there a way to switch it off forever?
Add the following line to /etc/sysctl.conf
net.ipv4.tcp_ecn = 1
RHL/Fedora reads this file on every startup and adjusts settings
accordingly. If you make changes to /etc/sysctl.conf, you can run
`sysctl -p` to force the system to re-read the file without a reboot.
As root you can run `sysctl -a | less` to read all about the other
settings available through this cool facility (such as for firewalls,
etc.). I prefer this as opposed to echoing to proc, but it may not be
portable to other flavors of Linux. sysctl.conf settings take effect
earlier than they would if placed in rc.local.
HTH,
Paul
20 years, 5 months