users December 2005

users@lists.fedoraproject.org

673 participants
1095 discussions

Did I miss a PHP update for FC4?
by Gilbert Sebenste 03 Dec '05

03 Dec '05

All, A few weeks ago some major security holes were announced in regards to the PHP packages. Was there an update I missed? I see the current version on a system I am running it on is php-5.0.4-10.5. ******************************************************************************* Gilbert Sebenste ******** (My opinions only!) ****** *******************************************************************************

3 2

changing default gcc include path
by Chris Jones 03 Dec '05

03 Dec '05

Hi, I'm running FC4 and trying to setup an environment to compile applications that require gcc 3.2.3 I've installed the various compat rpms, which have provided the gcc 3.2.3 compiler. This seems fine, however I am having problems with the standard include paths, as I get this error In file included from /usr/include/c++/4.0.2/string:52, from /data/lhcbSoft/lcg/external/AIDA/3.2.1/share/src/cpp/AIDA/IAnalysisFactory.h:15, from /data/lhcbSoft/lcg/external/AIDA/3.2.1/share/src/cpp/AIDA/AIDA.h:4, from /home/jonesc/cmtuser/GaudiPI/v0r3/dict/AFCreator.h:4: /usr/include/c++/4.0.2/bits/basic_string.h: In member function `bool std::basic_string<_CharT, _Traits, _Alloc>::_M_disjunct(const _CharT*) const ': etc. etc. So, I am actually including header files from /usr/include/c++/4.0.2/, whereas for this application, I need to use /usr/include/c++/3.2.3/. I have a script which I run in a terminal to setup various things for this case. Is there something I can put in this script to alter the default such path to find files from /usr/include/c++/3.2.3 ? cheers Chris

3 4

FC4: Ar problem
by Gilbert Sebenste 03 Dec '05

03 Dec '05

Hello all, When compiling a program, some object modules that are supposed to be added to the program'slibraries aren't getting added. According to a colleague, this seems to be a problem with the 'ar' utility bundled with FC4 that gets tripped on fast machines. A workaround seems to be putting "sleeps" after 'ar' and 'ranlib' invocations. Is there a bugzilla ticket for this? ******************************************************************************* Gilbert Sebenste ******** (My opinions only!) ****** *******************************************************************************

1 0

RE: Giving back for the holidays
by Chris Norman 03 Dec '05

03 Dec '05

I think the guys at Speakup (www.linux-speakup.ort) would be greatful, don't know if they have paypal or anything though. Cheers, -----Original Message----- From: fedora-list-bounces(a)redhat.com [mailto:fedora-list-bounces@redhat.com] On Behalf Of Skunk Worx Sent: Friday, December 02, 2005 5:09 AM To: For users of Fedora Core releases Subject: Giving back for the holidays Over the last two years Fedora has solved some problems for me quickly and easily. Prior to using Fedora, I bought and used a RedHat box set once or twice a year since around RedHat 4. I'm going to donate $250 across one or more open source projects this month. The Kadischi LiveCD and Sourceforge Unichrome projects helped me out a lot, but neither have paypal or donations. Does anyone have opinions on what developers or projects were unusually helpful to Fedora over the last year or two, and have some means of accepting donations? Thanks, John -- fedora-list mailing list fedora-list(a)redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list

2 1

why doesn't apache start?
by Dave Stevens 03 Dec '05

03 Dec '05

This is what I get when trying to start apache. httpd failed. The error was: Starting httpd: Syntax error on line 10 of /etc/httpd/conf.d/perl.conf: Cannot load /etc/httpd/modules/mod_perl.so into server: /etc/httpd/modules/mod_perl.so: undefined symbol: Perl_Ippid_ptr [FAILED] I don't understand this or what to do. dave -- A. "No, I went voluntarily." Q. "Mr. Presley, were you kidnapped by aliens?"

2 1

DNS Problem
by Khalid 03 Dec '05

03 Dec '05

Hi named service is stop automatically giving this error named[1018]: task.c:395: REQUIRE((((task) != ((void *)0)) && (((const isc__magic_t *)(task))->magic == ((('T') << 24 | ('A') << 16 | ('S') << 8 | ('K')))))) failed named[1018]: exiting (due to assertion failure) Thanks Mohd Khalid

2 1

Suppressing Real Player on RPM downloads
by Robert G. (Doc) Savage 02 Dec '05

02 Dec '05

I know I've seen this, but now can't find it on a fresh FC4 install. Where do I tell Real Player to ignore RPM files when I click on them to download? -- Doc Robert G. (Doc) Savage, BSE(EE), CISSP, RHCE | Fairview Heights, IL Fedora Core 4 kernel 2.6.14-1.1644_FC4 on a P-III/M IBM Thinkpad A22p ** Bob Costas for Baseball Commissioner **

7 9

Read/write access on nfs mounted file systems (corrected subject line)
by Dr. Peter Bladon 02 Dec '05

02 Dec '05

Dear colleagues, My apologies for my previous posting where the subject line was meaningless!! I have a problem in that some of the file systems that I wish to mount via nfs will not do so with write permissions set. I have acquired 4 (substantially) identical machines that I intend to use as a cluster (using gigabit ethernet), and also have linked to my other Unix (IRIX) machines (using 100Mbit ethernet). The basic configuration of the new machines is: AMD64 Athlon/1Gbyte RAM/80Gbyte disk/Gigabit ethernet/. Each machine is running Fedora Core 4 obtained from the five x86-64 iso discs. I chose the option to load "everything". The four machines are named Daffodil1, Daffodil2, Daffodil3, Daffodil4 Each of the four discs is partitioned by the loading process thus: /dev/hda1 10 Gbytes / (System) /dev/shm 5 Gbytes (Swapspace) /dev/hda3 65 Gbytes /home (working area) It is my wish to allow each of these four machines to be able to read and write in each other's "working areas". To this end I constructed the (relevant parts of) /etc/fstab thus: . . LABEL=/home1 /home ext3 defaults 1 2 # From the loading process . . Rose:/userdisc12 /userdisc12 nfs rw,bg 0 0 # Mounting an SGI disk . . Daffodil1:/home /daffodil01 nfs rw,bg 0 0 # Added by me Daffodil2:/home /daffodil02 nfs rw,bg 0 0 Daffodil3:/home /daffodil03 nfs rw,bg 0 0 Daffodil4:/home /daffodil04 nfs rw,bg 0 0 . . etc THE PROBLEM Despite requesting that the (remote) mounts are read/write (rw) it has proved impossible for (say) Daffodil1 to be able to write to the file system daffodil02. It is also not possible for daffodil1 to write to daffodil01 even though daffodil01 is effectively an alias for /home on daffodil1. (daffodil1 can write to /home on daffodil1) The machines (daffodil1 --- daffodil4) have no problem in writing on the remote mounted discs on SGI machines (e.g. /userdisc12) What am I missing? Is there some crucial flag that I have not set? Is there a fundamental difference in the way that "mount" works under UNIX and LINUX? Note that machines are named Daffodil1 etc, while file systems are named daffodil01 etc. (extra 0). Any help would be appreciated. Yours truly Peter Bladon

2 1

Re: mounted multiple times
by David G. Miller (aka DaveAtFraud) 02 Dec '05

02 Dec '05

jvian10(a)charter.net wrote: >On Wed, 2005-11-30 at 13:26 -0400, Trevor TeC Christian wrote: > > >>> On Wed, November 30, 2005 12:55 pm, Jack Tanner said: >> >> >>>> > On an FC4 client, I'm mounting an smbfs filesystem over the network. In >>>> > /etc/fstab, I have >>>> > >>>> > //server/share /mount/point smbfs >>>> > credentials=/home/user/.smbcreds,noauto,user 0 0 >>>> > >>>> > The user's .bash_profile contains the actual mount command, which goes off >>>> > without a hitch. Consequently, the user can issue >>>> > >>>> > $ umount /mount/point >>>> > >>>> > when necessary. >>>> > >>>> > It's all well and good, until the user connects to the FC4 client twice, >>>> > and >>>> > runs two xterms which both source .bash_profile. >>>> > >>>> > Now the mount command has been issued twice, so here's what happens: >>>> > >>> >>> >>> >>> Should script such that the share isn't mounted unless it isn't already >>> mounted. >>> >> >> >>>> > $ umount /mount/point >>>> > umount: it seems /mount/point is mounted multiple times >>>> > >>>> > Well, yes, it has been mounted multiple times; but how can the user >>>> > unmount >>>> > it now? >>>> > >>> >>> >>> >>> One should be able to use ps to find the PID of the mounts and kill them >>> directly i believe. >>> >>> ps ax | grep mount >>> >>> then kill -9 the corresponding PID's >>> >> >> > >Mount does not leave an active process running. Thus this would have no >affect. Did you test your idea? > > > >>> Hope this helps. >>> >> >> >>>> > >>>> > -- >>>> > fedora-list mailing list >>>> > fedora-list(a)redhat.com >>>> > To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list >>>> > >>> >>> >>> >>> >>> -- >>> *Trevor "TeC" Christian* >>> Cell #: (767) 225 4472 >>> IM >>> MSN: trevorc98(a)hotmail.com / trevor(a)bouyon.dalive.com >>> YM : trevorc01 >>> AIM: DoubleOTeC >>> GoogleTalk: doubleOTeC(a)gmail.com >>> ------------------------------------------------------------------------ >>> Home Page <http://trevor.bouyon.dalive.com//> >>> Curriculum Vitae <http://bouyon.dalive.com/cv/> >>> Dalive Market <http://www.dalivemarket.com/> >>> >> >> Question, does umount still unmount "one instance" of the mount? This is the behavior under Red Hat Enterprise 3 (2.4 kernel) which also doesn't complain. The application I QA hasn't managed to get confused enough under RHEL4 (we start supporting RHEL4 under our next release) to mount one of our mount points multiple times so I don't have a data point for RHEL4 (2.6 kernel). Under RHEL3, I just keep issuing umounts for the mount point until I get rid of all of them. Would this work for your situation? Any idea when this "feature" got added to mount/umount? Cheers, Dave

3 2

Slow Writing to Lexar Jump Drive
by Charlie McVeigh 02 Dec '05

02 Dec '05

I have a 128 MB USB Lexar Jump Drive. It worked reliably for me for close to a year. About a month ago it became painfully slow when writing large files to it (both Fedora and Windows) - It takes about 24 minutes to write a 4.5 MB file to it. Read speed from the device appears to be as fast as always. I contacted Lexar and they suggested that I reformat the device with FAT32 format. After reformatting to FAT32 read/write speed when using Windows appears to be back to normal. However when automounted on my FC3 desktop the write speed is still 24 minutes to write the same 4.5 MB file. Fast on Windows slow on Fedora - any one have any suggestions? Of course Lexar is no help when it comes to Linux.

9 12

Jump to page: