On Mon, Dec 06, 2004 at 05:27:03PM -0500, A1tmblwd@xxxxxxxxxxxx wrote: > What has your experience been using the ATRPMS, http://atrpms.net , repository? Excellent ;) > Mine has not been very favorable. I regret creating atrpms.repo and adding it > to the /etc/yum.repo.d/ directory. > > BEGIN VENT > ========== > > Here are the problems that plague my FC3 system after I used "yum update": > > 1. lm_sensors-2.8.7-2 was replaced with a buggy lm_sensors.i386 2.8.8-0_35.rhfc3.at. No, as explained in the bug report your config is buggy, when you ask for lm_sensors to start w/o configuring it. lm_sensors did not self-activate. > I discovered that when I rebooted my system and encountered a "cannot access > procfs/sysfs" error message. I restored the older lm_sensors rpm file and posted > a bug report #329. As of this posting rhgb still has does not work but I have not > implemented the steps to fix problem 3 and replace/repair upgraded packages. Your rhgb issues will most certainly be unrelated to both lm_sensors and ATrpms. > 2. yum-2.1.11-3 was replaced by yum.noarch-2.1.12-43.rhfc3.at. The install overwrote > the existing yum.conf file. Not very nice. Should not happen. Another bug > report #330. Axel Thimms comments and says that he finds no fault with his script. Not "his script", but the tagging of configuration files. If something is tagged %config(noreplace) then it passes the information to rpm and rpm does what it's got to do. If you had manual changes and rpm blasted your config files, then that would have been a severe regression of rpm proper, but I don't believe that this is the case. You should still report this to the rpm developers, if you think the bug is there. > 3. "0check4updates" was inserted into /etc/cron.d/cron.daily without my > permission or awareness. It runs a script to update APT and YUM. Bug report #331. > > "locate check4updates" yielded three hits. No hits using "rpm -q check4updates". > A little googling for "0check4updates" pointed me to > http://download.atrpms.net/production/sources/common/atrpms/atrpms.spec. Leading > me to try: > > # rpm -q atrpms > atrpms-61-1.at > > # rpm -e atrpms > error: Failed dependencies: > atrpms-perl-module-helper is needed by (installed) perl-XML-Parser-2.34-5_2.rhfc3.at.i386 > atrpms-perl-module-helper is needed by (installed) perl-HTML-Parser-3.36-0_2.rhfc3.at.i386 > atrpms-perl-module-helper is needed by (installed) perl-XML-Twig-3.15-0_8.rhfc3.at.noarch > atrpms-perl-module-helper is needed by (installed) perl-XML-LibXML-1.58-2.rhfc3.at.i386 > atrpms-perl-module-helper is needed by (installed) razor-agents-2.61-7.rhfc3.at.i386 > atrpms-perl-module-helper is needed by (installed) perl-IO-Socket-SSL-0.95-1.rhfc3.at.noarch > atrpms-perl-module-helper is needed by (installed) spamassassin-3.0.1-0_19.rhfc3.at.i386 > > Man, this guy really wants to keep his code installed. The > atrpms-perl-module-helper is dug in so deep yum lists 56 other > packages must be removed to clear the dependencies. Please, not that FUD again. Since you like googling, check the discussion on site vs vendor vs perl directories. You'll find lots of Debian related parts, which equally apply to Red Hat/Fedora. Bottom line is, you need the versioned site dir at the top of the search path, otherwise perl modules under the site dir may not be seeing the proper modules. > Why did he sneak check4updates into the atrpms-61-1.at package? What makes you think check4updates wasn't in the atrpms package from the beginning more than two years ago? What's so conspirative of a method deployed also under up2date and yum under different names? check4updates is just a notifier for apt/yum on new packages. And it is tunable by simple settings under /etc/sysconfig/check4updates. </consipracy> > It is certainly not required for the proper operation of yum or > apt-get. > > I believe that I have just encountered my first Linux trojan! > > I am definitely not happy about the turn of events. Axel has > commented on bugs 330 and 331. He has not commented on 329. In each > of his comments he finds no fault with his code. Sigh. Programmer's > denial? > > My advice stay well away if you know what is good for your system. It is sad to see someone quite clueless (which is not something to be ashamed of, everyone is doing his first steps once) to jump to conclusions and shout out wrong advices to others. Sorry to say, but your post doesn't really contribute in any constructive way. You are frustrated because your system hangs on rhgb and you try to find a scapegoat. Nevertheless the bug reports (even though these were invalid) are welcome. Better an invalid report too much than missing important ones. > END VENT > ======== > > Peace and joy! > > Kam Leo -- Axel.Thimm at ATrpms.net
Attachment:
pgprE3HCkFHx4.pgp
Description: PGP signature
