On Tue, May 08, 2001 at 02:21:17PM +0200, Lenz Grimmer wrote:
2) The RPM update for apache conflicts with the update for mod_ssl; the update for apache also conflicts with various Perl apache modules. Easily gotten around with --nodeps, but annoying nonetheless.
Hmm, does it conflict, or are dependencies missing? These are two different things.
Sorry, yes, the dependencies are wrong.
4) The DM option in /etc/sendmail.cf is set to YAST_ASK, which is really annoying, because YaST never asked. So when sendmail didn't work I had to go track this down.
I assume, SuSEconfig is using a default value in this case, but I have no real clue here.
Clearly it is. One of my pet peeves is the use of default values that will have to be changed 100% of the time, with no automatic notification or warning of the values set. In the case of sendmail.cf, you might as well leave the DM option blank. It often isn't needed at all.
6) If you want ypbind, and try to configure it using YaST1, it won't let you specify that you want to find an NIS server by broadcast instead of by IP. This has been a problem since at least SuSE 6.4.
Our YP maintainer (and original author of NIS for Linux) is against doing this - I assume he has a valid reason for it.
Unfortunately, I don't make the decisions in my company's IT department, so this just means that I (and others like me) can't reliably use YP.
8) Right after my installation, I downloaded and installed the updates that SuSE had posted on the website. I later discovered the YaST2 online update. It's a great tool (and free, unlike RH up2date!), but it doesn't recognize when I've manually updated a package. So I had to let the program re-download and re-install all the patches that I'd already installed. Waste of bandwidth, really.
Well, there is not much we can do here :)
Why not? The online update tool should be checking for new package version > old package version, not >=. It doesn't seem like it'd be that hard to do.
9) When I configured my static IP (no special routing or anything) I found that the default route wasn't ever added. So I had to go into the "expert" configuration and add it myself.
How are we supposed to find your default route, if you do not provide it? Black magic? Or are you rather saying, the user interface is not optimal here?
I'm saying the UI isn't optimal. If you don't want to guess a user's gateway based on IP and netmask, then put a box for "Gateway:" on the main screen. Don't make users have to click on "Expert" just because they have a static IP and want to get to the Internet.
So those are my nitpicks. I also want to express my happiness that the init scripts are now in /etc, and that the runlevels now match Redhat's.
No, they now match the current LSB specification draft, which happens to be similar to Red Hat and other distributions. :)
Point taken. :) (I didn't *want* to mention RH, but didn't know what else to say...) -tara