https://qa.mandrakesoft.com/show_bug.cgi?id=581
------- Additional Comments From [EMAIL PROTECTED] 2002-11-30 12:35 ------- > Of course, if *you* want to go and track every single package, and be able to > predict which package would break dependencies and which not *in the future*, > you can take over packaging all the RPMS. Practically, there is no way to do > this. Yes, others have made the same sort of silly comment to me via email and they all seem to be based on the same peculiar idea. Its the old "If we can't achieve perfection then we shouldn't do anything at all." argument all over again. Of course I don't believe that it is possible to get all of the dependencies for all of the RPMs correct. And I also realise that Mandrake staff can't spend their time hunting down such dependencies outside of releases. However, isn't the entire point of Cooker that it allows the user community to find and resolve problems so as to save Mandrake the effort? If through the efforts of users SOME of the incorrect dependencies in SOME of the RPMs were fixed then would this be such a bad thing? Its not as if (as one person suggested) Mandrake don't believe in sub-version dependencies. Let's take a couple of examples. $ rpm -q --requires gimp | grep = | grep -v rpm gtk+ >= 1.2.8 glib >= 1.2.8 libgimp1.2 = 1.2.3-17mdk $ rpm -q --requires kdelibs | grep = | grep -v rpm libarts >= 1.0.4 libqt3 >= 3.0.5 libarts >= 1.0.3 Note how in the latter case kdelibs wants libarts to be greater than 1.0.4 AND 1.0.3? I didn't have to hunt for long to find this. Perhaps the Mandrake dependencies are not a shining example of software quality which would be sullied by updating one dependency for Mozilla? If Mandrake made this change to the Mozilla RPM then who would lose? All they have to do is make a one line change to the RPM SPEC file. Its not as if they have to test the change, since it simply makes explicit the dependency which is implicit in the requirement that one not mix cooker and 9.0. Plus, if they decide to release Mozilla 1.2 as an update to 9.0, then one potential bug will already have been fixed. I have to say that the whole thing smacks of the "We have a fix for this bug but we're not going to apply it because it doesn't say in the contract that we have to." attitude that drives me crazy when I deal with software contractors at work. On a related topic how do we know that packages don't have dependencies which they don't actually need? The result of this is that we end up having to install lots of unnecessary software. Example: recently I wanted to install kdegraphics and I ended up having to install heaps of software for digital cameras, scanners and usb ports together with all their associated libraries, despite the fact that I don't have any of that hardware. Now kdegraphics is a complex system and perhaps it is difficult to separate it from these requirements, but I can't help suspecting that there are lots of other small cases where libraries are being installed when they are no longer actually needed. If you think about it for a while, you will see that as long as one adds new dependencies and doesn't remove ones that aren't needed then eventually the distribution will turn into one amorphous blob. BTW, don't forget that I didn't suggest adding a dependency to Mozilla, simply updating one that was already there. ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
