On Mon, 2005-04-04 at 05:00 +0200, Axel Thimm wrote: > > Uh, I wasn't making a comment about you, > > Commenting on my work is commenting on me. Great. Then all we do all day long on this list is comment on Seth? I'm sure he appreciates knowing that. I don't know about the rest of people here, but yum's achievements or shortcomings have absolutely no influence on my opinion of its developers. You, on the other hand, just called me "stupid." Therefore, if I come off as a little snippy, there is a reason for that. > > I was saying that when it's impossible for yum to figure out > > dependency issues (a situation commonly arising from broken > > repositories, of which there are plenty, especially when mixed and > > matched), trying to program in some fuzzy logic to acommodate for > > configuration and/or repository error will result in fuzzy systems. > > Yes, nicely worded FUD and still completely unrelated to the issue, FUD stands for "fear, uncertainty, and doubt," and I don't see how my little clarification above is any of these three words. > still implying that there is a "configuration and/or repository > error". There is/was no broken repo/configuration involved in > this. Period. Fine, take a cookie. If you have found a bug in yum/rpm/programmer's dna/whatever, you by all means deserve recognition. When I find bugs in software (such as, say, Python), I just report the problem upstream and then work around the issue in the meantime, but that's probably just me. Now, before Seth drives over here and kicks me repeatedly in the kidneys for the silliness that is this thread, allow me to reiterate: 1. I, Konstantin Ryabitsev, have no opinion on ATrpms, nor you, Axel Thimm, personally. In fact, this message was sent using madwifi drivers from the project, for which I shall add a "powered by ATRpms" signature at the bottom, since you seem to take great personal pride in your work. 2. It is, nonetheless, my opinion that when yum fails to resolve a dependency, it should abort and not try to make important decisions on the part of the admin, thus probably causing more problems in the long- run. 3. Since such dependency issues usually are the fault of packaging (of course, excluding rare bugs allegedly encountered by ATRpms and you, Axel Thimm, personally), the general rule applied by yum developers is that they will not devote their time to trying to work out some heuristic mechanism to second-guess a solution. Henceforth my response to the original message. There you have it. So, now, can we get along with our lives? (I will only further respond to issues dealing with the question whether it's appropriate for yum to fall back to an older version of a package should installation of the latest one available be impossible.) --icon -- Konstantin Ryabitsev <icon@xxxxxxxxxxxxxx> Duke University Physics This message was powered by ATRpms and by Axel Thimm, personally.
