On 5/23/08, Ciaran McCreesh <[EMAIL PROTECTED]> wrote: > > It is great that package manager forces a specific solution. > > But unfortunately there are some different views and I don't > > understand why you force a specific solution (network implementation). > > > We don't. Do whatever you want to do. No-one's going to stop you, even > when what you do is really really bad.
How do you not stopping me if you don't provide the option? Strange... I may miss something. > Honestly, if we're going to solve the "lots of systems" problem at some > point, we're going to do a hell of a better job of it than "have a > package manager on every system and install from binaries on every > system". But it's not been a priority for anyone. Which "hell of a better job" solution have you thought about? I am truly interested. > If you seriously think that anything Gentoo provides is a "good > solution" for large scale deployment, you need to work with real good > solutions until you realise that doing what you're describing is a > really bad way of managing anything. Did not say anything... I am here it suggest something... But at least Gentoo initiators provided some tools, they are not perfect, but at least they provide a solution until better can be implemented. > > > Uh, the correct thing to verify is the process as a whole. > > > > The right answer can be given even if the logic of ones though is > > wrong. If we go into details, there are white box checks and black box > > checks, you are forcing developers to use black box only, while the > > white box tests are the ones that finds most of bugs. > > > Uh... I'm not forcing anything. You can do both kinds of testing quite > easily using Paludis. How? Can you please tell me how you do white test without using "ebuild" command, and without modify the ebuild file before each test (so there will be no leftovers)? BTW: How do you change USE flag on command line: USE="X" paludis --install --preserve-world package Does not set use X for this command. This important if you want to test all USE combinations without modifying persistant configuration. so you can always return to known baseline after tests. > > Again, you force your QA view on users, although it may be incorrect, > > without providing the ability of users to use the process the choose. > > > I also don't provide a "remove glibc from / without asking anyone" > option... This is for the package to decide, not for the package management. > > Had you said that you don't provide phases as it complex for you to do > > this, I would have understood, but if you do this just to force your > > (wrongly) view of QA management, I think this is incorrect. > > > It's not complex at all... It's just that no-one's done it because it > isn't an essential or especially useful feature and won't become one no > matter how many times you say it is. OK got it... Although I don't believe I am the first one who raise this quetion. If I understand correctly, even if I write patch for it you won't merge it as you think this feature should not be available... > > > Why would it make migration any easier? You might as well install > > > dselect for all the difference it would make... > > > > Package management is the major components, as you can see from our > > discussion. > > > Uh... So? What's Paludis on Gentoo got to do with Paludis on Exherbo? As you can see, none of the feature we discuss are distribution specific. On both systems you won't be able to do white test easily, on both you won't be able to distribute packages from compiler to stations, I guess the cross compile stuff will be solved somehow not as part of the package manager (crossdev script like). Thank you, Alon. _______________________________________________ paludis-user mailing list [email protected] http://lists.pioto.org/mailman/listinfo/paludis-user
