Fernando Cassia wrote: > > It is just RIDICULOUS not to have an automated way to create RPMs. If > I can download some source file, and then ./configure make and make > install, this automated tool is perfectly capable of finding the base > system, the files installed during the process, and the requirements. > > I just don't understand this *nix attitude that "complex is good". End > users might want to create their own RPMs, too. > There is not a *nix attitude that "complex is good". Until the event of fancy GUIs, the trend was for simple packages that did one thing well, and could be linked to other programs to do complex tasks. (There are a couple of exceptions to this.) A lot of GUIs are front ends for one or more cli programs, so they are also following the building block tradition. Now, creating an RPM may be simple, or it may be complex. A lot depends on the package you are working with. One thing that does add a complication is that you are NOT compiling the package on the target system. Another is that you are providing a way to easily update, or remove the files. Add to that the fact that you have to keep tract of the package requirements, as well as the fact that the same source RPM may generate more then one binary RPM, and things start to get complicated. This is almost unavoidable is you want good a package management system. Another thing to keep in mind is that the same source RPM is often used on more then one architecture. When I was still running my Sparc Station, I used to build binary RPMs from source RPMs that were written for i386 system. For a well written RPM, all I had to do was run equivalent of "rpmbuild --rebuild <source RPM>". Mikkel -- Do not meddle in the affairs of dragons, for thou art crunchy and taste good with Ketchup!
Attachment:
signature.asc
Description: OpenPGP digital signature
-- fedora-list mailing list fedora-list@xxxxxxxxxx To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
