Historical bug tracker for the Pacman package manager.
The pacman bug tracker has moved to gitlab:
https://gitlab.archlinux.org/pacman/pacman/-/issues
This tracker remains open for interaction with historical bugs during the transition period. Any new bugs reports will be closed without further action.
The pacman bug tracker has moved to gitlab:
https://gitlab.archlinux.org/pacman/pacman/-/issues
This tracker remains open for interaction with historical bugs during the transition period. Any new bugs reports will be closed without further action.
FS#17689 - libfetch fork inconsistent with freebsd's libfetch
Attached to Project:
Pacman
Opened by Nathan Phillip Brink (ohnobinki) - Sunday, 03 January 2010, 03:54 GMT
Last edited by Xavier (shining) - Tuesday, 19 January 2010, 21:46 GMT
Opened by Nathan Phillip Brink (ohnobinki) - Sunday, 03 January 2010, 03:54 GMT
Last edited by Xavier (shining) - Tuesday, 19 January 2010, 21:46 GMT
|
DetailsI am interested in packaging libfetch for a distro that is not archlinux. In the process of attempting to scout out libfetch, I have encountered two versions so far; the latest (SVN) freebsd version and archlinux's hosted tarballs.
freebsd: http://svn.freebsd.org/viewvc/base/head/lib/libfetch/ (SVN: http://svn.freebsd.org/base/head/lib/libfetch/ ) archlinux: ftp://ftp.archlinux.org/other/libfetch/ diff -ru <freebsd's latest> <archlinux's 2.26 release>: http://ohnopub.net/~ohnobinki/tmp/libfetch.svn.diff These differences in API would be confusing to any developer aiming for portability. Such differences would also suggest that developers include their own tarballs of libfetch in their own source code repositories and set their build scripts to compile and statically link against their own copies of these libs. (it was suggested to me on #archlinux to just ``< Ghost1227> apostrophe: then statically compile the version of your choice into whatever you're building''). ( why encouraging this is bad: https://bugs.gentoo.org/251464 , http://blog.flameeyes.eu/2009/01/02/bundling-libraries-for-despair-and-insecurity ) I think that the best course of action is to establish a relationship with freebsd and recognize their position as ``upstream''. They could possibly be benefited by your expansion of the libfetch API. However, it will be harder for them to deal with the fetchIO typedef you have introduced (if I'm reading the deps correctly. Please recognize that I don't have time to become a history major with a concentration in libfetch development ;-)). |
This task depends upon
http://repos.archlinux.org/wsvn/packages/libfetch/trunk/PKGBUILD
http://cvsweb.netbsd.org/bsdweb.cgi/pkgsrc/net/libfetch/?only_with_tag=MAIN
We use NetBSD code and not FreeBSD; thus if you look how our tarballs are created you will see we aren't inventing our own library. So we are in fact following an upstream library and have in fact reported bugs (and gotten them fixed) there.