Re: Bug#466487: libauthen-smb-perl: should this package be orphaned or removed?
Ugh...
Xavier Oswald dijo [Thu, Feb 21, 2008 at 02:34:59PM +0100]:
> > I don't have a compelling reason to jump and adopt this package;
> > however, I know many modules in the Authen::* namespace were recently
> > grabbed by Xavier Oswald. Xavier, or anybody in the group: Are you
> > interested? :)
>
> I just had a look and it's not that easy.
> It seems that libapache2-authenntlm-perl includes all smb files of
> libauthen-sbm-perl. That's why #432809!
>
> libapache2-authenntlm-perl should depends on libauthen-smb-perl.
> What do you think ?
>
> I don't know why libapache2-authenntlm-perl includes smb files :/
There is a reason - From libapache2-authenntlm-perl-0.02/smb/README:
This is a slightly modified version of Authen::Smb for use with
Apache::AuthenNTLM
Authen::Smb allows you to authenticate users against an NT server. See
Smb.pm POD documentation for details.
And, no, differences are not _that_ trivial (see the attached file for
a simple "diff -ru libauthen-smb-perl-0.91/ libapache2-authenntlm-perl-0.02/smb/"
- "only" 524 lines!).
It seems to me (I might be wrong) that Gerald Richter got tired of not
getting feedback from the Authen::Smb and decided to hijack
it... inside this package :-/ Even the version number has been
bumped (0.91 -> 0.96). At a quick glance, the additions look like
perfectly legitimate bug fixes and enhancements for upstream, and in
any case, any NTLM-specific content AFAICT is in smbval/valid.c,
with a nice comment preceding it:
/* the following addtions are taken from mod_ntlm, see http://sourceforge.net/projects/modntlm/ */
So... With the situation as it stands:
- Tests should be performed. AFAICT, libapache2-authenntlm-perl could
provide: and conflict: libauthen-smb-perl.
- libauthen-smb-perl does not really hold any value over
libapache2-authenntlm-perl, as it is just an older version. In any
case, newer Authen::Smb should be decoupled from the (upstream)
Apache2::AuthenNTLM package.
Should we mark this as upstream and forward the hot potato to the CPAN
RT? :-/ It does not make me too happy anyway...
Greetings,
--
Gunnar Wolf - gwolf@gwolf.org - (+52-55)5623-0154 / 1451-2244
PGP key 1024D/8BB527AF 2001-10-23
Fingerprint: 0C79 D2D1 2C4E 9CE4 5973 F800 D80E F35A 8BB5 27AF
Reply to: