On Wed, 5 Sep 2007 11:35:34 -0300 "Nelson A. de Oliveira" <naoliv@gmail.com> wrote: > Unfortunately it's too much subjective to define what is a break. A > user may think that a missing help file is a break, while another one > just don't care about this missing file. A user who deletes /usr/share/doc and then complains about a missing file in a package that informs the user and continues to operate normally isn't being entirely logical! I don't think a break - in the context of this part of Policy - is that hard to define. If the package has a command line option or GUI menu command that would normally activate/load/display/otherwise process a file in /usr/share/doc for whatever reason, then if the file is missing, that package must explain to the user why the command cannot be performed and then carry on entirely as before. Otherwise it is an RC bug. A corrupt file in /usr/share/doc is entirely different - a crash resulting from a corrupt file is a normal bug (or maybe important if data is lost) but it cannot be RC without other more dire consequences (like breaking other packages in the process). -- Neil Williams ============= http://www.data-freedom.org/ http://www.nosoftwarepatents.com/ http://www.linux.codehelp.co.uk/
Attachment:
pgp_xxeZzqTBm.pgp
Description: PGP signature