On Sat, Sep 10, 2011 at 10:38:55PM +0200, Matthias Andree wrote: > Am 10.09.2011 18:17, schrieb Chad Perrin: > > On Sat, Sep 10, 2011 at 12:09:16PM +0200, Matthias Andree wrote: > >> > >> On the other hand, you're pointing out a problem of dead ports in the > >> first place: if the API of (usually library) port Y changes, and port X > >> is unmaintained, that's typically a situation where port X needs to be > >> deprecated and removed (and also will no longer build and/or work). > > > > I want to understand all the reasoning behind this stuff. Please explain > > the reason that library Y changing means that dependent port X should be > > deprecated and removed, regardless of whether it no longer builds and/or > > works. Note that I'm working on the assumption that your assertion it > > should be deprecated and removed does not rely on it no longer building > > and/or working because of the way you mentioned no longering building > > and/or working as a parenthetical addendum rather than a condition of > > deprecation and removal. > > I suppose you missed the meaning of "if the API of port Y changes". > API = application programming interface. This implies that either the > application no longer builds, or it is known that it would behave > inappropriately with the new library (because semantics changed).
That would have been an unwarranted assumption. If the API changes, it might mean a couple of changes were made -- and they do not affect the parts of the API that port X uses. I chose to take what you said at face value, rather than make a bunch of assumptions about it. Thanks for clearing up your intended meaning, though. -- Chad Perrin [ original content licensed OWL: http://owl.apotheon.org ]
pgp66Owl5ETND.pgp
Description: PGP signature
