On Fri, 2008-11-14 at 01:24 -0500, Seth Vidal wrote: > > On Thu, 13 Nov 2008, Bill Nottingham wrote: > Look, for the desktop in particular NM makes a lot of sense, I am not > arguing otherwise. Well, I would not key NM to separating desktops vs. servers. IMO, NM is addressing "machines w/ dynamic network connections", but it's ineffective/unnecessary overhead on machines with permanent/static connections, independently if these are servers or desktops (e.g. classic workstation desktop machine pools). Closely connected is another aspect, where I feel NM (and other tools in Fedora) has conceptional weaknesses: It doesn't take into account "user roles" and "machine roles", but only considers "personal machines", i.e. the classic "personal laptop". In practice, this clashes with concepts of central vs. decentral administration and with user roles. > For the server it is a solution looking for a problem. Agreed. > The reaction you're seeing is people who don't care about the desktop > trying to figure out why desktop and/or developer oriented features are > causing them to have to change their server deployment/config habits. Not necessarily. IMO, if NM was as easy to use as some people try to advertise it, people weren't complaining. Fact is: People are complaining for years. To me, the reasons to complain about NM are quite simple: I have too often been confronted with situations, where NM didn't not work as advertised. Worse, due to the fact "NM is wanting to be clever" and it's total lack of documentation, it often left me clueless about the causes. Ralf -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
