Re: the role of Custom in XEmacs (setting != saving)

Per Abrahamsen writes: Per> I think I'm getting it now. You want more `setq' variants. Per> 1: The user has explicitly asked for this variable to be changed for the Per> rest of this session, but don't want it to be saved. Per> 2: The user has explicitly set this variable, and also ensured that it Per> will remain set. Per> Customize could certainly use that information. Something like this. The point is that we should also make it clear for the users what it is all about: there are lisp variables, specifiers, whatever, and part of this stuff has the property that the user can modify it and possibly save the modifications for all sessions. Now Custom really belongs to the core of XEmacs. However, from outside, it still looks like an add-on package and people don't realize that the notion of user option is not really related to custom (as a package) but something general in Emacs. The 'custom- prefix everywhere doesn't help them much. Also, there are other things that I was thinking about, which I find feasable: say you have customized a variable, saved it some time ago. Maybe you also have it set from X resources on several machines. Now you realize that your setting is not valid when you're on a tty or anything, and then you decide to actually write it in init.el, inside a `window-system' check, with a nice comment saying why you do this. I think it would be nice to get a warning at the next session that an init.el setting might override a customized value, and possibly that there are conflicts with X resources as well. Per> Any profession that used to have two names now only has one name, because Per> having two names implies that women are not capable of doing the same job Per> as men. Per> Any profession that used to have a single name now has two names, because Per> having one name implied that the profession was reserved for a single Per> gender, thus degrading women. Geeze, we're not the only twisted minds on earth ;-) -- / / _ _ Didier Verna http://www.inf.enst.fr/~verna/ - / / - / / /_/ / E.N.S.T. INF C201.1 mailto:verna＠inf.enst.fr /_/ / /_/ / /__ / 46 rue Barrault Tel. (33) 01 45 81 73 46 75634 Paris cedex 13 Fax. (33) 01 45 81 31 19