Re: Don't bail out on :package-version in `defcustom'
"Stephen J. Turnbull" <[email protected]> Sun, 28 Sep 2014 02:09:54 +0900
| Newsgroups | gmane.emacs.xemacs.patches |
|---|---|
| Message-ID | <[email protected]> |
Michael Sperber writes: > > "Stephen J. Turnbull" <[email protected]> writes: > > > Michael Sperber writes: > > > > > +:package-version > > > + VALUE should be a list with the form (PACKAGE . VERSION) > > > + specifying that the variable was first introduced, or its > > > + default value was changed, in PACKAGE version VERSION. This > > > + keyword takes priority over :version. The PACKAGE and VERSION > > > + must appear in the alist `customize-package-emacs-version-alist'. > > > > Will an error be signaled if not? Where? > > No. I hate Elisp (almost as bad as X) for being so shoddy about validating arguments.... > > This really should document the format of VERSION. > > Yes, I was wondering about that, too, but the format seems to be > unspecified in GNU Emacs, as well. Should we pick one? No, I mean whether it's a cons or a string. In the part somebody cut custom-add-package-version takes a VERSION argument. However, I suspect the VERSION argument is not just a version string, but rather a (PACKAGE . VERSION) cons as documented above for the :package-version key. I think that picking a format for either PACKAGE or VERSION is not going to work. Just let them be strings and we'll deal with inconsistencies in `customize-package-emacs-version-alist' if and when it ever bugs somebody.