Re: Emacs 22.0.91 working, lazy-lock documentation inconsistency?

[Juanma Barranquero]

On 12/3/06, Stefan Monnier <address@hidden> wrote:

> Then why bother distributing it with Emacs at all?

Because it is much easier to enable a package you have already in the
Emacs distribution, than going around shopping for an obsolete package
whose "right" (i.e., newest) version can be hard to determine (witness
the repeated troubles trying to find what was the right gnuclient
release to use).

> The idea of the `obsolete' directory (the way I understood it) was to have
> an equivalent of the `make-obsolete' function: the package is marked as
> obsolete and we encourage users not to use it any more, but it is still 100%
> backward compatible.

Obsolete *autoloaded* packages do not generate any warning on using
them, so there's less incentive to switch (for people who's already
using them; new users of course will see that the package is in
obsolete). I agree with Richard: the error message is a good reminder
to switch to better alternatives.

Put another way: if they're always autoloaded, and someday we want to
remove them from the distribution, it will be an abrupt change, from
"working as usual" to "non-present". Not autoloading them makes this
gradual: at least the user will know at some point that he is using a
package considered obsolete.

                   /L/e/k/t/u