>>>> "Jan" == Jan Vroonhof
<vroonhof(a)math.ethz.ch> writes:
Jan>
Jan> greg(a)alphatech.com (Greg Klanderman) writes:
> However, I'd like it to be the case that you get the error on
entry to
> the package update stuff (pui-list-packages, etc) rather than later
> when you go to update a package.
Jan>
Jan> Then it should not be an error in my case. Often I do just want to
Jan> look at the installed versions and there is no need to go the net for
Jan> that? Note the "List" part of "List & Install"
But if you've customized package-get-remote as you previously stated,
entering pui-list-packages will grab a remote copy of the db with your
last patch.
It just seems the default behavior should favor keeping a less
experienced user from shooting himself in the foot. Usually when you
use pui-list-packages you want to install new packages. You should
get the current list.
Jan> Maybe we should change `pui-toggle-package' to warn in this case.
Jan> And a add package-get-require-base with a "force-current = t" argument
Jan> to it too. Additionally one can add a warning and an "Update" button
Jan> to the top the *Packages* buffer.
seems overly complicated. this stuff is already too hairy. also,
with the code your way, (package-get-require-base t) may succeed but
not actually get a current database.
Jan> Note that the
Jan> current package-get code does correctly first try to fetch the latest
Jan> version if the package-get-remote has since become non-nil.
No, pui-list-packages does (package-get-require-base t) so it thinks
it already has gotten an up to date package index, when in fact it
loaded the one from the etc directory, and has set package-get-was-current
to true.