On Wed, 2 Sep 2015 07:51:41 +0100 Neil Bothwick <n...@digimed.co.uk> wrote:
> On Tue, 1 Sep 2015 18:06:05 -0700, walt wrote: > > > Calculating dependencies... done! > > [ebuild U ] sys-libs/ncurses-5.9-r5 [5.9-r3] > > [ebuild NS ] sys-libs/ncurses-5.9-r99 [5.9-r3] USE="cxx gpm > > unicode -ada -static-libs -tinfo" ABI_X86="32 (64) (-x32)" > > > > Is this going to install two different (slotted) versions of > > ncurses-5.9 on my stable machine? > > > > If not, then this proposed update must be an experiment or a hack > > or a workaround of some kind. On my *stable* machine! > > It's a workaround, not an experiment as it's been tested for over a > week. It's to get round an issue with subslots and some ebuilds not > handling changes as expected (I won't say incorrectly as there seems > to be some dispute over exactly what is defined in the PMS).AIUI the > second ebuild simply depends on the first, so that all the ebuilds > requiring ncurses work properly. > > See https://bugs.gentoo.org/show_bug.cgi?id=558952 for more > information, probably more than you want to read. Thanks also to wabe and Fernando for your replies. Just for the record I did the update this morning, which completed without errors. qlop shows that both updates completed, but eix shows that I now have only ncurses-5.9-r5 installed. (This is apparently the desired result, but I'm only guessing what the desired result really is.) I think every portage tool should announce very clearly whether a package is slotted/subslotted, and exactly which slot and subslot the package belongs in. The subject of slots is way too confusing to withhold such information. If the devs can't explain slots to their users then they don't understand it themselves. (Hm. That phrase sounds familiar. Where did I get that?)
