On Wed, Dec 17, 2014 at 02:23:13PM -0500, Greg Troxel wrote:
> > Why don't we mark it BROKEN now (for this coming branch), delete krb4
> > whenever, and remove it after the next branch if nobody fixes it? I
> > don't expect anyone will but you never know.
>
> Right now, with krb4, I think it works. So it's really
> marking-broken/deleting when krb4 is removed.
Well, saying it's broken because it depends on krb4 seems perfectly
reasonable :-)
But, since the goal is to summon anyone who's actually using it,
marking it broken early seems like a better approach. Of course,
updating it seems better still.
> It seems we are leaning to removal rather than marking BROKEN for things
> that aren't super likely to get fixed, because the history is there
> anyway and in this case the new upstream version of zephyr has changed a
> fair bit anyway.
Once it's removed, it gets forgotten, though. So I guess in general it
depends on the extent to which we care whether it never comes back.
And I suppose in this case while zephyr used to be fairly important,
it probably isn't any more. Does it still get used to any extent at
MIT?