Re: [PROPOSAL]drop LSB.usersgroups/commands/passwd/T.passwd 2
hi John
(My thoughts were based on some parts of POSIX where they
omitted the group password feature intentionally) I agree that the
key here is that any change be decided by the specification
owners and not the test folks, who need to be careful
to follow the specification.
There needs to be a process for defect reporting and resolution;
as you say this feature may be necessary for some, is it in
scope for defect reports to remove such a feature?
I agree also that the spec owners ought to consider an obsolescence path for
deprecating features. Some specifications I have worked on have
marked features as "To be withdrawn" for one version of the
specification prior to dropping a feature.
regards
Andrew
On Jul 30, 10:38am in "Re: [PROPOSAL]drop L", John H Terpstra wrote:
> Andrew,
>
> I know of sites that still use the group password feature. Even if we
> offer a waiver here and we assent to the demise of this "feature", for a
> very small minority this is still a needed behaviour. As such I recommend
> that we retain this. We should however flag all features that are
> earmarked for obsolesnce. A suitable notation should be made in the spec.
>
> - John T.
>
Reply to: