Re: [gentoo-dev] [RFC] ACCEPT_RESTRICT for questionable values of RESTRICT

[Kent Fredric]

On 1/11/07, Chris Gianelloni <wolf31o2@g.o> wrote:
> getting quite hostile.  The only thing I can possibly gather from this
> is you're intentionally being fucking dense, so it's not worth my time.
> How is it that you can ignore half an email and only respond to
> something out of context and then still fuck *that* up?

Chill fellas.

Imo, FEATURES = things we say 'portage, mutate everything like so' ,
and "RESTRICT" to my way of thinking is per-package dependant, more
like USE flags, except more general and apply to all packages.

I think one of the arguments is that it provides a level of
communication between the package and portage/user as to what types of
things a package is permitted to do.

Say for example, we have a package called "child" ( forgive me If I've
also missunderstood the point of this feature ) . Now by default, say
all packages are not allowed "to go outside", but package "child" has
a unique situation where it needs to perform "go outside" in order to
merge. The "child" package of course is naïeve and knows nothing about
the outside environment that its trying to install into. So the child
reports a "RESTRICT='go outside' ( if i understand correctly ) , and
it can only "go outside" by doing this 'RESTRICT" request. By default,
the environment has ACCEPT_RESTRICT="go outside" in it. Now the
environment owner may be running a situation where they dont want apps
'going outside' and potentially trashing thier manicured lawn so they
say ACCEPT_RESTRICT="-go outside" , and after that instruction, all
ebuilds requesting to go outside will be bluntly denied.

Is that analogy of any sence to anybody?, or Have I completely missed
the plot too :S

-
Kent

-- 
gentoo-dev@g.o mailing list