On Sun, 27 Feb 2011 15:13:23 +0100
Bernd Steinhauser <[email protected]> wrote:
> > No, that's not backwards compatible, and it doesn't fit in nicely
> > with repositories whose masks don't support tokens or untokened
> > masks in repositories which do support tokens.
>
> Isn't that just a configuration issue (we're talking about
> package_unmask.conf, right?)? IMO, Paludis should moan if there is
> such a mask and tell the user about that.

It's a pretty big deal. It's completely wrong to assume that every mask
will be tokened straight away, or that every mask will eventually be
tokened. Tokens are an optional extra, not critical functionality.

> > Whatever replaces PLATFORMS will be based upon whitelisting, not
> > blacklisting, so it'll be a different mechanism anyway...
>
> Doesn't that require both whitelisting and blacklisting anyway (at
> least from a user's point of view)?

For the user, yes. For the repository, hopefully not.

> BTW, are packages grouped in the usual way?
> (i.e. (cat/foo1 cat/foo2) [[ ... ]])

That's not the usual way. [[ ]] applied to a ( ) block does *not*
usually apply to every child. But yes, there seems to be a need to do
something like:

(
    cat/first[>=2.0_beta]
    cat/second[>=2.0_beta]
) [[
    comment = [ mask cat testing packages ]
    token = [ testing ]
]]

It's important to remember, though, that [[ ]] applied to a ( ) applies
to the ( ), not to the things inside it.

-- 
Ciaran McCreesh

Attachment: signature.asc
Description: PGP signature

_______________________________________________
Exherbo-dev mailing list
[email protected]
http://lists.exherbo.org/mailman/listinfo/exherbo-dev

Reply via email to