On Oct 3, 2007, at 12:44 AM, Peter Kriens wrote:
The reason why we are considering allowing multiple hosts to use the
same fragments is because we run in version problems. Allowing only
one host to use a fragment, means that you have to disallow some hosts
to use a fragment or connect them to an older version. I tried to
describe this and it was terribly ugly and sounded very arbitrarily.
These seem to be descriptions of the "how" and not the "why". Are
you at liberty to provide the use cases that are driving this effort?
I don't quite understand your phrase "connect them to an older
version". Can you provide more detail?
Regards,
Alan
We discarded the multiplicity because we could not figure out a number
of details in time for R4 and felt it was easier to later relax the
spec than to narrow it one day. Believe me, we spent an insane amount
of time on these issues because we all want to keep the spec clean and
small ...
Kind regards,
Peter Kriens
AC> My humble thoughts:
AC> Like any good work of art it is hard to know when enough is enough
AC> and it's time to leave the piece as it is. I am worried that what
AC> was once a clean and elegant spec will become muddied to support
AC> features that, imo, tools should have taken the burden; take
required
AC> bundles for example.
AC> I'm not aware of the compelling use cases that call for an
extension
AC> to the number of hosts for a fragment but can think of one, again
AC> imo, not so compelling one. This being the patching of a set of
AC> bundles with one fragment. I'm happy to go into detail as to
why I
AC> think poorly of this use case if it is the impetus for adding this
AC> extension and if this is the appropriate forum.
_______________________________________________
OSGi Developer Mail List
[email protected]
http://www2.osgi.org/mailman/listinfo/osgi-dev
