Proliferation of small packages

[William Reade]

On Thu, May 30, 2013 at 9:59 AM, Tim Penhey <tim.penhey at canonical.com>wrote:

> My agreement was around the meta comment of considering existing
> packages rather than creating a new one for new work by default.


I'm happy to follow consensus here; my only concern is that everything in a
package is (in c++ terms) a friend of everything else. In the absence of
more sophisticated access control, this makes it hard to decompose packages
once they've grown beyond a certain point. So I'm inclined to call
implementor's choice, in general, but ask that you all bear this tendency
in mind as you go.

Cheers
William
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ubuntu.com/archives/juju-dev/attachments/20130530/f82d9c73/attachment.html>