Martijn Faassen wrote:
Note that Malthe (cc-ed too) is interested in:
"Architecting and implementing a Zope component that can extend a
running application in a way that plays nicely with both the component
architecture and setuptools. The extensions would typically be pulled
in over the network as eggs."
which is a project he proposed as a possible summer of code project. I
see a lot of possible connections between the project you mention and
this project. I think much of this project could be architected as a
generic Python library and/or Zope 3 component (that offers some
abstractions to describe, and possibly also run, a buildout), and then
you could do the UI in Grok or whatever.
When I first read about the buildout builder I had associations to the
plone.devel debate on genesis, a code-generation tool for Plone (not
sure about the scope of it).
If people are supposed to read the output of this tool, then I don't see
the point of using a tool to generate it. Assuming the output is
actually readable, why was it then necessary to generate it?
It can be a daunting task to put together a buildout from scratch that
builds, say, a full Plone site with extension points like the canonical
Plone buildout configuration does. But why would you?
I'd like to question the objective of the buildout builder project.
I remember reading something like "I'd like to cache my site using
varnish" and the builder could then make that happen. But you could also
just google "buildout varnish" and get:
and off you are.
Zope3-users mailing list