Re: New version of gep-2.html in cvs

[Michael Meeks <michael ximian com>]

Hi Bill,

On Fri, 2002-08-30 at 17:35, Bill Haneman wrote:
> It's still more-or-less an action gep but I started a Requirements
> bullet list.

	To quote from the (new) README that I added; and to expand my comments
on what an Action / Requirements GEP is here is the section:

** Action or Requirements **

        + The answer to this is pretty much always requirements,
          unless you have some really, really simple action to
          propose [ eg. removing a certain module ].


	Your changes are a great improvement towards codifying a set of
requirements; however - this is clearly not an 'action' gep - so it
would be most expiditious to change it to a 'requirements' GEP - so that
it's consistant with the framework, and good practice.

	There are two '2.2' sections somehow :-)  the latter is also not a
requirement. A requirement would be more solution neutral something
like:

2.3 Theme configuration / security

	While it is vital for themes to be able to have their own configuration
options, there are namespacing and security issues entailed with this -
these are XYZ.

	NB. I don't believe this is a very good requirement personally -
installing un-trusted themes will always subvert your computer - since
they may contain code. But I'd be _really_ interested in cutting through
the fug of proposed implementation complexity to the basic requirement
you are trying to drive at. Can you re-phrase that ?

	IMHO it would be good to hoist some of the list of things to theme
section 3.1 into the 'mininum' specified by the requirements; clearly
it's no good having a metatheme thing that doesn't theme anything ;-)
It's great to be considering the Java LAF etc. as well BTW.

	Lastly - I believe that in order to ensure that no 'astro-turfing' goes
on - it would be nice to dramaticaly pith the 'Issues Raised during
Discussion' - and attribute themwhere possible to individuals, in
'comment' blocks, - cf. gep-1.html, although clearly the additional
structure you present is most helpful.

	So - in summary; great work, with a bit more polishing, this can truly
be a model GEP to point people at in future.

	Regards,

		Michael.

-- 
 mmeeks gnu org  <><, Pseudo Engineer, itinerant idiot