[Framework-Team] Re: ploneenv - Or how using workingenv for a common Zope2 project might look like ;-)

whit d.w.morriss at gmail.com
Wed Feb 14 14:42:56 UTC 2007


> Well, yes, but if one developer does edit buildout.cfg (we added a new 
> requirement or a new step in the build process), everyone else svn up's 
> that file and runs bin/buildout and their environments are the same 
> again. The --requirements file is, as far as I understand, a one-off 
> bootstrap.

the only difference between buildout.cfg and a -r file that I see is you 
can specify builds of non-python based products in a buildout.cfg. 
Either can live in an svn repo and be updated.

-r files can be combined with links pages and used as a sort of external 
way to manage distribution; requirements file let you specify what to 
install as a development egg(similar to buildout), why changing the 
links page will actually change what code is checked out as a 
development egg.

   In the case of building zope and plone, I don't think there is much 
advantage one way or another; most of the dependencies could be or 
should be eggs(and I think this should be the emphasis rather than 
forcing workingenv or buildout down anybodies throat).

If eggs work, you can use what you like; making a zc.buildout recipe to 
read requirements file would be pretty easy(see this script: 
https://svn.openplans.org/svn/PoachEggs/trunk/).

The situation changes case when looking at things like building a real 
Plone deployment setup: squid, pound, nginx, postfix for listen, etc. 
The little bit of work I've down with trying to split deployment and dev 
builds is that having the option to run just some of the build is really 
nice(either through options on a script or multiple scripts).  It also 
eliminates a bit of the fear factor that comes from installing a monolith.



-w


-- 

------ d. whit morriss ------
- senior engineer, opencore -
- http://www.openplans.org  -
- m: 415-710-8975           -

"If you don't know where you are,
  you don't know anything at all" 


Dr. Edgar Spencer, Ph.D., 1995





More information about the Framework-Team mailing list