On Jun 3, 2013, at 2:36 PM, Greg Stein <[email protected]> wrote: > On Mon, Jun 3, 2013 at 4:23 PM, Alan Cabrera <[email protected]> wrote: >> Here's my current thinking. We would have two separate projects, one for a >> base library and command line tools and another for a gunicorn based web >> site. >> >> Thoughts? > > I assume you mean something like: > > repos/asf/steve/ > trunk/ > tags/ > branches/ > site/ > > Where the Steve code goes into trunk/tags/branches as usual, and the > site/ directory contains the web site tooling/content. > > The only question here, I believe, is whether we want to ship > documentation, or just refer people to our website. If we want to ship > docco, then it would go into trunk/. (and possibly incorporated > somehow into site/ by the tooling)
I think that the Sphinx-licious docs should be on our website, read-the-docs, and delivered bundled in with the code so that consumers can generate their own Sphinx-licious docs. > Note that steve.apache.org must be static (all sites at the ASF must > be static). We won't be allowed to use gunicorn, or similar. Sorry, I should have been more specific. By website I meant the voting website that's in this project. I thought that the release cadences of both may be different and the way I'm used to delivering Python code is that command line and library code gets shipped together, like boto, and gunicorn websites get their own project. If there's a clean way to ship both together then I'm eager to see how. The project website, I agree, should be somewhat static though I imagine it will include Sphinx-licious docs. Regards, Alan
