Forum OpenACS Q&A: Re: Instant gratification OpenACS installation documentation needed

Yesterday I installed OpenACS into the 'template1' database. I was hoping to use this for development, and create a new database just by using createdb as the user who will own the db.

It seems with a binary build on a popular OS, and a copy of what is in /usr/local/pgsql and /usr/local/aolserver, you would pretty much have a quick install OpenACS.

I hope this is the case, because I want to develop an app and quickly trash everything to load my packages up again for testing.

I like the idea of a prebuilt binary including aolserver and postgres. So you would do something like
tar xjzf /tmp/openacs-eval.tar.bz2
mv /tmp/openacs-eval/aolserver /usr/local
mv /tmp/openacs-eval/pgsql /usr/local
mkdir /web
mv /tmp/openacs-eval/service0 /web
./tmp/openacs-eval/run
And the run script could start postgres and aolserver. What's the minimum amount of user creation and permissions needed for this, and can we stuff it into a script and have it work reliably? Can we stuff all of that into a script? Are there other packages that do something similar (agata reporting tool comes with an optional version that includes php and php-gtk)? Should we put stuff into the standard versions or put it all within one directory like /usr/local/openacs-eval (ie, /usr/local/openacs-eval/pgsql) and all under one user openacs-eval? Unfortunately it looks like the lower-risk we make this, the more we guarantee that it doesn't look like the recommended configuration.
I really like this idea. God, it would make things so much easier!

Then we could just pay attention to the install scripts, and making sure they work well, rather than on the details of how to get everything set up.

If we did this well, then the production-level instructions could just expand upon this easy install.

I can help with testing, but I don't really have the time right now to help with setting this up.

This would rock!