that is intriguing. I am one of the creators of BioNode, and, in fact, we merged with CBL last summer. You can create a BioNode using CBL flavors, as described here (for a minimal flavor):
https://github.com/pjotrp/cloudbiolinux/blob/master/doc/hacking.md
and
https://github.com/pjotrp/cloudbiolinux/tree/master/contrib/bionodeThe default CBL builds all packages (indeed). Flavors have been developed so you can do a subset, by selection packages in a YAML file, i.e.
https://github.com/pjotrp/cloudbiolinux/blob/master/contrib/bionode/main.yamland configuration defined with
https://github.com/pjotrp/cloudbiolinux/blob/master/contrib/bionode/fabricrc_bionode.txtIt is all really straightforward (kudos to Brad for that). And because it is Python and YAML it is quite hackable. The reason we have merged with BioNode is to be able to share development resources. Once you define your own flavor, you benefit from the work that is being done on CBL.
The only thing I don't like about CBL is that Cloudman, a separate target, has a high integration with CBL code - I think that is unnecessary. Cloudman should be designed as a CBL flavor.
With time I am building more installations using CBL flavors. I plan to add support for Chef or Cfruby (the latter I also created myself) as a post-install facility. These tools are being used, but are as yet not integrated as an option. But other than that, if you see CBL as a flexible VM builder, CBL delivers the goods on many platforms.
If you invest in CBL, it will pay off in time. That is my experience.
Pj.