Assuming that "module" is a set of RPM packages then it's exactly what Insim does - it tracks module sizes (installation size, download size, and other numbers) across time and distro versions. Someone would need to define what base module is (which packages it consists of) and then Insim will track its sizes.

For example, the Docker base image is defined by:

https://git.fedorahosted.org/cgit/spin-kickstarts.git/tree/fedora-docker-base.ks

Those %packages with coreutils-single added would be a helpful metric.

I've just added fedora-docker-base module. Note that due to nature of kickstart files Insim may not be 100% accurate (Anaconda may install additional packages not listed in %packages, and kickstart file removes some files too.)

How do you go about adding new modules to insim? Is it a database thing entry? Or a config file? or..?

Is it something that we could delegate out to some modularity working-group members so that they could add new modules on the fly as we work on them?

Way further down the road, we may someday have a whole namespace of modules that packagers can work on in a 'dist-git' style format, much like the way they currently work on rpm spec files. Could we automate something (some day) where newly created Fedora modules get automatically added/updated in insim?

(I can file separate tickets for much of the above if that would help..)

Insim is proof of concept, for now new modules need to be added directly to database. If people like the project then I'll implement some way of adding modules directly by users. Until it's done please ping me to have them added.

I guess that in future modules will be somehow defined within Fedora itself. Insim could then import their contents.

Thanks! We'll be talking about it more in the modularity working group, I expect.