We applied as a mentoring organization for the Google Summer of Code 2011, but have unfortunately not been accepted this year. Perhaps 2012 ?

Here a couple of project ideas students could pick from / expand upon. Feel free to suggest more in comments.

Projects will require a decent knowledge of Java and various libraries, as well as a good understand of web applications. TODO: details
Necessary Magnolia introduction/training can be provided by the mentor.
Subversion and Maven experience a plus, but shouldn't be too much of a burden.

Unless otherwise stated, students will be mentored by Greg. TODO: details

In no particular order:

Infrastructure integration with our "Store"

Our recently born forge is a one-stop shop for developers, where they can develop/host their modules. Our "store" (integrated in the Magnolia AdminCentral, but also visible at http://store.magnolia-cms.com) is currently manually maintained; we'd love for modules to be automatically published in the store.

This project could require a good understanding of Maven and its various subsystems. (We've been thinking about using our Maven repository as the source of information for all things showing up in the store, rather than human input)

Other ideas in that area include improvements on the "client side" of the store, allowing Magnolia instances to download modules directly from the store.

Module downloader updater

Our current module mechanism relies on discovering modules (Magnolia plug-ins, i.e jar or zip files) on the classpath. We'd like to be able to host those outside the web application, and be able to download, install and start them without having to restart the container. This might or might not involve OSGi. This will involve close-to-core work within Magnolia, its module mechanism and version handling mechanism.

It is also related to the store project described above.

See Concept Module downloader updater for some more notes and ideas.

Standalone bundles

To improve the discovering user's experience, we'd like to have "double-click"-able artifacts.

Some more details and ideas at Concept Standalone bundles.

OAuth integration

We have an experimental OpenId module, allowing users to login to a Magnolia instance using their OpenId/Google/Yahoo credentials.

We'd like the same thing to happen for OAuth ("login with your twitter/facebook account"), and potentially consolidate those 2 in a single module.

This project would perhaps also be an open door for improving our whole login infrastructure. (Replacing an ugly jaas-based infrastructure; JSecurity is in the radar, for instance)

OpenSocial

We currently have an OpenSocial integration module, allowing OpenSocial gadgets to be embedded in Magnolia pages. How about exposing some Magnolia functionality as gadgets too ? (usage stats, various feeds, ...)

Social publishing

The current publishing in Magnolia publishes documents/pages between Magnolia "author" and "public" instances. It would be neat if users had the ability to publish stuff to social networks (twitter, facebook, reddit, ... but also youtube or flickr for instance) from within Magnolia as well.

With that, gathering the feedback from these platforms will also be very important for authors, so a parallel goal should be to integrate it. This might or might not mean using a platform/api that does this and integrate it in the Magnolia UI.

Google docs integration

There are at least two aspects to take into account when we think about integration Google Docs with Magnolia:

• Using google docs as a (shared and concurrent) text editor
• Publishing/embedding docs into Magnolia pages.

The aspects of sharing the docs are particularly interesting (the relationship between my Magnolia account, my Google account, and the users I want to share a document with).

Other interesting aspects: getting data back out of those documents into the admin UI of Magnolia. Say I embed a form in a Magnolia page. As an admin/editor, I'd like to see the results of that form right into Magnolia.

Your ideas !

There are plethoras of potential module ideas out there, we're sure you could come up with something we haven't thought of writing down here. Impress us and come up with integration ideas we haven't thought of !
(random hint: we did think of Wave, but have no idea what to do with it)