|Version 59 (modified by anonymous, 10 years ago) (diff)|
Google's Summer of Code 2006
Django (and the Lawrence Journal-World) is a sponsoring organization for the 2006 Google Summer of Code. (Read Google's page for more information on how the program works.)
If you're interesting in mentoring -- supervising a student in work on Django-related activities -- sign up over at Google and then add your name below:
- James Bennett
- Simon Blanchard
Student applications open May 1st. Watch django-users for an announcement when the signups open.
If you'd like to get started on your proposal early, we'll be looking for a few things. You'll need to have a concrete task in mind (some ideas are below) along with a solid idea of what will constitute "success" (you tell us). We'll want to know a bit about you -- links to previous work are great, if any. You'll also need to provide us with a rough schedule of milestones so your mentor can know if and when to nag you :)
Note that none of the ideas below are good enough to be submissions in their own right (so don't copy and paste)! We'll want to know not just what you want to do but how you plan to pull it off.
And don't feel limited to the ideas below -- if you've got a cool project you want to work on, we'll probably be able to find you a mentor. We plan on approving as many projects as we possibly can.
Here are some suggestions for projects students may want to propose (lazyweb: please add to this list!). This isn't by any means the be-all and end-all of ideas; please feel free to submit proposals for things not on this list.
Try to scope ideas/proposals to the 4-month timeline -- simply proposing to fix a ticket or two will probably result in your proposal being rejected in favor of a more ambitious one. The SOC does not cover activities other than coding, so certain ideas ("Write a more detailed tutorial" or "Create demonstration screencasts" or "Add a pony") are not suitable for inclusion here.
- Build a public repository for reusable Django applications. This probably includes:
- a standard way of packaging Django apps for distribution
- infrastructure for automatically downloading and installing packaged Django apps.
- Create Django packages and buildscripts for Linux (various), FreeBSD (There already is one www/py-django), OS X (Fink and Darwinports), etc.
- Create a generic, reusable search/indexer component (capable of htmlentities).
- Enhance the admin's support of rich media:
- richtextfield using tinymce integrated with file explorer mentioned below
- create a file explorer for MEDIA_ROOT in the admin: browsing and making directories, uploading files
- automatically make thumbnails from ImageFields (with programmable "PIL modes" (i.e. adding borders, overlaying png's))
- validation of image type, sizes, etc
- support for uploading large files through the admin with a pretty progress indicator
- Implement ModelInheritance.
- Make something like models.RequestModel allowing to pass request to the model, so it would be possible to access the current user etc. from the model, the admin should be made to pass request if the model inherits from models.RequestModel
- Implement SchemaEvolution.
- Implement a set of reusable AJAX components using Dojo and Django as a backend, including:
- Non-admin widgets
- Admin widgets to bring Django closer to, e.g., MicrosoftAccess (NB: if Django ever becomes anything LIKE Access, I'm leaving. -- JKM) for creating database applications:
- MultiSelectComboBox that works for thousands of rows and can display multi-column results
- Easily create any levels of in-lined editing widgets
- Compose admin pages of AjaxPortlets that can be connected together
- Add spell checking to CharField and TextField with a well thought out interface (getting the interface right will be significantly harder than the backend code).
- Change the permissions system to be more flexible (ACL's, Zope-style role based authorization, etc.)
- Implement custom redirects and field templates for the admin system. (This isn't very ambitious as-is.)
- Refactor package settings so things like django.template are usable outside of django. details
- A better built-in webserver and an officially supported methodology to connect to lighttpd and similar servers. While the apache+modpython solution is highly performant there are numerous situations where that setup is unfeasible or overkill. It is great that Django scales up well, if would be great if it scaled downwards as far as setup and deployment goes.
- Streaming uploads and downloads. Uploading a large file can break django as the files are stored in the request. This is a security flaw as well, denial of service attacks could be as simple as uploading one large or more smaller files simultaneously.
- Parametrized templates. Currently the only way to use the content of another template in a place that does not contain blocks is to include it. But that includes the whole template. It would be very nice if one could include blocks defined in other templates to be evaluated in a certain context. (this turned out to be harder to explain than I thought, basically what I'm saying is to introduce a new kind of block that works like a function, where you could call/import a block and pass it some parameters)
- add dabbledb (http://dabbledb.com/utr/) like functionality to the django admin. e.g. create a project and super user with one command line statement, fire up the development web server, and import/create/refine your model using live data from directly within the django auto admin.
- A pet store application, similar to the ones which are shipped with java and .net. This server should be built as a showcase for other people to learn how to implement django features. (mysql's Sakila DVD rental app is also good)
- Code a themeable generic theme (or two) for use in people's Django applications.
- Build a Django CMS application with on-the-fly multiple language support for content (specific reference to i10n)
- A Row/Object level authorisation system. so you can say 'User A has read- access to customer 123' and user B has write access to customer 123', and user C has read/write to customer 345
- An option in the model to keep a copy of past states of rows/objects that have been modified. If this option is set, it becomes possible in the admin to see a diff between two past states of an object and get back to a past state of an object.