Hello, My name is Christopher Baines and I am looking at applying to participate in the Google Summer of Code program. Since hearing about the program while participating in the Google Code In, I have been thinking of a few project ideas which I thought I would run by you.
I have been running Debian for over a year now, since moving from Ubuntu when I decided to try and get involved with the development of the OS. In that time I have been packaging, or attempting to package a number of applications, I currently have one package in Debian ant-contrib-cpptasks, and hopefully a few more on the way. While learning to package software for Debian, I found the process fragmented in a sense, many tools and programs that come together to make it all work (hopefully). While I see the enormous flexibility this gives it made the learning curve that little bit steeper. Thus I was thinking that for my project I could try and begin a Debian package IDE. After a bit of research I see this has already been attempted in a sense, however unlike the wizards (eg: http://sourceforge.net/projects/packin/), or the more complex programs (eg: https://code.google.com/p/debianpackagemaker/), it would be like a good programming IDE, less hiding, more helping. I was thinking of offering either a graphical (text fields, tickboxes) or text editor (gedit, kedit) style for editing the files, with lots of inline help (from both the program, and the relevant Debian documentation eg: the policy manual and the developer guide). Combined with lots of automated testing features from pbuilder and lintian, all usable directly from the program, it could speed up the process of learning to make packages and packaging itself by both reducing errors and bringing together the many applications that can help with the process. My second idea came from my recent usage of debdelta, a amazing tool to reduce download size and increase speed when updating Debian, perfect if like me you are on a slow internet connection. I would be looking to integrate it with apt-get and/or aptitude and improve the program its self by implementing a method of predicting the best approach for fetching the debs, and possibly looking at how to provide debdeltas in a better more comprehensive way, perhaps by allowing normal apt repositories to provide them? I am also quite open to other ideas if you would like to suggest any? Any thoughts? Thanks, Chris _______________________________________________ Soc-coordination mailing list [email protected] http://lists.alioth.debian.org/mailman/listinfo/soc-coordination
