One of the best things that anyone ever said was, "not enough gets said
about the importance of abandoning crap". Mr Glass was probably talking
about writing, but his words could well have been aimed squarely at any
well-established software development process.
If you see a small bug in a program on your desktop, you want to
something like this:
Figure out what package it's in
Look for an existing bug, if not file one
Get the source
Build it, try to reproduce the bug
Actually fix the bug
Test to see that it's actually fixed
Submit it then shepherd it through whatever processes come next
Much of this is monkey work if you know how to do it, or
incomprehensible arcana if you don't. Either way, it's important that we
abandon (or *automate*) this essentially uninteresting work.
In that spirit, at UDS I wrote a script that a lot of us have been
talking about for a long time. I called it
[start-hacking](https://launchpad.net/start-hacking), because I am
terrible at names and wanted to make a project. Right now, when you run
the script on your Ubuntu desktop, your cursor becomes a cross-hair.
When you click on an application, start-hacking will tell you the source
package that the application belongs to and where you can get the source
(both Ubuntu source and latest upstream if available). It's very
simple, but that's because it builds on a very large amount of work that
Canonical has done, mostly through Launchpad and Apport.
Eventually, we want to make the script a much more beautiful application
that gets you a built, runnable copy of any Ubuntu application or
library in a test environment. Our goal is to make fixing a typo (or
typo-level bug) as fast and as smooth as possible for any application on
All we need is a bit of time and a few volunteers.