Can upstream authors really be trusted to provide decent quality software?
Traditionally, a lot of the QA work that goes on in open source software is done by distributions (such as Debian). They take the `upstream’ code from the software authors, and then provide a packaged version to their users. There are then two classes of bugs to worry about: upstream bugs and packaging bugs. Since it’s not usually clear to users which bugs are in which category, they tend to report all bugs to the packagers, who often then fix even upstream bugs in their own packages. The effect of this is that distribution-provided packages are often more reliable than upstream ones (since upstream don’t get to hear about many of the bugs), and different distributions have fixed different bugs, with no coordination between them. With Zero Install, bugs get fixed upstream. So, the ‘Debian developer’ who currently fixes Gimp bugs would still do the same job, but as a ‘Gimp developer’ instead. Thus, the fixes would benefit everyone, not just Debian users. Of course, the other reason why
