On Tue, Dec 04, 2007 at 03:05:09PM -0800, Andrew Lentvorski wrote:
David Brown wrote:
Setting up an issue tracking system is overkill for many projects.
Wrong. Just dead wrong.
If it's worth publishing a repository, it's worth setting up an issue
tracking system.
Nonsense (just keeping the spirit of the discussion up). You are assuming
one particular development model. Whether a particular tool helps a
development model is a decision the maintainers of that project need to
use. Issue tracking makes the same kinds of assumptions that most revision
control systems make, of a centralized development model. Something like
linux kernel development, with a very different model.
If your attitude is like Linus, he doesn't want to see bug reports, he
wants patches, the tracking system would just be filled with noise.
This si even *more* true if the project only has one person who works on it
intermittently. That tracking system becomes a refresher list when that
person gets back to it.
That depends on that developer. I personally would find bugs being hidden
off in some database much more difficult to find than email messages
describing the problems, or even that just end up in a text file called
BUGS at the top of the code tree.
However, I agree that issue tracking systems are indeed quite useful, when
used properly. I have just seen too many times where bugs just vanish into
them to declare that they should always be used.
Dave
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
