Searching for bugs can be very time consuming. It's hard to imagine what it's like for the developers, but it seems the end result is an ever-increasing and more confusing list of reports.
I would think there needs to be some kind of front-end weeding out process. One particular category that tends to stockpile is for feature requests. While many seem well intentioned, they can be difficult to understand, ill-conceived, and probably best seen in the context of other requests of a similar nature as well as the roadmap. Also, extreme things, like Sig 11 crashes, at this point (in my experience) tend to be very transient problems with a particular version that get fixed as soon as they come to attention. Yet they can lead to repetitive reports when someone uses that brief "bad" version. Perhaps there could be a list of at least recent cvs versions with date that could be labeled as "bad" or "do not use - update". When I have reported bugs, I find some of the decisions about labels to be difficult: what category, how severe. Maybe there could be a rethinking about how bugs are reported to help with these decisions. Greg
