[cvsnt] CVSNT gets more and more unstable (too much releases)

[Tony Hoyle]

Gerhard Fiedler wrote:
> Makes sense -- but I wouldn't say that this is a bad thing. You got
> notified of the issue. Now you have basically four cases:

The list also gives this plus the fifth case, where the reporter can't 
help but others replicate the problem and help nail it down.

Bug databases tend to be isolating.. you end up with a conversation 
between two people and little or no external input (in the worst case we 
disagree about the severity/existence of the bug and it just becomes an 
argument.. that happened a couple of times too).

> I'm not trying to push you into any direction here :)  I'm just throwing in
> my POV. But I never worked with a bug database open to the public for
> reporting, so there may be issues I have yet to experience. OTOH quite a
> number of projects do this, and it seems to work well in many cases. But
> then, this is a matter of work style, too...

It depends.  For example if you look at the mozilla bug database, 
probably 90% of what's there is repeats of other stuff/insufficient 
information, etc.  they have several full-time staff able to handle it 
so are able to jump on these things but it's obviously taking 
maintenence time to do so.  That also has the advantage of being a 
high-profile project, and will get more volunteer input anyway.

Tony