It is my impression (backed up by a hasty count) that the rate of bug
detection is not diminishing. There was a brief lull but there has been a lot
of activity from about 2.0.10 on. Assuming a fixed proportion of
the existing bugs get discovered in any fixed interval, that would indicate
that the number of undiscovered bugs is much larger than the number
discovered so far. I.e. we haven't made a dent.
I would think it a bit dangerous to leave a "stable" kernel in that
state ... but your guess is better than mine about the true state of
affairs. Is Alan keeping count of the dates of bug reports as well as
the actual fixes? If so he could graph the reports against time and
tell us what the curve looks like. Logarithmic paper time.
Hmm .. another point is that as soon as you cease work on 2.0.x it will
be taken as being the last word in stable kernels by everyone and their
billygoat, which means that both will immediately install it. That
will multiply the number of volunteer testers by about 100, which will
accordingly multiply the rate of bug discoveries by 100.
Peter T. Breuer
,---------------------------------------------------------------------------
|Departamento de Ingenieria de Sistemas Telematicos, Universidad Politecnica
|de Madrid, Escuela Tecnica Superior de Ingenieros de Telecomunicacion,
|Ciudad Universitaria, E--28040 Madrid, SPAIN.
|Tel. Office : +34 (1)336 6831
| Fax : +34 (1)543 2077 or 336 7333
|Internet : <ptb@eng.cam.ac.uk, ptb@comlab.ox.ac.uk, ptb@dit.upm.es>
| URL : http://www.dit.upm.es:80/~ptb/
|Other : GCS/E/M d H- s-: !g !p+ !au a w+ v+ C++ UL++++ UH+ UV+
| : P-- L+++ 3+ N+++ E--- K++ W+ M-- V-- po- Y t+ !5 !j R- G?
| : tv++ b- D+ !B e+++++ -u h+ !f r* n+ y?
`---------------------------------------------------------------------------