Friday December 14th, 2001
As you may have noticed, we've taken the forums down, as we are hitting our bandwidth limit with them. We're working on getting them up and running again as soon as possible.
#38 Strauss-bashers... can we chill just a bit?
by rickst29 <email@example.com>
Monday December 17th, 2001 11:25 PM
You are replying to this message
I don\'t think it is Mozilla\'s interest to recommend that interested people with thought provoking questions should go away (and/or use IE). I personally agree with comments on both sides of the question:
1) Strauss is correct in recommending a statistical analysis to find if new problems are being exposed within particular \"old\" objects/functions at a higher than other code. It sounds paradoxical, but it is generally accepted in the software Q/A business that the number of UNEXPOSED bugs remaining in a module is highly cooordinated to the number you\'ve ALREADY FIXED. (If certain objects and methods keep showing up in problems, they should perhaps be divided into smaller components).
2) However, Strauss may be going to far in declaring it to be a \'system level\' problem (I assume that Strauss has not actually performed a design review from the repository.)
3) EVERYONE is correct (and agrees!) that the nightly build of the last couple weeks have been very unstable until just the last couple of days.
Where the disagreement appears to be is that Strauss (and several other end users) feel that the Nightly Builds should now be approaching alpha/beta quality. While the development team and several other users are staying with the declaration \"they might work and they might not\".
I am not sure what capabilities exist within Mozilla Development and Q/A for running automated regression tests. Again, EVERYONE agrees that the number of features to be tested is far to large for typing and mouse-clicking by humans. But, it doesn\'t seem reasonable that \"end users\" should be suffering destroyed profiles so regularly.