roberts
Group: Members
Posts: 4983
Joined: Oct. 2003 |
|
Posted: Aug. 04 2005,05:07 |
|
This feature was requested from QA yet I never received any feedback and therefore assumed all is well. Your is the first feedback on this particular feature.
This suggests to me that maybe we should consider wider beta testing or even public beta testing. I see that some of the other distros now have public beta testing and some even make formal announcments that a public beta testing version will soon be forthcomming.
I am sure that there will be some who say that our release cycles are too often. Yet in reality the cycles have not changed in frequency for a very long time. Yet too, it is not too cool to have to issue a plus .0.0.1 bug fix release, as has been done that last couple of releases.
I also realise that my contributions to DSL has increased it flexibility tremendously and therefore its complexity resulting in quite a burden for formal QA testing. I even made a formal testing grid. Its size is intimidating and nearly impossible for anyone person to test every new feature against every possible install type. I also realize that the change logs for the last couple of releases have been quite large, again a burder to test everything in a timely manner. I had completed all the major features in this release on July 26 and had been waiting for feedback. This date was choosen for two reasons. One a full week had gone by, and two I wanted to release well before LinuxWord as I will be gone attending that event and not around to fix things or may not have readily available access to even resond in a timely manner. This is not to be construed as a complaint against those who currently help QA my pre-releases. Nor is this my excuse. It is only observations that I am public expressing.
The multiple wmet is not a show-stopper, as you have mentioned you can easily kill the pids of the left over wmnets.
Thanks for reporting.
If others want to weight in on the ideas expressed, please do.
Robert
|