On Thu, Apr 12, 2007 at 07:48:56AM -0700, JT Moree wrote: > > I think a solution that avoids hacking Dirvish itself is much better > > until such time as the feature can be incorporated in a full release. > > This is key.. 'until such time'. We can't use the excuse that hacking > dirvish is a pain so 'let's never add a new feature'. > > I agree that we should not hack a stable release but let us not use that > as an excuse to never move forward. > > for example, what is the experimental code doing? is it the stable > branch with a few bugfixes? or are new features being added?
The big problem with any change is that changes need to be tested, and we don't have a mechanism for that. Testing means more than "does my patch solve my problem". The dirvish code is pretty twisty right now, and needs to be straightened out, but until it is, "small" changes in one part of the code can have unforseen consequences elsewhere. There is an experimental version 1.3.1 which refactors common code into a module, and replaces the shell script builder with Module::Build to make it easier to add automated tests. But that is not ready for use. Testing will be difficult, since we are wrapping around rsync, and we want to test that behavior without screwing up existing backups. The "not screwing up existing backups" is the key. We are installed at some important places, such as OSUOSL, and are backing up projects such as Mozilla and Gentoo and Drupal. I want to make sure we deliver 1.2.x or 1.4.x versions that do not damage backups for these projects. But I also want to see dirvish move forward. So if you want to move forward, help figure out a testing method. To start with, testing will be manual. We find some volunteers, build a few different test machines, design some observation methods and report results back to the list. Later, we can automate that. I will give commit access to the subversion server to individuals who take responsibility for testing their changes in multiple environments. I will help with the testing myself, and can devote an experimental dirvish server to trying out new code. Who else will participate in testing? Keith -- Keith Lofstrom [EMAIL PROTECTED] Voice (503)-520-1993 KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon" Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs _______________________________________________ Dirvish mailing list [EMAIL PROTECTED] http://www.dirvish.org/mailman/listinfo/dirvish
