kfogel@collab.net writes:
> "C. Michael Pilato" <cmpilato@collab.net> writes:
> > As an aside (not directed only at Greg): These little weekend code
> > bombs might get folks all giddy for a few moments, but when the time
> > comes to support that code, it generally means that exactly one person
> > (Greg himself) has any idea what's going on there. This is not to the
> > benefit of the development community. Ideas developed and committed
> > piecemeal to the repository can be more easily digested over time by
> > everyone -- we all get to see not just the final functionality, but
> > the implementation decisions and steps, successes and failures, made
> > along the way. There's a wealth of wisdom to be gained by keeping the
> > process transparent in this way.
>
> Agreed, but the new native fs backend is not a code bomb. Greg posted
> a design, just like anyone would ask, and requested comments. (Maybe
> the --deltas dumpfile thing was sort of a code bomb, but that's
> smaller and more restricted than a whole new fs).
I never said this was an example of such a thing. Just advising
against posting the design, requesting comments, and then disappearing
into the night only to emerge with a "finished" product.
> And to be fair, svnserve/ra_dav have proven to be maintainable by
> others. It took a while for the rest of us to get up to speed, but
> nowadays Greg is not a bottleneck there.
And this will be the case for all such code bombs of non-negligible
size/complexity. Just don't discount as irrelevant the "while" it
took for the rest of us to get up to speed.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Wed Mar 31 00:00:06 2004