C. Michael Pilato wrote on Wed, Aug 22, 2012 at 14:23:39 -0400: > On 08/22/2012 02:02 PM, Branko Čibej wrote: > > On 22.08.2012 18:28, C. Michael Pilato wrote: > >> * besides, philip's freeze/unfreeze offers a far more logical way to > >> get honest-to-goodness-read-only-ness. > > > > The only caveat here being that using repo freeze could not be used > > instead of start-commit blocking commits whilst importing code from a > > dump file, which ASF infra does. However, I'm almost certain that the > > changed start-commit semantics would be sufficient to that end. > > Agreed.
+1. I think we can live with a stream of created-then-immediately-aborted txns whilst running a load.
