On 2/5/2017 15:45, Stefan Fuhrmann wrote: > On 04.02.2017 17:53, Daniel Shahaf wrote: >> I wrote the following in a thread on private@, but the issue need not be >> discussed confidentially: >> >> Daniel Shahaf wrote on Fri, Feb 03, 2017 at 18:05:27 +0000: >>> I've noticed that some threads don't happen. >>> >>> Examples: >>> >>> - stefan2 solicited reviews of his authz branch. None happened. >>> >>> - SVN-4670 was filed with a trivial patch. Nothing happened. >>> >>> What worries me isn't the reduced activity — that's to be expected — >>> but >>> the complete *lack* of activity around these and other threads. That >>> activity level is lower than I would expect, even taking into account >>> that we're now mostly volunteer-run. > > Since more or less all committers contribute in their > spare time these days, not only the "intensity" of > interaction will go down. > > What we do see instead is that people appear to be > very active for a short period of time - a day or two - > and then become silent for a longer period of time, > maybe for weeks at a time. Those periods of activity > often don't overlap, which makes interaction harder. > The amount of back and forth discussion will probably > go down. > > I think that as a community, we need to adapt our > expectations / communication to that new pattern. > Things that might help: > > * Allow for at least 2 weeks of reaction time for > silent consensus etc. > > * Send an notification post to dev@ before starting > some larger work - people may not follow commits > closely anymore. Not as a vote or anything but simply > to keep people in the loop. > > * "Ping" a thread that you _really_ want feedback on > after 2+ weeks of inactivity. > > None of these need to be codified; they seem like > pretty common sense for a project with much more > asynchronism. > > -- Stefan^2.
Wouldn't it be possible to also consider getting some funding/sponsorship/donations? If enough money gets donated maybe some of the existing committers could arrange something with their dayjobs (as in being able to work like 20% on SVN) and therefore be able to contribute more time on SVN than they can right now? Also if some specific features would be put up as examples for committers who would volunteer to work on a feature, that might be caught up by companies who are willing to pay for getting the feature into the subversion core (LuaJIT used that approach, for instance, quite successfully). Regards, Stefan
smime.p7s
Description: S/MIME Cryptographic Signature