Mark Brouwer wrote:
other issues but I'll give it a try this evening. IIRC there was some
discussion on incubator about mentioning the issues for graduation, if
people can help me out with listing them here I would be grateful. Those
I can think of is the namespace com.sun -> org.apache change and
attracting fresh blood.
There are a number of low-hanging fruits that come to mind:
- unifying and retiring the Sun mailing lists at river-user/-dev
- maybe conversion of the mail archives? this would be nice to have, as
many postings still contain important discussions.
- I don't know if this has been discussed previously, but maintaining two
separate websites usually does not work. The often-cited "community" does
not seem to exist. Maybe everybody is just hiding behind corporate walls,
but then again other Apache projects do not have this problem either.
- adoption is unlikely to increase as long as the only practical way to
get started is to _not_ use the river distribution, but rather the starter
kit. I recently pointed a friend to river and told him to give it a shot
(he was interested in the Jini service model). After two days he gave up
and told me that "this stuff is unusable" - this from someone with a lot
of experience and an open mind. I have therefore started to adapt &
refactor some of the generated starter-kit configurations to an
out-of-the-box river distribution with separate one-click .bat files for
"all services" and the browser. I still need to work out some paths and
make it pretty (scripts in bin, configs in config, auto-declare RIVER_HOME
etc..) but then again I had to start somewhere. The project *must* pass
what I call the 60-seconds test, because that's the acceptance hurdle you
have to take. Download, unzip, run. Anything else is a usability fail.
I'll skip the usual suspects of longer-term goals for now, since they are
already 5 years overdue.
Ultimately, everybody who has invested effort, time or money in Jini
should ask themselves what an incubation failure would mean.
Holger
