On Tue, Sep 25, 2012 at 10:34 AM, <roger.moff...@gmail.com> wrote: >> >> Just to come back to your user mindset argument. I would argue that the >> people we really need to convince ARE the developers. Because the >> developers are the people who pick the tooling. And if CouchDB/apps do not >> look attractive, they will be ignored. > > > I'd just like to jump in and echo this. I'm the architect for a number of > commercial projects and will often be the person specifying the datastore > technology. I'm happy to see the progress that has been made with 1.2 in > recent months as this makes it viable for me to continue to advocate > CouchDB as a perfect backend for mobile projects in particular. > > But. > > When it comes to datastores for a backend API, the landscape is different. > For years my tooling of choice would have been mod_perl and MySQL, but Perl > has become fragmented and less relevant and is generally perceived as old > technology with a long list of deficiencies. > > The driver for change has therefore been the backend development language, > not the database per se. > > I'm finding now that Scala is a superb choice for replacing Perl for > reasons that aren't at all relevant to this discussion, but what is highly > relevant is that the choice of language can often drive the choice of > database depending on what is best supported. > > With Scala, and particularly when using the popular Play! framework, that > seems to mean Mongo - for example if you search the play site for CouchDB, > you'll get zero results, whereas with Mongo ... > > http://scala.playframework.org/documentation/2.0.3/Modules > > And so I find myself doing new projects with Scala and Mongo ... > > The detail of this isn't important, but I guess what I'm saying is that > it's probably more important now for the CouchDB team to focus on > advocating a great solution than it is to add new features to it. Building, > maintaining and promoting some great client libraries for emerging stacks > like Scala / Play would not be the worst place to start. > > I know that actions speak louder than words though, and to that end I will > be finishing my presentations on why Couch is such a great fit for mobile > and will be happy to contribute that to the project if it is in anyway > useful. > > Great to see the group active with passion! > > Roger
Nothing stop people to write cliens in scala. Imo that not the role of a server to also propose clients to access to it. We already have one of the easiest transport and API available (HTTP) . If none have written one in scala that maybe just because no scala users are using us or did the job to promote its lib. Personnaly before that I would dream to have a cool page presenting languages like redis.io does : http://redis.io/clients It would help a lot. As a side note I find the redis website a lot more useful than ours for the beginner or the guy that is just looking for a link ;) The wiki doesn't replace that . - benoît
