Just one reply inline for now On Wed, Aug 27, 2014 at 9:08 AM, Bertrand Delacretaz <[email protected] > wrote:
> On Wed, Aug 27, 2014 at 1:22 AM, eberhard speer jr. <[email protected]> > wrote: > > ...Considering goings-on on the incubator list, I'd like the Mentors to > > weigh in on this... > > I'm not sure what you are asking, IIUC the issue is how to manage the > device data in a way that offers some database-like features (which > ones exactly?) and how to publish the results. > > IMO the best way by far to manage the evolution of such data is via > structured text files, where one can check diffs to see what changes > between versions, and which make it easy (via source code control, svn > or Git) for multiple contributions to be integrated in a traceable > way. > > I know, that is an easy way (or as Eberhard phrased it recently "cheap bait" I believe;-) and goes along the lines of traditional "committer" views and models, Apache might be best used to. Similar to community-driven projects at Eclipse like Babel, continuous update and generation of a file based structure based on such input (to some sort of DB like structure at first) is possible, someone just has to work on that. AFAIK Adobe did a great deal of Eclipse Babel, but now it's only maintained by the few "webmasters" at Eclipse.org. If we stick to traditional checking out from SVN of XML files, we will not get a lot more people to update it than those who currently write code, too;-| Werner > Distribution is then just a matter of releasing those structured text > files, along with documentation and/or test cases that demonstrate how > to use it, like the devicemap-data and devicemap-client releases that > have been done. > > If this is not what you are asking, best might be to create a brief > list of requirements or use cases for that data management, and get > consensus on that first, before talking about how to implement it. > > -Bertrand >
