As far as I see, the "2.0" epic in JIRA and related efforts would start with a design document, even if it's small and "lean". The Apache project Wiki could be best to keep things in a place where people would look for it, unless you feel much better with Google Docs (or even JIRA, the 5/6 versions provide all sorts of tickets like "requirement" which we use in projects like the current one over here, too[?])
Any code change resulting in the "data" repository, would require a branch IMHO, either for the new "2.0" effort or by creating a "release branch" (not just a tag) for existing state so that any updates, patches or new device data that comes up can be maintained in a safe way. And not have a moving target of XML or other file structures. We probably won't get Git before actual 2.x "code" work is likely to start, are we? None of that must have Git, two separate SVN branches, one for the "stable 1.x" data, the other for the new direction seems fine. Either of them can be done on trunk, and we won't be able to just have a single Jenkins job soon, either. See many other Apache projects, they have several JDKs or repository branches built on different jobs. If the interactive GUI or service Eberhard also mentioned (we heard suggestions earlier, but never saw an actual web app for it) caters to device data, then of course doing so for more than one release (or migrating data, either by this service or on a lower level in SVN/Git) Werner On Tue, Aug 5, 2014 at 3:55 AM, Reza <[email protected]> wrote: > Cool. So something to think about between now and when we start 2.0 is: > > What are the core attributes for a device? > > So obviously things like id, model, marketing name, manufacture, screen > size, screen resolution, cpu type???, etc. > > Things to exclude are attributes which are specific to the browser or OS. > But we need to repeat this exercise for them too. > > An issue, how do we bridge the legacy devices which have some kind of > embedded browser and we can only deduce the browser/OS attributes from the > device. Also, how do we classify desktops and bots as devices. > > Just some food for thought... :) > > > ________________________________ > From: eberhard speer jr. <[email protected]> > To: "'[email protected]'" < > [email protected]> > Sent: Monday, August 4, 2014 9:19 PM > Subject: Re: Tracking new phones for DeviceMap data > > > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > I think once the 'tables'/views of the RDBMS I mentioned earlier are > visible etc the task of overhauling the data will be considerably > simplified, mainly because of the many ad hoc queries you can run on it. > > For example a query over 'versions' reveals that certain properties > were added to all or sub-sets of all devices 'in one go', all with the > same default values. > > My main point is : we then have a way to approach this systematically. > > By the end of the week... > > esjr > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v2.0.22 (MingW32) > Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ > > iQEcBAEBAgAGBQJT4DEjAAoJEOxywXcFLKYc7sMIAIJL8HO1T9H3uwrWnreJ7qfc > US5/3KfS054jFfQ/R+HMpJzfqizF0KQKT5eacr7Gl8OpTXrGYE6V5TtpDcqpX3dz > oZzscoyZQcWhAAOFss4TxRIsHf+vckc0+yprRVacAHSoT4tuaa/byK07CTVq7oAk > g2YMYXXvIsZ22g1SQI/SvQjHSl8X7STaMR++57gbgaOVdP6GxdgP5ylbwfCczkfS > lAEv0iWM7eNtO0qZHJCjzqHWynhm8mpv3UCwok35LkyYkW1XJD6+449LxVkcJ4Fd > pRqi1cLWgZMmpMtJnyHd0wVTgZ/NMuPF0GjFqat/fRg50bD6I3NiN6KogJ8j/UQ= > =TJNt > -----END PGP SIGNATURE----- >
