Thanks for the thoughts on this. Here's what I will do, then.
* create a new "mothballed" directory under trunk (alongside core, components, examples etc). * move the html viewer, sql security and sql security into this directory. I'll preserve the directory structure for these, so that moving it back, if we ever wanted to, would be really easy * remove the references to these modules from the root pom * update the root README.md, and create a mothballed/README.md for anyone who explores this stuff via github. * update the Isis website to indicate that these components have been mothballed. Chip in if you have any further thoughts/improvements/refinements on this. Cheers Dan On 27 March 2013 19:29, Kevin Meyer - KMZ <ke...@kmz.co.za> wrote: > I understand the desire to cut back on components that are not > contributing to the codebase. > > There is no reason to distract (potential) contributors with unsupported > code. > > I suppose I would prefer it if the code were still accessible (remove it > from the Isis pom framework, sure) so that if someone develops a > strong desire to resurrect and maintain it, they can! > > PS: I like some of the features of the HTML viewer (e.g. automatically > contributing items that have been recently viewed, to action method > parameters). And Dan has provided a suggestion on how to add this > functionality to Wicket... > > > Regarding dropping units that don't receive updates - I wouldn't use a > plain time-based criterion.. it should only kick in if the component in > question requires developer effort (e.g. an accepted change > elsewhere in the framework requires effort to update - beyond > automatic refactoring, that is). > > > > > On 21 Mar 2013 at 12:34, Dan Haywood wrote: > > > All, > > Something I've been meaning to raise for a while is whether we should > > retire the HTML viewer. (I did mail Kevin about this offline because he > > has historically been the most active user of this viewer; I'll let him > > respond with his views rather than me summarizing them) . > > > > My view is that we're only actively developing Scimpi and Wicket, and > that > > both are more functional than the HTML viewer that is not being > developed. > > I'm keen that we trim our codebase of stuff that isn't in active > > development. I think that we should therefore retire this viewer. > > > > To generalize the point, it might also make sense to institute some sort > of > > rule that we retire stuff that hasn't had any updates after a certain > > period of time, eg 12 or 18 months. (Perhaps such a rule might require a > > formal vote, but probably worth discussing here) > > > > Anyway, your thoughts welcome. > > > > Thx > > Dan > >
