Hi Tim, Mark, etc Thought I might also chime in in advance of our Dev meeting.
While I sympathize with the sentiments expressed here, I'm a little unclear on what a vote would really buy us. I have no particular attachment to our legacy service implementations (e.g. ConfigurationManager, PluginManager, etc), nor any particular aversion to the dspace-services work, but I would observe that: (1) the new services aren't (yet) functional peers of legacy, nor as stable or tested. I can get into specifics, but I noticed at least 3 or 4 patches in the last month alone to new services. (2) although I have not closely examined the implementation of dspace-services, I have no reason to contest some of Graham Triggs' observations about stack complexity. But mainly I want to express a point about how development works: it is 'Darwinian' in the sense that the code that prevails is what individuals choose to adopt and work on given local needs or interests. We vote with our feet (or keyboards, in this case), rather than by ballot. A related example can illustrate my point: over the course of the last few years, we have had many spirited discussions about whether to 'deprecate' one UI platform or another. And while these discussions were (and are) interesting, and raise good questions about whether we have the resources to handle A or B, in the end what is deciding the issue is whether folks put calories into the maintenance and development of a UI platform. If we had 'officially voted' to deprecate work on all but Manakin, would that have prevented the creation of the Discovery UI, or Web MVC UI? Would that have been a good thing? I'd say that those interested in seeing the dspace-services platform advance should nourish it by incorporating it into their code, extend it, test it, etc. If it proves it's worth, it will gradually (or quickly!) displace the older idiom. But this is a quite different sort of commitment than a deprecation vote. My 2 cents, Richard R. On Mar 30, 2011, at 2:16 PM, Tim Donohue wrote: > Mark W, > > Thanks for getting down to the heart of the issue/question. > > I agree completely, that we should officially vote on our commitment to > DSpace Services, and officially mark all classes that will be replaced > with @deprecated notes. > > I've added a note about this to today's DSpace Developer Meeting agenda. > I cannot guarantee we'll get to it today (as we have a lot of topics > already), but I'll make sure to carry it over to next week's agenda as > necessary. > > https://wiki.duraspace.org/display/DSPACE/DevMtg+2011-03-30 > > Personally, I feel we should commit to Services Framework for all new > code additions, and deprecating the older ways of doing things. However, > I'd like to also hear any concerns others may have around going this > route, and whether there are ways to alleviate those concerns. For > instance, is there a way we can bring Services more "into the light", > rather than being in the "shadows", as Mark W puts it. > > https://wiki.duraspace.org/display/DSDOC/DSpace+Services+Framework > > So, if anyone has concerns/questions to pose around the Services > Framework, please feel free to do so in this email thread. > > To get this issue behind us, I'd like to call for an official 'vote' on > deprecating older classes in very near future (likely in our meeting > next week, April 6, if we can fit it in). So, I ask that we please try > to discuss any concerns/questions via email, if possible. > > - Tim > > On 3/30/2011 9:03 AM, Mark H. Wood wrote: >> There are several things going on here. >> >> o The dark side of greater modularity and Maven's magical >> behind-your-back dependency management is that we lose sight of >> some of the pieces. The Services Framework and core services have >> been In There for some time but we never notice them; "DSpace" is >> whatever you get when you check out svn/repo/dspace/trunk, plus >> some shadowy stuff that Maven takes care of (which now includes >> Services and will surely grow to encompass other parts of the >> product). >> >> o I don't believe there has ever been a completed call for consensus >> on Services. The developer community as a whole have not signed up >> to the notion that this is the way forward. Without a group >> commitment, inertia will rule and Services will languish. We've >> talked about it quite a lot but the talk has never advanced to a >> decision. >> >> We need to reach that decision and, assuming "yes", mark the >> ConfigurationManager and PluginManager as @deprecated so we're >> constantly reminded, "oh, yes, I should use the Whatsit Service now." >> If there are reservations then let's get them on the schedule and >> address them. We need to get beyond this point, one way or the other. >> >> >> >> >> ------------------------------------------------------------------------------ >> Create and publish websites with WebMatrix >> Use the most popular FREE web apps or write code yourself; >> WebMatrix provides all the features you need to develop and >> publish your website. http://p.sf.net/sfu/ms-webmatrix-sf >> >> >> >> _______________________________________________ >> Dspace-devel mailing list >> Dspace-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/dspace-devel > > ------------------------------------------------------------------------------ > Create and publish websites with WebMatrix > Use the most popular FREE web apps or write code yourself; > WebMatrix provides all the features you need to develop and > publish your website. http://p.sf.net/sfu/ms-webmatrix-sf > _______________________________________________ > Dspace-devel mailing list > Dspace-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/dspace-devel ------------------------------------------------------------------------------ Xperia(TM) PLAY It's a major breakthrough. An authentic gaming smartphone on the nation's most reliable network. And it wants your games. http://p.sf.net/sfu/verizon-sfdev _______________________________________________ Dspace-devel mailing list Dspace-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dspace-devel
