Hi Michael, At least for my part, I'm not resisting getting rid of GXP - I even advocated it in my earlier response. What I'm resisting is churn. If someone has some meaningful progress to contribute and GXP is standing in their way then please by all means change it to Velocity.
If however it's a matter of new developers struggling, then we should definitely do something about this. We could either invest a bit of effort to add a few pointers to get people going with GXP, or invest some effort (probably a lot more) to change all our templates to something else. But looking at the mailing list archives (searching for "gxp" and "wave-dev" in my email) the last question that came up was in January from someone trying to embed the wave editor and it was more a side question? If people are having trouble they should email the list, we'll do our best to help them out and it would soon become obvious that we need to do something about this asap. All that said, this discussion is kind of academic unless we have someone willing to actually go and do the work. There's no philosophical objection from me. The next time someone needs to go do some extensive changes in the servlet code, if they feel up to it, they can switch the template system. If no one can commit to make that change, I'd be personally willing to write up some pointers to help ease the immediate pain. Dan Στις 18 Μαΐου 2011 12:28 π.μ., ο χρήστης Michael MacFadden <[email protected]> έγραψε: > It's not just the documentation. The only release we have is a Beta release > from 2008 and there hasn't been any activity on the project since 2009. For > all we know, Google isn't using or advancing this technology. There > certainly isn't a large and thriving community around it. Google could at > any moment decide to abandon it (if they haven't already). It just seems > risky. > > I know the general reaction to everything is to resist change, but on the > other hand if we didn't have a template system in place and we were doing an > evaluation of GXP and other template engines, I doubt that without an > insiders view in to Google, we would use GXP, based on the perceived lack of > support and activity. > > ~Michael > > > On May 17, 2011, at 1:43 AM, Yuri Z wrote: > >> I think that if GXP is good enough for Google, it is also good enough for >> us. Regarding lack of public documentation - maybe we just need to try and >> ask Google to open some more of it. After all we have some "google >> connections" in this project. >> >> 2011/5/17 Joseph Gentle <[email protected]> >> >>> On Sat, May 14, 2011 at 11:45 PM, Lennard de Rijk <[email protected]> >>> wrote: >>>> Hi, >>>> >>>> I also think it is something that was chosen because it was a Google >>>> project. However the amount of templates should still be pretty limited >>> (7?) >>>> and we might want to switch over to something more publicly documented? >>>> >>>> Greetings, >>>> Lennard >>> >>> Last I checked (a few months ago), most of the documentation for GXPs >>> is sitting on google's intranet. I think if google isn't going to >>> opensource the GXP documentation, it doesn't make sense for wiab to >>> use it. >>> >>> The principle of least surprise suggests we ditch GXP for something >>> more standard. >>> >>> -J >>> > >
