I highly encourage doing an experiment and coming up with a demo. This is how innovation happens!
But it seems to me that it's way too early to make migration plans. Even if the new thing is super-cool, we still have to support old code at least until all of Google moves off of it and that will take quite a while. Even with Super Dev Mode (where in theory there are no API changes) it is taking a while. So expect that it will have to run side-by-side with the old way for quite some time. I would also recommend keeping compiler and tool improvements separate from library improvements. If you come up with a new compiler that doesn't change any API's, there is a chance that everyone will be able to use it soonish and you'll get a lot of existing GWT users quickly. If the API's have changed then it's not really GWT anymore as far as existing projects are concerned, and if they're going to do a big migration, why not something else entirely like Dart? When you reinvent the API then you're limiting your audience to a small subset of GWT developers who are starting new projects and basically regrowing your userbase from scratch. - Brian On Saturday, April 6, 2013 7:25:36 AM UTC-7, Stephen Haberman wrote: > > Hey, > > Thought I would move this to gwt-contrib; during a compiler > improvements discussion in the last steering committee meeting, the > concept of having a dev mode run incrementally in Eclipse came up (e.g. > on save, reuses the existing AST), and Ray made a side comment > describing it as a "moonshot", which is I think is an apt description: > it's a big, hairy change, but with a huge payoff. > > I've been mulling this for awhile, and so have some notes here: > > > https://docs.google.com/document/d/1CQpfpsYq5Tf-oeDmwp9ETI9qNNt1ABIqu-N18nxTt9g/edit#heading=h.x1h3hdf7ui8l > > > If anyone wants to discuss it here (preferable) or in the doc, please > feel free. > > I am likely being naive in thinking this won't be *that* hard (I don't > think it's anything fundamentally new, just moving things around and > applying a heavy refactoring hand), but there are a few things that > would make this a "v3.0" change and likely require changes to user code > (not widget APIs/JSNI/etc., but things like super sourcing). > > But IMO, in the long term, e.g. if you want to be using GWT in 3-5 > years (like I very much do), and restore some of it's luster/sexiness, > I think now is a good time to make this investment. > > - Stephen > -- -- http://groups.google.com/group/Google-Web-Toolkit-Contributors --- You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/groups/opt_out.
