Agreed, it's all a matter of perspective. I use GWT-RPC alot, I actually don't see anything wrong with the underlying mechanisms, except for the SerializationPolicy and its generator. It's the SerializationPolicy generation that is the main culprit for the deficiencies perceived in GWT-RPC. However, I think that they don't have to be part and parcel and in GWT 3.0, I would personally like to see them separated and the SerializationPolicy made pluggable, instead of ripping out GWT-RPC. IMO, if the Serialization policy generation were made pluggable, then we could have different solutions that provide different tradeoffs and GWT devs could choose the solution that best suits them.
On Tuesday, July 19, 2016 at 1:34:28 PM UTC-7, [email protected] wrote: > > I was contrasting the blog post to the presentations Ray/Daniel have given > in the past, hence the Singular example preceding what you are quoting. > These are on youtube and I've also attended their presentations in person. > > I'm not trying to argue the compilation efficiency of GWT-RPC or get an > explanation of why it is so slow or say there aren't alternatives. If I am > giving off those impressions, I apologize. > > I am attempting to point out (poorly I suppose) that the value of the > feature largely stems from one's perspective or point of view. When one > uses GWT-RPC, one trades the compile-hit for the functionality. Similarly > with SOYC reports or massive numbers of permutations, you can generate > them, however they incur a time/space cost. These are choices, they have > costs, and my point has largely been to let the users decide if the > trade-offs fit their needs or not. (Please don't explain how SOYC/Perms > don't impact incremental compile!) > > Most of the GWT-RPC criticism here (and in the blog post) are purely > technical detail about some of its shortcomings and its impact on > incremental compile. From the detailed technical point of view, > Daniel/Thomas make accurate observations and there is not much to discuss. > However, at the feature level of GWT-RPC, given its flaws/quirks, many > folks would still opt to use it in at least some situations with the cost > it incurs. > -- You received this message because you are subscribed to the Google Groups "GWT Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/google-web-toolkit. For more options, visit https://groups.google.com/d/optout.
