Towards 0.1 release

[Grzegorz Kossakowski]

Hi,

My personal goal for 0.1 release was to get the full language support. After a fairly extensive testing I can see that this has been achieved. Thanks to all that helped me getting there!

I'd like to cut 0.1 release before Scala days, probably on Monday. I can see two other things that could be included in 0.1 release:

• UiBinder support: https://github.com/nbauernfeind/scalagwt-gwt/tree/UiBinderSupport
• jsni support: https://github.com/maffoo/scalagwt-gwt/commits/scalagwt

I think that jsni support is fairly straightforward patch that includes tests and probably is ready to be merged in. It even goes with samples! :-)

When it comes to UiBinder there's a bit more work left to be done but I hope it can be finished by Saturday (when I plan to prepare binaries).

Also, I worked fairly a lot on updating showcase sample. The code has been synchronized with GWT's trunk but it still needs of cleaning up work. I relied on Intellij's automatic Java2Scala conversion which spits out not very idiomatic Scala code. I'll try to clean up everything for 0.1 release.

--
Grzegorz Kossakowski

[Simon Ochsenreither]

Shouldn't be a 1.0 release, seriously? :-)

We can use Scala with GWT, which involves converting Scala to JavaScript. Additionally, a working Eclpsie plugin wtih GWTs devmode. This is damn impressive ... marking it with 0.1 wouldn't do justice to that, imho.

[Nate Bauernfeind]

Don't forget there was no additional work needed to get it to work
with Maven or intellij's GWT support (only required an additional mvn
compile before running GWT in dev-mode as usual).

Maybe a better question is, what do you feel is missing to call it a
1.0 release?

[Matthew Pocock]

There's no shame in releasing it as 0.1, waiting a few weeks and then re-badging it as 1.0 if in the real world it does appear to be production-ready and feature complete. if for any reason you find there is still some work to do then you still have 0.2 on up to do it ;)

Matthew

--

Dr Matthew Pocock

Integrative Bioinformatics Group, School of Computing Science, Newcastle University

mailto: turingate...@gmail.com

gchat: turingate...@gmail.com

msn: matthew...@yahoo.co.uk

irc.freenode.net: drdozer

skype: matthew.pocock

tel: (0191) 2566550

mob: +447535664143

[Lex Spoon]

I don't know how everyone else feels, but I would propose that we
think of "1.0" as getting merged into Scala and GWT. I put "1.0" in
quotes because, if we do that, we will no longer have our own version
numbers so much. Instead, we will track the Scala and GWT releases.
When Scala releases their version 2.10, we will release
scala-library-gwt-2.10.jar, etc.

Lex

[Grzegorz Kossakowski]

I agree with Lex. The only problem I can see is that we still don't know when GWT team is going to agree to merge our patch.

I exchanged a couple of emails with their new techlead and they seem to be quite positive about Scala+GWT. However, they went through major reorganization. It's my own impression but I think most of the original GWT team is not working on GWT anymore. Therefore they are short on people (especially with compiler expertise) to even review our patch.

I'll try to find out more. Meanwhile, let's continue with 0.1.

--
Grzegorz Kossakowski

[Simon Ochsenreither]

Imho 1.0 would be nice, because it tends to lead to a higher number of people testing your code. You will get your changes pulled faster, because public interest often helps speeding things up in my experience.

If I where on a the GWT team and would learn that you intended to release 0.1 I wouldn't care as much as when you would release 1.0, because the connotations between 0.1 and 1.0 are _very_ different.

=> Cheap win through better marketing, imho.

[Grzegorz Kossakowski]

Well, I didn't want to go into details what's missing for 1.0 but let me give you a few examples that come to my mind:

• Optimized JS generation doesn't work yet due to some obscure bug I couldn't track down. This means we generate really huge JS. Remember what happened to Dart when they announced it and people found out that they generate 18k of JavaScript for hello world?
• Documentation is lacking. It's not something unique but we really need to do a bit more work on it.
• We don't ship any Maven artifacts. Therefore, even if technically it's possible to use Maven for building Scala+GWT projects then in practice it's not feasible yet.
• We still don't have continuous builds set up (meaning I don't want to have storm of pull request yet)
• Nobody apart from me knows how to build Eclipse plugin that supports Scala+GWT. Exact steps got a lot better due to work of Eclipse plugin team but still I need to document it.

Apart from first bullet others are fairly straightforward to address. I don't say everything has to be perfect for 1.0 release but I'd like to get at least a few things addressed here. Help is always welcome.

Also, I'd like to release more often from now.

--
Grzegorz Kossakowski

[Simon Ochsenreither]

Hi Grzegorz,

thaks for the information, that explains a lot!

• Optimized JS generation doesn't work yet due to some obscure bug I couldn't track down. This means we generate really huge JS. Remember what happened to Dart when they announced it and people found out that they generate 18k of JavaScript for hello world?

Absolutely!

• Documentation is lacking. It's not something unique but we really need to do a bit more work on it. 
  

Isn't that in line with standard Scala documentation practices? :-)

• We don't ship any Maven artifacts. Therefore, even if technically it's possible to use Maven for building Scala+GWT projects then in practice it's not feasible yet.
• We still don't have continuous builds set up (meaning I don't want to have storm of pull request yet)
• Nobody apart from me knows how to build Eclipse plugin that supports Scala+GWT. Exact steps got a lot better due to work of Eclipse plugin team but still I need to document it.

Apart from first bullet others are fairly straightforward to address. I don't say everything has to be perfect for 1.0 release but I'd like to get at least a few things addressed here. Help is always welcome.

 

Still trying to build current trunk :-/ I guess it just gets magnitudes easier to test when the code is merged into trunk...

Also, I'd like to release more often from now.

Great idea!


Thanks a lot for your great work!

Simon

[Ellis Breen]

Interesting, perhaps the lack of optimisation was partially the cause of runtime slowness I witnessed... Certainly, I've seen vanilla GWT run very quickly.

[Nate Bauernfeind]

On Fri, Apr 13, 2012 at 6:39 AM, Grzegorz Kossakowski
<grzegorz.k...@gmail.com> wrote:
> We don't ship any Maven artifacts. Therefore, even if technically it's
> possible to use Maven for building Scala+GWT projects then in practice it's
> not feasible yet.

It's very easy to ship maven artifacts. We don't even need to host
them in common maven repository. People are used to adding additional
maven repositories to get the artifacts that they care about.

This guy hosts his maven artifacts on github:
https://github.com/mpeltonen/sbt-idea

Maven repo:
http://mpeltonen.github.com/maven/

It works like a charm.

Nate
P.S. If you wanted to host them in scala-tools, sonatype, or maven
central, it will be more difficult since we need to re-use groupId's
that belong to other teams (org.scala-lang, and com.google.gwt).

[Grzegorz Kossakowski]

Nate,

It Smseems like we have the right person to work on this? :-)

-- Sent from Android

[Grzegorz Kossakowski]

[Nate Bauernfeind]

On Fri, Apr 13, 2012 at 6:39 AM, Grzegorz Kossakowski
<grzegorz.k...@gmail.com> wrote:

> We still don't have continuous builds set up (meaning I don't want to have
> storm of pull request yet)

I also have had good experience with http://travis-ci.org/ for
open-source continuous builds.

And yes, I'd be happy to do the setup for maven gwt. I'll start on it
as soon as you tag a release. =)