Which CVS branch to work on?

[tpeierls]

I lost track of what was going on in the DWR CVS repository, and I'm trying to find my way back so I can address some long-standing issues with the DWR-Guice integration.

I notice that Joe has edited many if not all of the files in org.directwebremoting.guice since the Version_2_0_2_RC3 tag was applied. The revision message is "adding missing comments, and removing warnings from guice".

Are these being considered for release as part of 2.0.2? I hope not, because I can't compile with against these revisions. I get this message about one of my extensions of AbstractMapContextScope:

[My class] is not abstract and does not override abstract method <T>remove(ConcurrentMap<?,?>,Key<T>, String, InstanceProvider<T>) in ContextRegistry

AbstractMapContextScope has a remove method declared with:

public <T> boolean remove(ConcurrentMap<Key<T>, InstanceProvider<T>> registry, Key<T> key, String keyString, InstanceProvider<T> creator) {...}

ConcurrentMap<?, ?> is not the same thing as ConcurrentMap<Key<T>, InstanceProvider<T>>, so that explains the error message. Not sure what to do about it, but I'll try to make things nice without resorting to raw types.

Should I be addressing this and other issues in the HEAD revisions, or is there a branch that I should be using?

What happened to Subversion?

--tim

[tpeierls]

I think I've figured out how to work on HEAD and Branch_2.0 in two separately checked out directories. I also have a fix for the code in HEAD.

So my question is now reduced to "Is it too late to check in DWR-Guice enhancements to Branch_2.0?"

--tim

tpeierls wrote:

I lost track of what was going on in the DWR CVS repository, and I'm trying
to find my way back so I can address some long-standing issues with the
DWR-Guice integration.

I notice that Joe has edited many if not all of the files in
org.directwebremoting.guice since the Version_2_0_2_RC3 tag was applied. The
revision message is "adding missing comments, and removing warnings from
guice".

...

Should I be addressing this and other issues in the HEAD revisions, or is
there a branch that I should be using?

[Joe Walker-3]

I was in the middle of writing a reply when I saw your update

On 9/7/07, tpeierls <tim@...> wrote:

I think I've figured out how to work on HEAD and Branch_2.0 in two separately
checked out directories. I also have a fix for the code in HEAD.

So my question is now reduced to "Is it too late to check in DWR-Guice
enhancements to Branch_2.0?"

Hmm - I don't what to get into having to do multiple releases of several branches, so I think we should keep new features to HEAD.
If the enhancements are:
- safe
- can be moved to HEAD easily
- not going to require many extra releases

Then it might make sense to put them into 2.0.

Joe.
 

--tim


tpeierls wrote:

>
> I lost track of what was going on in the DWR CVS repository, and I'm
> trying
> to find my way back so I can address some long-standing issues with the
> DWR-Guice integration.
>
> I notice that Joe has edited many if not all of the files in
> org.directwebremoting.guice since the Version_2_0_2_RC3 tag was applied.
> The
> revision message is "adding missing comments, and removing warnings from
> guice".
>
> ...
>
> Should I be addressing this and other issues in the HEAD revisions, or is
> there a branch that I should be using?
>
>

--
View this message in context: http://www.nabble.com/Which-CVS-branch-to-work-on--tf4387976.html#a12559290
Sent from the DWR - Dev mailing list archive at Nabble.com.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Joe Walker-3]

On 9/5/07, Tim Peierls <tim@...> wrote:

ConcurrentMap<?, ?> is not the same thing as ConcurrentMap<Key<T>, InstanceProvider<T>>, so that explains the error message. Not sure what to do about it, but I'll try to make things nice without resorting to raw types.

Sorry I didn't spend enough time testing for backwards compat, and just added the types in  where I could - if you are happy to fix it, that's great or you can tell me about problems.
We should perhaps add some Guice examples to DWR.WAR which should help prevent some of this.

Joe.

[tpeierls]

I have a fix and will check it in. It means adding an interface InstanceMap and class InstanceMapImpl -- this is to able to say InstanceMap<?> and get the effect of ConcurrentMap<Key<?>, InstanceProvider<?>> where the two ? wildcards refer to the same type.

I'll try to grok a pattern from the other examples in order to add a Guice example/test.

--tim

Joe Walker-3 wrote:

Sorry I didn't spend enough time testing for backwards compat, and just
added the types in  where I could - if you are happy to fix it, that's great
or you can tell me about problems.
We should perhaps add some Guice examples to DWR.WAR which should help
prevent some of this.

[tpeierls]

I'm inclined to wait, then, and leave 2.0.x alone. AFAIK nothing is actually broken, just lacking some enhancements a few people have requested.

But I still want to use these enhancements myself in production, and I'm not sure the right way to go about it. How bleeding-edge is HEAD?

--tim

Joe Walker-3 wrote:

On 9/7/07, tpeierls <[e-mail]> wrote:
> So my question is now reduced to "Is it too late to check in DWR-Guice
> enhancements to Branch_2.0?"

Hmm - I don't what to get into having to do multiple releases of several
branches, so I think we should keep new features to HEAD.
If the enhancements are:
- safe
- can be moved to HEAD easily
- not going to require many extra releases

Then it might make sense to put them into 2.0.