Re: Update on development for 1.5

[Alex Boisvert]

Nice, thanks for bringing zinc back.

On Mon, Aug 15, 2016 at 11:38 PM, Antoine Toulme 
wrote:

> Just brought zinc back in. Had to do a little dancing around with a Java
> class to call out to the Scala main object.
>
> Here is the commit:
>
> [master 749acfe] Bring back zinc compiler, using the latest version and a
> runner class
>  5 files changed, 183 insertions(+), 5 deletions(-)
>  create mode 100644 lib/buildr/scala/org/apache/buildr/ZincRunner.class
>  create mode 100644 lib/buildr/scala/org/apache/buildr/ZincRunner.java
>
> > On Aug 15, 2016, at 4:49 PM, Antoine Toulme  wrote:
> >
> > On Aug 15, 2016, at 3:47 PM, Alex Boisvert 
> wrote:
> >>
> >> On Mon, Aug 15, 2016 at 3:24 PM, Antoine Toulme 
> wrote:
> >>
> >>> Two things had to go to make it happen:
> >>> -the Zinc compiler’s last release was 6 years ago, and it doesn’t seem
> to
> >>> play well with the latest. I have removed it.
> >>>
> >>
> >> Zinc's last release was about a month ago (July 3rd);  I think you might
> >> have been looking at older artifacts / older organization in maven2
> repo?
> >>
> >> https://github.com/typesafehub/zinc
> >>
> >> No biggie, just mentioning it.
> >
> > OK, I missed it:
> > http://central.maven.org/maven2/com/typesafe/zinc/zinc/0.3.12/ <
> http://central.maven.org/maven2/com/typesafe/zinc/zinc/0.3.12/>
> >
> > OK, let me try to bring this back in.
> >
> > I looked at this new Zinc incarnation from SBT and found that it has a
> first 1.0.0-X1 build out there:
> > http://mvnrepository.com/artifact/org.scala-sbt/zinc_2.11 <
> http://mvnrepository.com/artifact/org.scala-sbt/zinc_2.11>
> >
> > Here is the code:
> > https://github.com/sbt/zinc 
> >
> > I tried to use it but could not find the main class like the old one.
>
>

Re: Update on development for 1.5

[Antoine Toulme]

Just brought zinc back in. Had to do a little dancing around with a Java class 
to call out to the Scala main object.

Here is the commit:

[master 749acfe] Bring back zinc compiler, using the latest version and a 
runner class
 5 files changed, 183 insertions(+), 5 deletions(-)
 create mode 100644 lib/buildr/scala/org/apache/buildr/ZincRunner.class
 create mode 100644 lib/buildr/scala/org/apache/buildr/ZincRunner.java

> On Aug 15, 2016, at 4:49 PM, Antoine Toulme  wrote:
> 
> On Aug 15, 2016, at 3:47 PM, Alex Boisvert  wrote:
>> 
>> On Mon, Aug 15, 2016 at 3:24 PM, Antoine Toulme  wrote:
>> 
>>> Two things had to go to make it happen:
>>> -the Zinc compiler’s last release was 6 years ago, and it doesn’t seem to
>>> play well with the latest. I have removed it.
>>> 
>> 
>> Zinc's last release was about a month ago (July 3rd);  I think you might
>> have been looking at older artifacts / older organization in maven2 repo?
>> 
>> https://github.com/typesafehub/zinc
>> 
>> No biggie, just mentioning it.
> 
> OK, I missed it:
> http://central.maven.org/maven2/com/typesafe/zinc/zinc/0.3.12/ 
> 
> 
> OK, let me try to bring this back in.
> 
> I looked at this new Zinc incarnation from SBT and found that it has a first 
> 1.0.0-X1 build out there:
> http://mvnrepository.com/artifact/org.scala-sbt/zinc_2.11 
> 
> 
> Here is the code:
> https://github.com/sbt/zinc 
> 
> I tried to use it but could not find the main class like the old one.

Re: Update on development for 1.5

[Antoine Toulme]

OK,

we now have support for Scala 2.11.

Two things had to go to make it happen:
-the Zinc compiler’s last release was 6 years ago, and it doesn’t seem to play 
well with the latest. I have removed it.
-specs vs specs2. specs would not work with 2.11. I have removed it.

Also, as a bonus we now run our test suite on JDK8. Happy to report all tests 
pass and we have a great streak of green builds:
https://travis-ci.org/apache/buildr/builds 


We’re down to 8 issues for 1.5, 4 of them major ones.

Cheers,

Antoine

> On Aug 14, 2016, at 12:41 AM, Antoine Toulme  wrote:
> 
> Cool, thanks!
> 
> Further updates on development.
> 
> We have 6 major issues remaining on the 1.5 roadmap.
> Still no progress on the Scala 2.11 upgrade, because I’m lazy.
> 
> Cheers,
> 
> Antoine
> 
>> On May 22, 2016, at 3:47 AM, Peter Donald  wrote:
>> 
>> Actually went and updated custom_pom to do this as this shouuld be done 
>> anyway.
>> 
>> On Sun, May 22, 2016 at 8:02 PM, Peter Donald  wrote:
>>> We may also need to change the way we output dependencies from
>>> 
>>> 
>>>   com.biz
>>>   app
>>>   1.0
>>> 
>>> 
>>> to
>>> 
>>> 
>>>   com.biz
>>>   app
>>>   1.0
>>>   
>>>   
>>>   *
>>>   *
>>>   
>>>   
>>> 
>>> 
>>> As Maven assumes transitive dependencies and buildr assumes no
>>> transitive dependencies.
>> 
>> 
>> -- 
>> Cheers,
>> 
>> Peter Donald
> 

Re: Update on development for 1.5

[Peter Donald]

On Fri, May 20, 2016 at 4:05 PM, Antoine Toulme  wrote:
> I got a few things in for 1.5. The changelog reads:

Excellent work.

> Additionally, I have started reviewing the issues in jira and as best I could 
> tried to perform some triage. Some issues are old and don’t apply so much and 
> I’ve started closing them. If needed we can reopen issues if they should be 
> looked at again.

Sounds good.

> I have other things slowly brewing but I’d really like to knock off the old 
> issues and update to Java 8 before anything else happens.

The one thing I would like to see and are willing to do the work for
is moving 'buildr/custom_pom' addon back into the core. I would
ideally like to be generating useful poms with correctly defined poms
for consumption in other build systems. Besides adding a truckload of
tests the only other thing that I think would be useful to add is the
ability to specify "provided" scoped dependencies. I am not really
sure about how to go about this.

I had thought of adding something like

compile.provides << :mydependency

This would be equivalent to

project.compile.with :mydependency
project.pom.provided_dependencies << :mydependency

We may also need to change the way we output dependencies from


com.biz
app
1.0


to


com.biz
app
1.0


*
*




As Maven assumes transitive dependencies and buildr assumes no
transitive dependencies.

Thoughts?

-- 
Cheers,

Peter Donald