You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@uima.apache.org by Richard Eckart de Castilho <re...@apache.org> on 2020/02/01 08:14:43 UTC
Re: doing a uima-wide parent pom release?
On 31. Jan 2020, at 17:31, Marshall Schor <ms...@schor.com> wrote:
>
> I'd like to release this. Any objections? The release would have just this one
> change.
We have the profile "enforce-compatibility" in the uima-uimaj parent pom which could
also be moved up into the parent. It is activated by a marker file, so it shouldn't
break anything if it is added there. At least uimaFIT is also using this profile.
There is a "spotbugs" profile in the uimaFIT parent pom which could be moved up.
And there are the profiles "pmd" and "jacoco" in the uima-uimaj parent pom which could
be moved up.
"spotbugs", "pmd" and "jacoco" are all not active by default, but can be explicitly
activated if desired, e.g. during Jenkins builds.
-- Richard
Re: doing a uima-wide parent pom release?
Posted by Richard Eckart de Castilho <re...@apache.org>.
On 1. Feb 2020, at 20:42, Marshall Schor <ms...@schor.com> wrote:
>
> Anyone volunteering to do the work of moving enforce-compatibility, spotbugs,
> pmd, jacoco profiles?
I will move the spotbugs, pmd, and jacoco profiles.
The only problem with the "enforce-compatibility" profile is that it cannot
load its post-processing script from the classpath. That would require an
upstream fix/contribution - ... - or an unnecessarily complex Maven configuration
which would extract the post-processing script from e.g. the build-resources
artifact.
So I wonder if it is the right time to move the "enforce-compatibility" profile
or whether it would be better to wait until it can be done properly.
-- Richard
Re: doing a uima-wide parent pom release?
Posted by Marshall Schor <ms...@schor.com>.
Anyone volunteering to do the work of moving enforce-compatibility, spotbugs,
pmd, jacoco profiles?
-M
On 2/1/2020 3:14 AM, Richard Eckart de Castilho wrote:
> On 31. Jan 2020, at 17:31, Marshall Schor <ms...@schor.com> wrote:
>> I'd like to release this. Any objections? The release would have just this one
>> change.
> We have the profile "enforce-compatibility" in the uima-uimaj parent pom which could
> also be moved up into the parent. It is activated by a marker file, so it shouldn't
> break anything if it is added there. At least uimaFIT is also using this profile.
>
> There is a "spotbugs" profile in the uimaFIT parent pom which could be moved up.
>
> And there are the profiles "pmd" and "jacoco" in the uima-uimaj parent pom which could
> be moved up.
>
> "spotbugs", "pmd" and "jacoco" are all not active by default, but can be explicitly
> activated if desired, e.g. during Jenkins builds.
>
> -- Richard
>
>
Re: doing a uima-wide parent pom release?
Posted by Richard Eckart de Castilho <re...@apache.org>.
On 1. Feb 2020, at 09:32, Richard Eckart de Castilho <re...@apache.org> wrote:
>
>> We have the profile "enforce-compatibility" in the uima-uimaj parent pom which could
>> also be moved up into the parent. It is activated by a marker file, so it shouldn't
>> break anything if it is added there. At least uimaFIT is also using this profile.
>
> Hm, this profile involves at least the api-report.groovy post-processing file.
> One would need to check if that file can be moved to our uima-build-resources
> artifact...
Unfortunately, JApiCmpMojo.java is presently unable to retrieve the post-analysis script
from the plugin classpath - only from the file system.
But I'm not the first to notice ;)
https://github.com/siom79/japicmp/issues/257
-- Richard
Re: doing a uima-wide parent pom release?
Posted by Richard Eckart de Castilho <re...@apache.org>.
Anyway...
+1 for doing a parent-pom release.
Cheers,
-- Richard
Re: doing a uima-wide parent pom release?
Posted by Richard Eckart de Castilho <re...@apache.org>.
On 1. Feb 2020, at 09:14, Richard Eckart de Castilho <re...@apache.org> wrote:
>
> We have the profile "enforce-compatibility" in the uima-uimaj parent pom which could
> also be moved up into the parent. It is activated by a marker file, so it shouldn't
> break anything if it is added there. At least uimaFIT is also using this profile.
Hm, this profile involves at least the api-report.groovy post-processing file.
One would need to check if that file can be moved to our uima-build-resources
artifact...
-- Richard