Le mercredi 11 octobre 2017, 23:47:54 CEST Robert Scholte a écrit :
> I think the conclusion is: 3.5.1 is not correct
>
> it should either be:
> 3.5.2 without the 2 classloader related issues.
ideally with an option to activate the classloader changes (for people knowing
what they are doing or
Github user DaGeRe commented on the issue:
https://github.com/apache/maven-surefire/pull/167
I adjusted the current commit and PR to code style, could you re-review
this one?
---
-
To unsubscribe, e-mail:
I think the conclusion is: 3.5.1 is not correct
it should either be:
3.5.2 without the 2 classloader related issues.
or
3.6.0 including the classloader related issues, but probably improved with
a system property to switch back to the 3.5.0-behavior. We also need to
improve ITs and
Github user Tibor17 commented on the issue:
https://github.com/apache/maven-surefire/pull/167
Pls do it in a new PR. Here is link for Eclipse/Idea code style xml
https://maven.apache.org/developers/conventions/code.html
On Wed, Oct 11, 2017 at 8:24 PM, Tibor Digana
Github user Tibor17 commented on the issue:
https://github.com/apache/maven-surefire/pull/167
You completely changed code style.
Please use ASF Maven code style because here I do not see relevant changes.
On Wed, Oct 11, 2017 at 5:52 PM, DaGeRe
If we did that, then wouldn't the build get really slow every time you hit
that 5 minute expiration?
On Wed, Oct 11, 2017 at 11:49 AM Michael Osipov <1983-01...@gmx.net> wrote:
>
> > I just wanted to get thoughts on a possible new feature. Right now, my
> > company has thousands of Maven modules
I agree with Mark about the rule to split: 1 repo = 1 lifecycle
But like we said with the number of repo and the flat organisation in infra
side it won't be easy :(
About sub-modules or sub-trees I never used them myself but I always had
various bad feedbacks.
About dependencies cycles ... yeah
GitHub user DaGeRe opened a pull request:
https://github.com/apache/maven-surefire/pull/167
Speedup Standard Output if Tests
Currently, surefire creates a new byte array with size of the input *3 for
saving the unescaped input. This made surefire way slower than running a test
> I just wanted to get thoughts on a possible new feature. Right now, my
> company has thousands of Maven modules that are all on snapshots with an
> updatePolicy of "always" (basically everyone gets the latest build of
> everything, kind of like a poor man's monorepo).
Why "always" then and not
I just wanted to get thoughts on a possible new feature. Right now, my
company has thousands of Maven modules that are all on snapshots with an
updatePolicy of "always" (basically everyone gets the latest build of
everything, kind of like a poor man's monorepo). This mostly works fine,
but one of
I’d really like if somebody could draft the release notes for the two
different classloader changes. It will make it easier to decide whether to
roll with this or bump minor with (if I recall correctly) the corresponding
drop of Java 7 per our policy on JVMs with minor version bump
On Wed 11 Oct
+1
eventually adding a flag or system property to activate the new behaviour
this remembers me of the idea regarding flags to support new beta features
Regards,
Hervé
Le mercredi 11 octobre 2017, 08:13:13 CEST Anders Hammar a écrit :
> I'm feeling stronger and stronger about not having this
I'm feeling stronger and stronger about not having this change in a bugfix
release. Why not go for v3.6.0 if we decide that the class loading change
is the way to go? And keep bugfix releases for just bugfixes.
/Anders
On Wed, Oct 11, 2017 at 3:43 AM, Hervé BOUTEMY
wrote:
13 matches
Mail list logo