I read Michaels message the way like option 6:
> Since we have only two files: installation and user, I'd expect
that first the installation settings are parsed followed by user
settings (generic to specific).
Installation is more generic (global) than user.
Same for git: it will first parse the
Hello,
I am not able to dedicate the time to do bisecting this issue
-D
On Mon, Aug 9, 2021 at 3:03 AM Eric Lilja wrote:
> It's an impressive list of fixes and we are many that would like to see a
> 3.8.2 release appear. Question for me is if the particular change
> that appears to cause a ge
Looking only at the log, both writers existed before it was promoted from the
sandbox in 2007.
Counter was introduced by Karl Heinz while refactoring.
It could very well be there's a better solution for it right now.
I would expect enough coverage by ITs to confirm that.
Robert
On 9-8-2021 16:35:
When I see "global" I think scope (applies to all users) not precedence
(applies first).
I'm happy with option 5 as it stands. It allows configuration to be set
that can both overide or be overridden by the project, without changing
system files.
Delany
On Mon, 9 Aug 2021, 15:09 Benjamin Marwell,
I'll add it, but it doesn't match the order of GIT[1] as mentioned by Michael.
Maybe there should be a list of other comparable tools with their order
Robert
[1] https://git-scm.com/docs/git-config#FILES
On 9-8-2021 15:09:30, Benjamin Marwell wrote:
Hi Robert,
I would like to propose another
Howdy,
re maven-shade-plugin's Java package org.apache.maven.plugins.shade.pom:
1. What is the goal of this code? As there are 3 classes in here, and
Counter is used only by MavenJDOMWriter, while MavenJDOMWriter is used only
by PomWriter, while PomWriter is used only by ShadeMojo (to write out d
Hi Robert,
I would like to propose another solution.
According to the wiki page and to the documentation, we have
user < installation (aka global) < project.
where "<" means "is overridden by".
>From what I understand from other build systems, a more common approach would
>be
installation (ak
It's an impressive list of fixes and we are many that would like to see a
3.8.2 release appear. Question for me is if the particular change
that appears to cause a general performance degradation on the scale of ten
percent is worth it (if it indeed can be isolated to particular JIRA and
not due to
+1
Tested on several projects with no issue
I agree that the performance is a bit degraded but nothing blocker compared
the value of the fixes introduced here.
On Sun, Aug 8, 2021 at 10:53 PM Sylwester Lachiewicz
wrote:
> +1
>
> niedz., 8 sie 2021, 21:32 użytkownik Robert Scholte
> napisał:
>
>