I can only second that idea. I am using it for Android applications built
with Maven and blogged a few tips for setting it up recently. That might
help for normal stuff as well (e.g. keeping the map file if you
obfuscate..)
See more
http://www.simpligility.com/2010/12/proguard-for-android-with-ma
http://pyx4me.com/pyx4me-maven-plugins/proguard-maven-plugin/.
Obfuscate the jars that go into the war or obfuscate the classes
directly before they are packaged into the war. The plugin can do
both. Deploy the zip to your repository as well as the war.
Kalle
On Fri, Jan 28, 2011 at 4:23 PM, Kar
lukewpatterson wrote:
>
> In this thread, things that seem to be in agreement so far:
>
> * Maven 2 and 3 have different interpolation behavior, and the behavior
> isn't documented in the compatibility notes
> * dynamic variable interpolation doesn't affect determinism, i.e. same
> input yields
Hi,
Part of my job is to maintain an Ant build to obfuscate a web app. This
happens in four steps:
- Ant builds the WAR file on PC
- WAR file gets uploaded to a server where it is obfuscated (which - due to
class and method renaming - also changes configuration files and accesses
the libraries)
-
On 28 January 2011 15:22, lukewpatterson wrote:
> * Maven 2 and 3 have different interpolation behavior, and the behavior
> isn't documented in the compatibility notes
> * dynamic variable interpolation doesn't affect determinism, i.e. same input
> yields same output
> * whether the ability furthe
In this thread, things that seem to be in agreement so far:
* Maven 2 and 3 have different interpolation behavior, and the behavior
isn't documented in the compatibility notes
* dynamic variable interpolation doesn't affect determinism, i.e. same input
yields same output
* whether the ability fur
Hi,
We saw a behavior where both Maven 2 and 3 don't bring a file larger than
2Gb from repository and truncate it to 2147483647 bytes, which is exactly
Integer.MAX_VALUE. An issue was opened
(http://jira.codehaus.org/browse/MNG-4977) but no responses so far.
I wonder if it happens to anybody els
Hilco Wijbenga-3 wrote:
>
> Why would this make it less repeatable? A build's
> repeatability/reproducibility is based on the whether it is
> deterministic. And whether a property can be overwritten/changed
> doesn't change the determinism of a process. If it did, then (single
> threaded) progra
I ended up checking file sizes as you suggested.
Thanks,
Gili
On 26/01/2011 4:04 PM, stephenconnolly [via Maven] wrote:
> up to you, but if I was downloading large stuff from non maven
> repositories,
> I'd add checks for interrupted downloads.
>
> - Stephen
>
> ---
> Sent from my Android
Hi Brian
You can add wagon-ssh libary to the M2_HOME/lib directory. I didn't
test this configuration but it should work.
Marcin
On 27 January 2011 00:53, Brian Ferris wrote:
> I discovered that with the upgrade to Maven 3, the ssh wagon is no longer
> included by default, which became an issue
On 28 January 2011 11:10, Evgeny Goldin wrote:
> lukewpatterson wrote:
>> The general argument against dynamic properties is that they make the
>> build configuration less declarative, right?
>
> It makes builds more dynamic and less repeatable and that could be the
> reason Maven 3 was changed to
lukewpatterson wrote:
>
> Yeah, it would be nice to have some middle ground for allowing plugins to
> help configure other plugins. Right now the "Maven Way" is to use the
> file system, right?
>
That's the main problem - the "Maven Way" is different in Maven 2 and 3 so I
still don't know if
Evgeny Goldin wrote:
>
> I think this issue opens a wider one: does Maven treat properties like a
> mutable variables ...
>
Yeah, it would be nice to have some middle ground for allowing plugins to
help configure other plugins. Right now the "Maven Way" is to use the file
system, right? The
Hi,
just try to add an dependency to plexus-utils 2.0.5 ...
I'm not sure if it will solve the problem...
Kind regards
Karl Heinz Marbaise
--
SoftwareEntwicklung Beratung SchulungTel.: +49 (0) 2405 / 415 893
Dipl.Ing.(FH) Karl Heinz MarbaiseICQ#: 135949029
Hauptstrasse 177
And yes, as a workaround I need not to define anything as , in
this case it works as expected. The problem with this approach is that POM
becomes very fragile, like in old Ant days: you forget the rule and add a
property that is overridden later .. Whops, but it's not.
I think this issue opens a
Thanks you, Marc. I hope dynamic behavior will be chosen here, Ant behavior
was known to cause a lot of pain and weird errors that were fixed by
Ant-Contrib task
(http://ant-contrib.sourceforge.net/tasks/tasks/variable_task.html),
defining properties dynamically at any time. Used that *a lot*.
C
Hi,
it looks like a conflict in the dependencies - one of the dependencies
(plexus-utils?) seems to be to old.
regards,
Oli
Am 28.01.2011 um 16:34 schrieb mremerson...@aim.com:
> Hi,
>
> I like to use the maven-invoker plugin in my own plugin, but it failes with
> an error...
> i use maven
Hi
My multi module project is structured with separate aggregate pom and
parent pom (aggregate pom is not a parent pom).
All modules are aggregated by aggragate pom (parent pom is also
submodule of aggregate pom). All submodules inherit from parent pom.
Aggregate pom:
http://code.google.com/p/m4e
Hi,
I like to use the maven-invoker plugin in my own plugin, but it failes with an
error...
i use maven 2.2.1, in the pom.xml i have
org.apache.maven.shared
maven-invoker
2.0.11
and the error is:
[INFO] [idl-builder:build {execution: default-cli}]
[FATAL ERROR] de.idlbuilder.IdlB
Hi, how can I tell the surefire plugin that I want to use a custom
JUnitResultFormatter? In Ant you would do it like this:
--
View this message in context:
http://old.nabble.com/using-custom-JUnitResultFormatter-tp30787256p30787256.html
Sent from the Surefire - Users mailing list archive a
I am trying to use the org.ops4j:maven-inherit-plugin:1.1, which has always
worked great for me in Maven2.
But now I am using Maven3, and I get this error:
[ERROR] Failed to execute goal org.ops4j:maven-inherit-plugin:1.1:inherit
(default) on project helios-product-package-plugin: assembly plugin
I'd say properties should be static - otherwise predefining them using
the command line switch '-D' wouldn't work properly. BTW: With this
Maven 3 show the same behaviour like Ant (I know, I shouldn't compare
them), where properties are always immutable.
I just checked it with the the Buildnum
22 matches
Mail list logo
