hey,
Artifactory (http://www.jfrog.org/products.php) has the possibility to copy an
artifact to another repository - about the if and how to use it, there are
smarter people here already discussing it
cheers
-Original Message-
From: Phillip Hellewell ssh...@gmail.com
To:
Well, at this point all I'm mainly after is the ability to copy anartifact
from snapshot
to release repo (removing -SNAPSHOT from theversion) without having to go
through
the rebuild process.
If thatfunctionality does not exist as a free plugin or anything at
themoment,
who knows,
thanks a lot, it works
regards
On Wed, Oct 6, 2010 at 5:03 PM, Stephen Connolly
stephen.alan.conno...@gmail.com wrote:
you need to add -DoldVersion=1.0
On 6 October 2010 15:24, Jo Support jo.supp...@gmail.com wrote:
I'm a bit puzzled by versions plugin. According to versions:set goal's
Hi,
I sent this request to repo-maintain...@maven.apache.org the other day (as
described below), but I haven't received a response yet.
Any ideas? This is currently holding up our 1.5.2 release, since Maven bundle
generation is part of our release process.
Thanks,
Greg Brown
Begin forwarded
Hi,
Is logging POM operations so difficult as described in
http://maven.apache.org/plugins/maven-antrun-plugin
So that every time yoy want to trace execution with ant echo command you
have to
declare thirty lines of XML clutter ?
Is it possible to code those xml lines only once and refer then
Bingo. I was just about to reply to say that. You generally never go from a
SNAPSHOT to release as the way the artifacts are generated are very different.
Not only are the names of the artifacts themselves different but the metadata
in the repository is also quite different. And I am sure there
On Thu, Oct 7, 2010 at 8:29 AM, Sakari Isoniemi
sakari.isoni...@gmail.com wrote:
Is logging POM operations so difficult as described in
http://maven.apache.org/plugins/maven-antrun-plugin
So that every time yoy want to trace execution with ant echo command you
have to declare thirty lines of
Yeah, I'll just keep it pretty simple like this, however it will be
even simpler because our build process will append the SVN tag name,
so we won't ever need to manually add like 01, 02, 03, etc.
What are you using as the SVN tag name? The release plugin will create the tag
for you from the
Is anyone using the Selenium plugin?
I'd like to use the selenese goal to run html test suites; but haven't
had any success getting it to work.
The plugin documentation recommends using JUnit to run Selenium tests
with the normal surefire plugin. In that case, I'm not sure of the
point of the
I'm looking to rename a couple of files before the actual WAR package
is created. For example, if I have a Javascript file named my.js, I
want to rename it to my-20100106.js then do a search and replace on
all references to my.js and have them reference the new file.
I would love to use the
2010/10/7 Greg Akins angryg...@gmail.com:
Is anyone using the Selenium plugin?
I'd like to use the selenese goal to run html test suites; but haven't
had any success getting it to work.
Tiles test webapp is connected to Selenium HTML Selenese tests.
It starts the container (Tomcat), runs the
Yes. We use it very extensively. Works great. I found the docs to be fairly
adequate. We write all our Selenium tests using junit.
Make sure you get the Selenium IDE plugin for firefox. It will greatly help you
write your junit tests.
The Maven plugin for Selenium start the selenium-rc (I
Hi Lukas,
On Thu, Oct 7, 2010 at 8:59 AM, Lukas Bradley
lu...@thrustinteractive.com wrote:
I would love to use the maven-antrun-plugin, but it doesn't look like
there is a stage where the war plugin would cede control, or allow
something to pre-process the files that have been copied to the
Thanks Todd and Antonio
I've used Selenium tests before, so I know the basic mechanism
Just not clear how the plugin fits in. I'll look at the Tiles tests
to get a better idea.
On Thu, Oct 7, 2010 at 9:20 AM, Thiessen, Todd (Todd)
tthies...@avaya.com wrote:
Yes. We use it very extensively.
Hello,
war:war is called during package phase, so at prepare-package phase you have
nothing available. I think the war-plugin should be updated to take that
into account.
A possible workaround is to bind an execution of war:exploded during the
prepare-phase, then you do your stuff with antrun,
2010/10/7 Greg Akins angryg...@gmail.com:
Just not clear how the plugin fits in. I'll look at the Tiles tests
to get a better idea.
Just to be clear, Tiles test webapp uses the selenese goal that runs
HTML Selenese tests, as you wanted.
Notice that JUnit suggestion is only valid if you are
I know I can use properties like ${project.build.finalName} to get some
information from the POM. But the information I want to get is the final name
of the artifact built by a dependency. So, for example suppose I have a
dependency like this:
dependencies
dependency
On Thu, Oct 7, 2010 at 9:37 AM, Antonio Petrelli
antonio.petre...@gmail.com wrote:
Just to be clear, Tiles test webapp uses the selenese goal that runs
HTML Selenese tests, as you wanted.
Notice that JUnit suggestion is only valid if you are using pure Java
commands, using Selenium RC.
In
2010/10/7 Greg Akins angryg...@gmail.com:
On Thu, Oct 7, 2010 at 9:37 AM, Antonio Petrelli
antonio.petre...@gmail.com wrote:
Just to be clear, Tiles test webapp uses the selenese goal that runs
HTML Selenese tests, as you wanted.
Notice that JUnit suggestion is only valid if you are using
Is is possible to use properties to get the final name of an artifact produced
by a dependency?
Not that I'm aware of. Unless you're using properties in the
dependency declaration, which you could reuse somewhere else in the
pom.
How do I do this?
Why do you need to do this?
Wayne
The problem is now fixed and it was user-error. Thanks to everyone for
their suggestions. The discussion about poms in the repository led me to
the answer. I noticed the pom for one of my dependent jars was basically
blank and had been generated by nexus. It was a minimal POM that included
no
On Thu, Oct 7, 2010 at 6:51 AM, Thiessen, Todd (Todd)
tthies...@avaya.com wrote:
Bingo. I was just about to reply to say that. You generally never go from a
SNAPSHOT to release as the way the artifacts are generated are very
different. Not only are the names of the artifacts themselves
But the main thing I'm thinking about is the fact that the package
hasn't changed. So why should I have to rebuild the code?
But it has changed. Very much so. A SNAPSHOT is not a release.
That not
only wastes time but more importantly it adds risks that the build may
not be identical to
On Thu, Oct 7, 2010 at 7:06 AM, Thiessen, Todd (Todd)
tthies...@avaya.com wrote:
Yeah, I'll just keep it pretty simple like this, however it will be
even simpler because our build process will append the SVN tag name,
so we won't ever need to manually add like 01, 02, 03, etc.
What are you
So this is one aspect where our setup will differ slightly. We don't
need the release plugin to create tags for us because our build
scripts on our build machines are already set up to create tags with
every single build (even snapshots builds).
I think we may be getting to crux of your.
On Thu, Oct 7, 2010 at 10:43 AM, Thiessen, Todd (Todd)
tthies...@avaya.com wrote:
But the main thing I'm thinking about is the fact that the package
hasn't changed. So why should I have to rebuild the code?
But it has changed. Very much so. A SNAPSHOT is not a release.
Why would the package
On Thu, Oct 7, 2010 at 11:08 AM, Thiessen, Todd (Todd)
tthies...@avaya.com wrote:
So this is one aspect where our setup will differ slightly. We don't
need the release plugin to create tags for us because our build
scripts on our build machines are already set up to create tags with
every
somethign that occurrs to me...
Take a step back and look at your processes... what determines when a build
goes to QA? Look for a way to have that trigger a mvn release which will
push the build to your release repo
On Thu, Oct 7, 2010 at 10:26 AM, Phillip Hellewell ssh...@gmail.com wrote:
On
Why would the package change? Were' using the assembly plugin to
create a zip of the binaries (dlls, libs, and headers). Why should
that change if the code it's built from hasn't changed?
The names of the artifacts would change. The meta data used to retrieve the
artifacts is also
This conversation turns on a classic dilemma in release management.
1. The important thing is to ship exactly the bits that have been
tested. Prepare a package, give it to QA. If they like it, ship it. If
they don't, rinse, lather, and repeat.
2. The important thing is to never, ever, ever
Hi all,
I'm pretty sure I must be configuring something wrong in my poms, but the
plugin update reports is telling me I don't have the latest versions when
reporting against a pom that's pulling plugins from it's parent. For
example:
*Parent Pom*
*
*
build
plugins
Hi
I implemented using the following:
In the parent pom:
properties
envTypelocal/envType !-- Default Environment value --
envProps../autotestProperties//envProps
/properties
build
resources
resource
directory${envProps}/resources/directory
On Thu, Oct 7, 2010 at 11:34 AM, Benson Margulies bimargul...@gmail.com wrote:
This conversation turns on a classic dilemma in release management.
1. The important thing is to ship exactly the bits that have been
tested. Prepare a package, give it to QA. If they like it, ship it. If
they
Thanks, I'll take the ., maven. and .version into consideration
for our pom, but that is unrelated to the issue at hand. Either way it
is done, the site 2.1.1 plug-in does not execute the site for the
modules, only for the parent projects site directory. When I get a
chance, I will put together a
Hi,
You can have a look at http://jira.codehaus.org/browse/MWAR-238.
I'm just an it test before pushing the patch.
So you can test it too :-)
2010/10/7 Lukas Bradley lu...@thrustinteractive.com:
I'm looking to rename a couple of files before the actual WAR package
is created. For example, if
I want to use the enforcer plugin to disallow deployment if a certain
property is not set to a certain value.
If I bind to the deploy phase it fails _after_ the actual deploy goal,
so that doesn't work.
If I bind to the install phase it fails after the install, which works
ok but you see an error
Hi Phillip,
On Thu, Oct 7, 2010 at 3:57 PM, Phillip Hellewell ssh...@gmail.com wrote:
What is the general approach to solving this conundrum? If there were
a post-install or a pre-deploy phase, that would solve it, but
those don't exist afaik :(
Please describe why the verify phase is not
On Oct 7, 2010, at 3:57 PM, Phillip Hellewell wrote:
What is the general approach to solving this conundrum? If there were
a post-install or a pre-deploy phase, that would solve it, but
those don't exist afaik :(
You can create a custom lifecycle for that build, then bind where you want.
On 2010-10-07 21:40, Andrew Robinson wrote:
Thanks, I'll take the ., maven. and .version into consideration
for our pom, but that is unrelated to the issue at hand. Either way it
is done, the site 2.1.1 plug-in does not execute the site for the
modules, only for the parent projects site
I have a dumb question. If I unpack dependencies to a deps subfolder,
should mvn clean be cleaning those up automatically? Or should I
attach a goal to the clean phase to do it? Or should I not be
worrying about this at all because unpack-dependencies already knows
how to replace files when
they will be being extracted into a sub-folder of ${basedir}/target if
you are following the Maven way... given some of your mail threads I
will not assume that you are following the Maven way ;-)
If you follow the Maven way then clean will tidy up as by default
clean removes the target folder
Can anyone recommend whether to SCM commit the generated MANIFEST.MF
files? Or should they just be ignored?
-
To unsubscribe, e-mail: users-unsubscr...@maven.apache.org
For additional commands, e-mail: users-h...@maven.apache.org
In general, you shouldn't commit any generated file to SCM.
On Oct 7, 2010, at 8:31 PM, Paul Benedict pbened...@apache.org wrote:
Can anyone recommend whether to SCM commit the generated MANIFEST.MF
files? Or should they just be ignored?
Aren't they generated into target usually, and rebuilt on each build? Doesn't
sound like something to check in.
On 08/10/2010, at 11:31 AM, Paul Benedict wrote:
Can anyone recommend whether to SCM commit the generated MANIFEST.MF
files? Or should they just be ignored?
44 matches
Mail list logo
