Thanks for working on this. Today is my last day before Christmas vacation
so I might not be able to test the staged version when the time comes. Happy
voting, and hopefully we'll be able to upgrade in 2011 :)
--
- Jan Fredrik Wedén
On Wed, Dec 22, 2010 at 10:23, Kristian Rosenvold <
kristian.ro
Mockito problem should be fixed, issue see
http://jira.codehaus.org/browse/SUREFIRE-673
There is a published 2.7.1-SNAPSHOT that you can use to re-test,
attach feedback to issue.
2.7.1 is fairly close, I'll just pack in one more session of fixing
low-hanging issues from the backlog. Vote will sta
It seems like this issue + the mockito problem
(http://jira.codehaus.org/browse/SUREFIRE-673) will make 2.7.1 happen
sooner rather than later. Sticking to your previous favourite version
is always an option.
-
To unsubscribe, e-ma
Hi,
Sorry for not taking the time to test this while the vote was in progress.
However, I think I've found a regression causing the failsafe plugin v2.7 to
report builds with broken integration tests as successful. I've created [1]
already - hope I got my reasoning correct in the description. Just
Mark,
I now have a second report of this and a test-project (via twitter ;).
I will track the problem from there.
I have been planning to do a 2.7.1 maintenance release in 2-4 weeks
time, so whatever it is will be fixed for 2.7.1
Kristian
> Looks like I had a rogue debugger process left around
Looks like I had a rogue debugger process left around, the mockito
issue is still haunting me tho.
This is on a line doing:
doReturn(something).when(spiedObject).doSomething();
I'll create an example project and see if I can that to break.
--
"Great artists are extremely selfish and arrogant
I was able to build Archiva with 2.7, which uses Mockito (no parallel /
debugging bits though)
- Brett
On 20/12/2010, at 6:23 PM, Kristian Rosenvold wrote:
> Technically we do not have any parallelism that works between forks,
> it's all inside each fork. Do keep quiet about that though, or som
Technically we do not have any parallelism that works between forks,
it's all inside each fork. Do keep quiet about that though, or someone
will start nagging for parallel running for forkMode=always :)
As for mockito, I thought one of my test projects was using that, but it
turned out they had g
Are you running tests in parallel... Attaching a debugger with tests in
parallel and fork mode always would not make sense if you were doing that
:-)
- Stephen
---
Sent from my Android phone, so random spelling mistakes, random nonsense
words and other nonsense are a direct result of using swype
Works fine here, in both "once" and "always" (tested on linux/windows).
Is port 5005 free ? On occasions I've seen the debugger hang at the OS
level that require all jvm's to be terminated before it unlocks (maybe
even a boot on windows).
But please do try to find out what the problem is, esp. the
Doh.
So I thought I'd run it thru the debugger and see whats going on
> mvn test -Dmaven.surefire.debug
>
>
> ~/IdeaProjects/secure
Hrm,
I'm seeing some tests fail under 2.7 that use Mockito that pass fine
when using 2.6. This is only in one of our modules and not uniform
across other things that use mockito. Will investigate further.
--
"Great artists are extremely selfish and arrogant things" — Steven
Wilson, Porcupine
Hi,
The vote has passed with the following result :
+1 (binding): Lukas Theussl, Benjamin Bentmann, Hervé Boutemy, Stephen
Connolly, Olivier Lamy, Vincent Siveton, Kristian Rosenvold
+1 (non binding): Tony Chemit, Jeff Jensen, Robert Scholte
I will promote the artifacts to the central repo.
+1 (non-binding)
-Robert
> From: ol...@apache.org
> Date: Sun, 19 Dec 2010 09:39:58 +0100
> Subject: Re: [VOTE] Release Maven Surefire Plugin version 2.7, Take 2
> To: dev@maven.apache.org
>
> +1
>
> 2010/12/16 Kristian Rosenvold :
> > Highlights of the rele
+1
2010/12/16 Kristian Rosenvold :
> Highlights of the release:
> * Pluggable providers are now supported, as well as manual provider
> selection
> * 2.7 is significantly smaller and lighter than earlier 2.x versions
> (Only 30% of the download size)
> * A large chunk of the highest voted issues h
+1
- Stephen
---
Sent from my Android phone, so random spelling mistakes, random nonsense
words and other nonsense are a direct result of using swype to type on the
screen
On 16 Dec 2010 10:47, "Kristian Rosenvold"
wrote:
> Highlights of the release:
> * Pluggable providers are now supported, as
+1
Hervé
Le jeudi 16 décembre 2010, Kristian Rosenvold a écrit :
> Highlights of the release:
> * Pluggable providers are now supported, as well as manual provider
> selection
> * 2.7 is significantly smaller and lighter than earlier 2.x versions
> (Only 30% of the download size)
> * A large chun
Kristian Rosenvold wrote:
Staging repo:
https://repository.apache.org/content/repositories/maven-026/
+1
Benjamin
-
To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
For additional commands, e-mail: dev-h...@maven.ap
+1
-Lukas
Kristian Rosenvold wrote:
Highlights of the release:
* Pluggable providers are now supported, as well as manual provider
selection
* 2.7 is significantly smaller and lighter than earlier 2.x versions
(Only 30% of the download size)
* A large chunk of the highest voted issues have be
+1 (non-binding)
This one works great too!
On Thu, Dec 16, 2010 at 5:38 AM, chemit wrote:
> On Thu, 16 Dec 2010 11:46:40 +0100
> Kristian Rosenvold wrote:
>
> +1 (non binding)
>
>
>
> --
> Tony Chemit
>
> tél: +33 (0) 2 40 50 29 28
> email: che...@codelutin.com
> http://ww
On Thu, 16 Dec 2010 11:46:40 +0100
Kristian Rosenvold wrote:
+1 (non binding)
--
Tony Chemit
tél: +33 (0) 2 40 50 29 28
email: che...@codelutin.com
http://www.codelutin.com
-
To unsubscribe, e-mail: dev-
Cool, I'll take a look.
Does this mean we can have -multiple- providers at once? i.e.
previously if you had TestNG on the class path you couldn't also have
JUnit4 tests.
( I'm thinking of writing a clojure-test provider to go along with the
clojure-maven-plugin ).
--
"Great artists are extreme
From the staged 2.7 docs:
http://maven.apache.org/plugins/maven-surefire-plugin-2.7/api.html
http://maven.apache.org/plugins/maven-surefire-plugin-2.7/examples/providers.html
The proposed api's can be seen in:
http://maven.apache.org/surefire/staging/surefire-api/apidocs/index.html
As can be se
Highlights of the release:
* Pluggable providers are now supported, as well as manual provider
selection
* 2.7 is significantly smaller and lighter than earlier 2.x versions
(Only 30% of the download size)
* A large chunk of the highest voted issues have been fixed.
* Some performance optimization
Kristian,
Where can I read about writing a pluggable provider? If it's what I
assume then gimme gimme gimme! :-)
Mark
--
"Great artists are extremely selfish and arrogant things" — Steven
Wilson, Porcupine Tree
On Wed, Dec 15, 2010 at 12:21 PM, Kristian Rosenvold
wrote:
> * Pluggable provid
I cancel this vote. Will be respinning shortly for SUREFIRE-665 and a
jdk1.3 (!) fix.
Kristian
Den 16.12.2010 08:44, skrev Lukas Theussl:
+1
-Lukas
Kristian Rosenvold wrote:
Highlights of the release:
* Pluggable providers are now supported, as well as manual provider
selection
* 2.7 is s
+1
-Lukas
Kristian Rosenvold wrote:
Highlights of the release:
* Pluggable providers are now supported, as well as manual provider
selection
* 2.7 is significantly smaller and lighter than earlier 2.x versions
(Only 30% of the download size)
* A large chunk of the highest voted issues have bee
Which clearly explains why, if we reduced the logging output level
from DEBUG to WARN, the OOM did not occur with smaller values of -Xmx.
That was baffling at the time... Wonderful fix. Thanks! :-) (this
problem was the worst for release via parent pom; not happy to
-DskipTests on release or i
Sorry for chatting on the vote thread here, but fixing the TeeStream
memory leak probably also has a noticeable performance impact for users
with large numbers of tests and/or lots of console output;
It turned out for test number N, any stdout/stderr output would be
duplicated into N tee-streams,
+1 (non-binding)
We can finally run a full build with tests from the parent with this
version (would OOM unless -Xmx was set to nearly 2G!). :-)
On Wed, Dec 15, 2010 at 2:01 AM, chemit wrote:
> On Wed, 15 Dec 2010 00:21:49 +0100
> Kristian Rosenvold wrote:
>
> +1 (none binding)
>
> Works fine
PEBKAC of the slightly embarrasing sort on my end.
So the vote is on; happy testing!
Kristian
-
To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
For additional commands, e-mail: dev-h...@maven.apache.org
I seem to have run across at least one case where failsafe does not
properly fail a build with test failures. I'm looking into the problem.
Although I'm not cancelling the vote right now, you may consider
postponing further testing until I have determined the exact
circumstances. It smells like a
On Wed, 15 Dec 2010 00:21:49 +0100
Kristian Rosenvold wrote:
+1 (none binding)
Works fine on our project.
Thanks!
> Highlights of the release:
> * Pluggable providers are now supported, as well as manual provider
> selection
> * 2.7 is significantly smaller and lighter than earlier 2.x version
Highlights of the release:
* Pluggable providers are now supported, as well as manual provider
selection
* 2.7 is significantly smaller and lighter than earlier 2.x versions
(Only 30% of the download size)
* A large chunk of the highest voted issues have been fixed.
* Some performance optimizations
34 matches
Mail list logo
