logging-log4j-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Remko Popma <remko.po...@gmail.com>
Subject Re: [VOTE] Log4j 2.0.1 candidate 1
Date Thu, 31 Jul 2014 23:52:54 GMT
No, CachedClock wasn't changed.  

I suppose the test is too strict in that it may fail on slower systems or systems that are
busy doing other things at the same time. 

I should just remove that test. Gary has mentioned it failing on his machine several times
in the past, I think. It is unfortunately not easy to test Clocks otherwise, but this has
been an annoyance for too long now. I'll get rid of it. 

Gary, with that in mind, can I ask you to reconsider your vote? 

Sent from my iPhone

> On 2014/08/01, at 8:06, Matt Sicker <boards@gmail.com> wrote:
> Wasn't the CachedClock stuff changed recently? Did that make it into this release?
>> On 31 July 2014 17:51, Gary Gregory <garydgregory@gmail.com> wrote:
>> -0:
>> I downloaded the src zip and ran "mvn clean test" and got:
>> Failed tests:
>>   CachedClockTest.testAfterWaitStillLessThan17Millis:33 diff too large: 43
>> Tests run: 684, Failures: 1, Errors: 0, Skipped: 28
>> [INFO] ------------------------------------------------------------------------
>> [INFO] Reactor Summary:
>> [INFO]
>> [INFO] Apache Log4j 2 ..................................... SUCCESS [  3.427 s]
>> [INFO] Apache Log4j API ................................... SUCCESS [ 38.578 s]
>> [INFO] Apache Log4j Core .................................. FAILURE [06:41 min]
>> Using:
>> Apache Maven 3.2.2 (45f7c06d68e745d05611f7fd14efb6594181933e; 2014-06-17T09:51:42-04:00)
>> Maven home: C:\Java\apache-maven-3.2.2
>> Java version: 1.7.0_60, vendor: Oracle Corporation
>> Java home: C:\Program Files\Java\jdk1.7.0_60\jre
>> Default locale: en_US, platform encoding: Cp1252
>> OS name: "windows 7", version: "6.1", arch: "amd64", family: "windows"
>> This seems like a test too strict or incorrect assumptions about clock granularity
on Windows.
>> Worse, is that it does not happen every time I run Maven.
>> Ideally, I think all of the issues Remko found should be fixed and another RC cut,
but they do not feel like "showstoppers", and I am not doing the RM work. Thank you Matt for
doing the work.
>> The version thing is a bit confusing and I've rewritten this section of my email
several times before settling on proposing to use the exact version number consistently: 2.0.1
for this release. Otherwise, talking about "Log4j" is fine, I do not think we need to talk
about "Log4j 2" as a general rule. In the migration doc we can talk about "Log4j 1" and "Log4j
2" since context is important. But I can also see that some additional migration docs might
only apply to later versions of Log4j where using a specific version will be important.
>> The perf module prints out screen after screen of warnings like:
>> [WARNING] /C:/temp/rc/apache-log4j-2.0.1-src/log4j-perf/target/generated-sources/annotations/org/apache/logging/log4j/perf/jmh/generated/FormatterLoggerBenchmark_latencyBaseline.java:[53,33]
found raw type: org.openjdk.jmh.logic.results.Re
>> ult
>>   missing type arguments for generic class org.openjdk.jmh.logic.results.Result<T>
>> What is that all about?
>> Gary
>>> On Wed, Jul 30, 2014 at 10:57 AM, Remko Popma <remko.popma@gmail.com> wrote:
>>> +1 Artifacts look good, site looks good.
>>> Minor issues found (no showstoppers IMO):
>>> Artifacts
>>> ---------
>>> log4j-distribution-2.0.1-bin.zip/apache-log4j-2.0.1-bin/RELEASE-NOTES.txt
>>> "This is the first GA release, after thirteen prior releases over the last 4
>>> ?
>>> Changes in this release..."
>>> -> Should be "...the second GA release..."
>>> -> The question mark "?" should probably be "Bug fixes and enhancements"
>>> Site
>>> ----
>>> Download page: "Apache Log4j 2 is signed by Ralph Goers B3D8E1BA"
>>> -> Is this correct for 2.0.1?
>>> Project Info > Dependency Convergence: logj2-perf uses disruptor-3.2.1, but
log4j-core uses disruptor-3.3.0.
>>> (Not an issue since log4j-perf is not a distribution artifact.)
>>> Javadoc: many package docs mention "Log4j 2.0". E.g.
>>> http://people.apache.org/~mattsicker/log4j/2.0.1/log4j-api/apidocs/index.html
>>> Sometime in the future we will release version 2.1, 2.2, etc.
>>> Should we update all Javadoc pages to "Log4j 2" to future-proof them?
>>> Similarly, the FAQ page mentions "the Log4j 2.0 implementation" and "the Log4j
2.0 API".
>>> We should probably change this to "the Log4j 2 implementation" and "the Log4j
2 API".
>>> Manual > Introduction (index.html)
>>> "The latest log4j version, including full-source code, class files and 
>>>  documentation can be found at http://logging.apache.org/log4j/2.0/index.html."
>>> -> We should change that link to http://logging.apache.org/log4j/2.x/index.html
>>>    (It looks like 2.0 and 2.x both work as URLs.)
>>> Manual > Log4j 1.x Migration also has a few mentions of "2.0".
>>> Manual > Configuration page has one mention of "2.0".
>>> Manual > JMX has six places where log4j-api-2.0.jar etc are mentioned. 
>>> 		-> Can we make these variables to they change with the ${project.version}?
>>> Manual > Custom Log Levels has two places where log4j-core.2.0.jar is mentioned.
>>> 		-> Can we make these variables to they change with the ${project.version}?
>>> Manual > Layouts > PatternLayout, %d pattern examples:
>>> %d{HH:mm:ss}{GMT+0}	18:34:02,781 <- this does not look right? The result has
millis instead of timezone...
>>> Manual > Architecture > LoggerConfig > Log Levels: (we may want to rewrite
the text below)
>>> The set of possible levels includes TRACE, DEBUG, INFO, WARN, ERROR, and FATAL.
>>> Note that in Log4j 2, the Level is an Enum and cannot be sub-classed.
>>> Users who desire more granularity are encouraged to use Markers instead.
>>> Manual > Extending Log4j > ContextSelector: missing AsyncLoggerContextSelector
>>>> On Wed, Jul 30, 2014 at 10:33 AM, Matt Sicker <boards@gmail.com> wrote:
>>>> This is a vote to release Log4j 2.0.1, the first bug fix release of the 2.0
branch of Log4j 2.
>>>> Please test and cast your votes.
>>>> [] +1, release the artifacts
>>>> [] -1, don't release because...
>>>> The vote will remain open for 72 hours (or more if required). Note that binding
votes, as usual, are done by PMC members. Non-binding votes are welcome from anyone and are
always welcome.
>>>> This is my first time acting as an RM, so if there are any issues with this,
please let me know as soon as possible! If you don't already have my GPG key, it's key ID
FA1C814D which is available on all the public GPG servers. Otherwise, you can download it
from my Apache site at: <http://people.apache.org/~mattsicker/FA1C814D.asc>
>>>> New features:
>>>> o LOG4J2-725:  Added WebLoggerContextUtils class to log4j-web for helper
methods useful for asynchronous servlets. 
>>>> o LOG4J2-710:  Added documentation for Custom Levels and Custom Loggers.

>>>> Fixed Bugs:
>>>> o LOG4J2-744:  Avoid unnecessary Clock calls when TimestampMessage is logged.
Thanks to Scott Harrington. 
>>>> o LOG4J2-704:  Improved error message if configuration file not found. 
>>>> o LOG4J2-750:  Webapp configuration page has incorrect class name. Thanks
to Mike Calmus. 
>>>> o LOG4J2-749:  Retain the default date pattern after fixing the ISO8601 pattern.
Thanks to Scott Harrington. 
>>>> o LOG4J2-670:  DatePatternConverter ISO8601_PATTERN now conforms to ISO8601.

>>>> o LOG4J2-741:  Reinstate the package configuration attribute for discovering
custom plugins. 
>>>> o LOG4J2-742:  XInclude not working with relative path. Thanks to Pascal
>>>> o LOG4J2-740:  Fixed typo in webapp manual regarding sample web.xml file.
Thanks to Kosta Krauth. 
>>>> o LOG4J2-738:  RollingFileManager deadlock if async action thread fails to
start. Thanks to Timothy Stack. 
>>>> o LOG4J2-736:  Fixed log4j-bom so that it won't specify a default scope on
any third party dependencies. 
>>>> o LOG4J2-735:  Fixed log4j-bom so that it won't interfere with spring-bom
and others. 
>>>> o LOG4J2-731:  Updated documentation regarding extensions to LoggerContextFactory
and Log4j 2 providers. 
>>>> o LOG4J2-373:  Fixed ClassLoader issues in loading Log4j providers in an
OSGi environment. 
>>>> o LOG4J2-719:  Correctly handle NetworkOnMainThreadException thrown on Android
during Log4j2 initialization. 
>>>> o LOG4J2-716:  Automatically disable log4j JMX when detecting we are running
on Android. 
>>>> o LOG4J2-657:  Fixed AbstractDatabaseManager to close connection on writeInternal
error. Thanks to Stefan Wehner. 
>>>> o LOG4J2-713:  Android: java.lang.VerifyError: org/apache/logging/log4j/core/util/Closer
Thanks to Nelson Melina. 
>>>> o LOG4J2-703:  Android: Could not find class 'javax.naming.InitialContext',
referenced from method org.apache.logging.log4j.core.lookup.JndiLookup.lookup. Thanks to Nelson
>>>> Tag: http://svn.apache.org/repos/asf/logging/log4j/log4j2/tags/log4j-2.0.1/
>>>> SVN revision: 1614534
>>>> Web Site: http://people.apache.org/~mattsicker/log4j/2.0.1/
>>>> Artifacts: https://repository.apache.org/content/repositories/orgapachelogging-1005/
>>>> You may download all the artifacts by executing:
>>>> wget -e robots=off --cut-dirs=3 -r -p -np --no-check-certificate https://repository.apache.org/content/repositories/orgapachelogging-1005/
>>>> -- 
>>>> Matt Sicker <boards@gmail.com>
>> -- 
>> E-Mail: garydgregory@gmail.com | ggregory@apache.org 
>> Java Persistence with Hibernate, Second Edition
>> JUnit in Action, Second Edition
>> Spring Batch in Action
>> Blog: http://garygregory.wordpress.com 
>> Home: http://garygregory.com/
>> Tweet! http://twitter.com/GaryGregory
> -- 
> Matt Sicker <boards@gmail.com>

View raw message