Markus Eisele is an Oracle ACE  Director, a champion of Java Enterprise Edition, knowledgeable in messaging systems and is a German author of note. He lives and works in Germany; and Markus is famous for sending scores of valuable tweets on Java EE and related technology as the user @myfear. I am very pleased to have Markus be an early access reviewer for book and he recently accept my invitation.

Now that cat has been released; and it is finally free. I can let you also know a few more things. I am using my time on the contractor bench of truth to dig deep and finish writing the remaining content of the book. The aim of this book project is to have it fully ready for September 2013; published, printed and shipped for  JavaOne.

Currently, the chapters sort of look like this:

• Chapter 1 Introduction
• Chapter 2 CDI
• Chapter 3 EJB
• Chapter 4 Essential JPA
• Chapter 5 Intermediate JPA
• Chapter 6 Java Servlets 3.1
• Chapter 7 WebSocket 1.0
• Chapter 8 JAX-RS 2.0
• Chapter 9 JMS 2
• Chapter 10 Bean Validation 1.1
• and more

The content is subject to change, of course, and all of the usual caveat emptors and legal clauses apply. Sadly, I decided to drop the JSF content out of the book, because we are rather too close for comfort to page limits for hardware printers.

+PP+

PS: I have a couple of other secret helpers out there and they know who they are. Thanks to you also.Tweet

Part One

Devoxx 2012 JUG Leaders BOF Part 1 from Peter Pilgrim on Vimeo.

and Part Two (Devoxx 4 Kids)

Devoxx 2012 JUG Leaders BOF Part 2 from Peter Pilgrim on Vimeo.

Enjoy

+PP+

PS: Thanks to Rabea Gransberger for reminding me that I did record the entire BOF at the last Devoxx UK conference.

Tweet

The biggest benefit I have from ACCU this year is talking to people about Agile, the industry, the prospects of software development and getting a uniquely British perspective on where we all could be going in the year. I was surprised by the interest of ACCU conference goers into the functional programming movement. Lots of the attendee are well informed about C++ initiatives of Lambdas  in the next upcoming standard. The ACCU audience tends to be knowledgeable about the lower levels of abstraction, which are most of time closer to the metal [the hardware]. A few delegate will definitely written assembler code from the ground up once upon time.

I also have to mention the Bloomberg Game Zone; I had great fun playing Defender, Galaga and Donkey Kong. After 25 years, it was embarrassing to see how bad I have become with Defender, especially. When I was fourteen or so, I spent my youth and multitudes of ten piece coins “getting good”, the planet exploded many times, regularly I clocked over 100,000 points per game, kill scores of Mutants and Baiters. Now I was paltry, I struggled to get over 20,000 points.  Hot Tip: Mr. John Lakos please bring over Mr. Do, Gorf and PacMan arcade consoles to 2014.

Find the slide-deck to my talk on Wednesday 10th April, here:

I must come back to ACCU and do it special on intermediate Scala and Play Framework, something that stretches way beyond the beginner and introduction. We will see in 2014 and beyond.

Special shout-out, in no particular order: Astrid Byro, Phil Nash, Ewan Milne, Brian Marick, Martin Waplington, Schalke Cronje, Russel Winder, Roger Orr, Kevlin Henney, Michel Grootjans, Steve Love, Frances Buontempo, Detlef Vollman, Beth McKenzie, Ann-Jayne Metcalfe, James Slaughter,  Jon Jagger and Kjersti Sendeberg; and finally Julie Archer and Belinda Wiacek.

+PP+Tweet

Here is the Slidedeck for the talk:

And you can find the entire codebase of the demonstrations exported to Github as promised: https://github.com/peterpilgrim/devoxxuk2013-tdd-javaee7.

As I said before during the presentation, actually, I am writing a JavaEE 7 Developer User Guide book for Packt Publishing. Please look out for it as it will hit the shelves in late Summer / early Autumn of 2013. Finally, don’t be afraid to get in touch to ask any question on my talk and presentation, JavaEE 7 and development testing.

See you all at ACCU 2013 in Bristol, next month or perhaps further ahead to JavaOne 2013, San Francisco.

+PP+

PS: A pleasure shout out to Aslak Knutsen and David Blevins from Red Hat JBoss team for their hints and tips on this talk.

Some Captured Tweets:

Markus Eisele ‏@myfear 2h
#devoxxuk @peter_pilgrim about #Arquillian and #JavaEE7 pic.twitter.com/CSv2rb5LTr

Holly Cummins ‏@holly_cummins 3h
‘The cloud is perhaps the heaviest container there is.’ @peter_pilgrim #DevoxxUK

Aslak Knutsen ‏@aslakknutsen 3h
http://yfrog.com/od34253901j Mocks are the old way of testing EE, it’s time to move on. @peter_pilgrim #Arquillian #DevoxxUK

Steven Van Impe ‏@sipofjava 3h
Attending “Test Driven Development with Java EE 7, …” by Peter Pilgrim (@peter_pilgrim) at #Devoxx.

Dan Allen and Aslak Knutsen retweeted you
20h: 13:30PM Weds, Test Driven Development with JavaEE 7, Arquillian & Embedded Containers http://www.devoxx.com/display/UK13/Test+Driven+Development+with+Java+EE+7%2C+Arquillian+and+Enterprise+Containers … #DevoxxUK “Let’s go UK!

Dan Allen and Aslak Knutsen retweeted you
20h: 13:30PM Weds, Test Driven Development with JavaEE 7, Arquillian & Embedded Containers http://www.devoxx.com/display/UK13/Test+Driven+Development+with+Java+EE+7%2C+Arquillian+and+Enterprise+Containers … #DevoxxUK “Let’s go UK!”

adoptajsr retweeted you
20h: Thanks to @aslakknutsen and @dblevins for a very insightful #JavaEE7 discussion around the hack tables #DevoxxUK

 Tweet

I found the official Instruction for FullBuild of GlassFish and then decided to build the server myself. Sometimes, you may not want to wait for the GlassFish build files to be promoted by the team. In this entry, I reference Artifactory as a private Maven repository, of course, you can use something else as well.

Checkout the source code for GlassFish 4.0 yourself from Subversion:

svn checkout https://svn.java.net/svn/glassfish~svn/trunk/main glassfish-main

You need to hack the Maven Settings file for your workstation to exclude Eclipse artifacts.Here is an example of the settings.xml, which I configured.

<settings xmlns="http://maven.apache.org/SETTINGS/1.0.0"
  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
  xsi:schemaLocation="http://maven.apache.org/SETTINGS/1.0.0

http://maven.apache.org/xsd/settings-1.0.0.xsd">

  <!--Maven http://maven.apache.org/settings.html -->
  <localRepository/>
  <interactiveMode/>
  <usePluginRegistry/>
  <offline/>
  <pluginGroups/>
  <servers>
      <server>
         <id>ACME-ARTIFACTORY-PRIVATE</id>
         <username>administrator</username>
      <password>password</password>
    </server>
  </servers>
	<mirrors>
		<mirror>
		  <id>maven-central</id>
		  <url>http://repo1.maven.org/maven2/</url>
		  <mirrorOf>central,!eclipselink.repository</mirrorOf>
		</mirror>
	</mirrors>
  <proxies/>
  <profiles/>
  <activeProfiles/>
</settings>

Before we can compile the entire GlassFish code, we need to hack POM files so that they install artifacts into our private Artifactory server instead of the Maven Central.

Add the following Stanza to the POM files in the distribution:

	<distributionManagement>
		<repository>
			<id>ACME-ARTIFACTORY-PRIVATE</id>
			<name>acme-releases</name>
			<url>http://peabody.internal.acme.com/artifactory/ACME-ARTIFACTORY-PRIVATE</url>
		</repository>
		<snapshotRepository>
			<id>ACME-ARTIFACTORY-PRIVATE</id>
			<name>acme-snapshots</name>
			<url>http://peabody.internal.acme.com/artifactory/ACME-ARTIFACTORY-PRIVATE</url>
		</snapshotRepository>
	</distributionManagement>

In the above Stanza, edit the definitions from ACME to the Artifactory server that you privately own, then copy it the following POM files:

• main/pom.xml
• main/appserver/javaee-api/pom.xml
• main/appserver/pom.xml
• main/nuclues/pom.xml

This is a nasty hack, because I don’t like that you can’t set change the deployment server and credentials from a configuration. Other source code allow configuration of the deployment server through Maven Profiles or even property files.

Make sure that your Maven settings are correct for Artifactory deployment and also we set up Maven build process. Set the environment variable MAVEN_OPTS so that there Maven has enough memory and the permanent generation is high enough to avoid out of memory exceptions during compilation.

MAVEN_OPTS=-Xmx1024m -Xms256m -XX:MaxPermSize=512m -XX:+UseConcMarkSweepGC -XX:+CMSClassUnloadingEnabled

If you have 16GB RAM workstation, why not just max it out for compiling the entire GlassFish? The Garbage Collection algorithm is changed to the concurrent mark and sweep algorithm and we also set the enabled class unloading enabled.

You are ready to compile, entering the following commands:

cd glassfish-main
svn update
mvn clean
mvn install -DskipTests=true

Make yourself a hot beverage and snack for about 20 minutes on a decent Intel Core i5/i7 machine (2012). Have a break. Notice that we avoid running the unit tests here, we skip the tests, because we just want a working release in repo, quickly, which is just not to say testing is bad.

After successful compilation of all of the modules, now you are ready to deploy to the private Maven repository. If you have followed the earlier instruction, about copying the stanza to the individual POM files, then you can execute this command from root.

cd glassfish-main
mvn deploy -DskipTests=true

After deploying the artifacts to Artifactory, check the repository for snapshot 4.0 release, they should all be there.

Now descend to the Java EE project folder. Hack the POM file, glassfish-main/appserver/javaee-api/javax.javaee-api/pom.xml. It is missing the maven source plugin in the build section, and therefore, by default, it does not generate the sources JAR, which is useful for seeing the new JavaEE 7 APIs!

Find the XPath project/build/plugins and append the following stanza to this POM.

            <plugin>
                <groupId>org.apache.maven.plugins</groupId>
                <artifactId>maven-source-plugin</artifactId>
                <executions>
                    <execution>
                        <id>attach-sources</id>
                        <phase>package</phase>
                        <goals>
                            <goal>jar-no-fork</goal>
                        </goals>
                    </execution>
                </executions>
            </plugin>

Execute the following command line, to deploy the Java EE api artifacts

cd glassfish-main/appserver/javaee
mvn deploy -DskipTests=true

For some reason, the main execution does not install javax.javaee-api artifacts automatically. Executing this line generates JAR and SOURCES JAR for the three underlying modules: javax.javaee-api, javax.javaee-web-api and javax.javaee-endorsed-api.

Go Artifactory and see that the artifacts have all been deployed. You can then write a Gradle build file like this:

repositories {
    maven {
        credentials {
            username 'administrator'
            password 'passowrd'
        }
        url 'http://peabody.internal.acme.com/artifactory/ACME-ARTIFACTORY-PRIVATE'
    }
    maven {
        url 'https://maven.java.net/content/groups/promoted'
    }
    maven {
        url 'http://repository.jboss.org/nexus/content/groups/public'
    }
}

dependencies {
    providedCompile 'org.glassfish.main.extras:glassfish-embedded-all:4.0-SNAPSHOT'
    providedCompile 'javax:javaee-api:7.0-bpeter-private'
    providedCompile 'javax:javaee-web-api:7.0-bpeter-private'

    compile     'org.glassfish.main.extras:glassfish-embedded-all:4.0-SNAPSHOT'
    compile     'javax:javaee-api:7.0-bpeter-private'

    testCompile 'junit:junit:4.10'
}

Especially, note that the build version are annotated as 7.0-bpeter-private.

The last piece of the puzzle, which I have not yet worked out is how to configure the build.id Maven property so that I can customize the build number. It is a mystery, still. If you happen to know the answer, please give me a bell. Cheers!

+PP+ 2013
 

PS: The EclipseLink uses it own Maven repository for artifacts: http://wiki.eclipse.org/EclipseLink/Maven

 Tweet

I think the best advice, which I can give to you, personally is the Devoxx conference is cheap compared to many others, because Stephan Janssen gets the sponsors to back the cost of organising, setting up and running these event. The annual Belgian conference is, probably, the second most important Java conference in the world. There are even those, some other Java Champions, who say that it is really now in position one, since 2010, and because JavaOne is no longer held in the Moscone Center, San Francisco.

There is finally no need to pack a passport, you do not have to fly to another country. There is no packing of travel luggage, you can just pack a laptop in a rucksack. There is no need to kiss the spouse goodbye, because you will back in the evening to see her/him after a long day. It’s be just like going live with a application release, taking risky development system into the production, except that you will enjoy the experience and have fun and most important of all; you will meet other like-minded people. That’s has to be better.

Even if you attend only one day, take a day out of your annual holiday allocation, it will be worth it. (If you are really naughty, paraphrasing the Brit Rock band, Oasis; Some Might Say, “one could throw a sicky on that day” and therefore save the allocation day). Besides attending a learning experience is much better fun than working than the gravy train (or the workaholicism of 8am to 8pm working in some non-Agile based  investment banks). Heaven forbid that a conference, actually, might open the eyes of developers and designers to see what else could be out there beyond just technology; and break out of the path of getting a heart attack, stroke or unhealthy long term experience.

I have rambled on enough. Workers unite. Do yourself a favour. Go, do it!

+PP+Tweet

Day 4 Sessions of JPR 2013

The second session was a follow session to the first in many ways. Dick hosted a session; it was called “Types: How Much Can Compiler Do?”. Given static compiled language like Scala enforce type safety, Dick wanted to find out from other people how to ensure code will execute correctly by push the burden of type verification with semantics on to the compiler. Dick is obviously influenced by functional programming languages such Haskell. This may be considered advanced developers and programmer only, when you listen it in the podcast.

The final session of the Java Posse Round-Up 2013 was the “Open Source Business Model” which proposed and hosted by Hans Dockter. Bruce Eckels, Fred Simon and, of course, Hans were the main contributers to this discussion. If you are interested in running a professional open source business in near future, I believe this will be worth you while, as they discuss the various business models on service, product and consultancy oriented operations.

This wrapped up the conference. In the afternoon, there were a bunch of us, who went up to the mountain for a downhill ski or ride on a snowboard. It was great being with Jeremy Cerise, DJ Hagberg, Chris Phelps and Chris Marks. In particular, Chris Mark and I tore the mountain up!

A sunset view of the Crested Butte mountain outside of the Yurt

Inside the Yurt

The last event of the conference, proper, was the Yurt dinner, which James Ward organised very successfully. It was very well attended. The Yurt is a Mongolian hunt in the country side a couple of kilometres from the Crested Butte town. In order to get to the hunt, because it is inaccessible by road, the group hike with showshoes from the Gronk area of town to the Yurt location. The three course dinner was cooked by a quality chef. It is not free, we all had to pay about 75 USD, but it was delicious and well worth it. The biggest bonus was not the dinner or wine, it was the remoteness, and the absence of town lights. When I say we could see the stars, I mean, in truth, we could see stars aplenty. The milky way was fascinating, it was a bit hard to star up in the nightsky, but eventually I saw a faint band of dense stars arching overhead.

Inside the Yurt #2

Inside the Yurt #3

Time to wrap and go home. The end of the Java Posse Round-Up 2013. It has been a fun experience, I am glad I had the chance to travel to this open space conference, despite the initial airplane and weather problems. You do meet some of the best quality minds and humans on this planet. I have come away refreshed and I know exactly what I am going to focus on for the rest of the year.

+PP+Tweet

Day 3 sessins

On return to the Parish Hall, I found myself in Build Pipelines, hosted by Justin Ryan, which someone added my original ticket to, which was titled Gradle Build Systems and Other Non-Maven Builders. I found this session more about continuous improvement and delivery. There lot of discussion on Jenkins and writing plug-ins using Chef and Puppet. This was interesting for Justin Ryan, because he works at Netflix, where they face deployment and being able to streamline dependencies. A lot of talk was about master of control: who owns ultimately the artifact between teams? Especially when said artifact is shared between lots of teams. Is the upstream team on the Maven coordinates (Group, Artifact, Version) able to bump up a version from 1.0 to 1.1 on a particular Jar just to see if it will smoke test the downstream? So perhaps this talk was a little out of my current daily work focus. I did suggest one idea having a master POM with properties, but it was unpopular. Hans arrived halfway through this session and of course being a proponent of Domain Driven Design brought his insight into pipelines.

The next session on my schedule was How to Teach What You Know? hosted by DJ Hagberg and myself with my title How to Be A Better Presenter?. Whilst everyone else, mostly, were downstairs in the Parish Hall, we had a really great session in the Stained Glass area. I really thought we packed a lot in to the session and when I came away from the discussion I could easily added some more hints and tips. I also learnt some ideas from Daniel Hinojosa on how to best to prepare content for sessions. Romain Pelisse also led his knowledge and experience on training client site to the session; he had some very valid suggestions and comments to prepare slides using LaTeX, or another tool to build PDF. One of his ideas is to prepare a book from a text file alone, which he just gives to the student an hour before the training. I contributed some opinions from my experience of running a Java User Group, standing up in front of audience, and especially some insights from my partner, who runs a business abd personal communication business. It was generally a great session: I can’t wait to hear it on the podcast later this year.

So the afternoon, on Thursday, was taken up by more snowboarding, this time on my own. I admit that I came a cropper on the last run of the afternoon. The day was overcast and there were some snowflakes, but not enough to make the piste a slushy mush. On the contrary, it was fairly icy on some parts of the mountains, because of the cold temperature. Certainly, riding with Chris Phleps, on the day before, was much better.

A list of Progressive Dinners house for JPR 2013

There were other events on ground, which obviously I did participate in, but you could if you want to, enjoy hacking with Clojure with Bill Robertson, or make some headway into Haskell with Dick Wall. Or you could have gone on cross-country skiing with Diane Marsh or DJ Hagberg. Those were choices made. That was it.

This year 2013, we continued the tradition of progressive dinners, where the Round-Up people moved from one house to another whilst get dinner and beverage. It was a great dinner event. My favourite was the Gothic house, where Chris Cobb had a Spanish theme of food, chicken enchiladas and a supremely tasty marguerita. I think he made it strong; it was powerful stuff in that bucket of limes, gin and juice. Goodness me, even writing this text about it is giving me the memories! Anyway, it was, indeed, lovely food.

Romain Pelisse is talking to Guy Gascoigne-Piggford (at least I hope it is the back of his head!). In the Background (L-to-R) are Duncan, Gene Smalls, Joe Kramer is talking to Dimitry Kalinsky.

Karoake Nights

Thursday also was famous for the Karaoke night that Bruce Eckel and Dick Wall helped to organise at the LoBar. The Karoake night is set up by Bruce’s friend and they only got paid by tips. If you tipped more money, then it bought your song higher up the list in priority. It was a fun night, my favourite was Andrew Neely who worked his socks off with Digital Underground of Humpty Hump.

Andrew Neely performing a Karaoke country song

+PP+
Tweet

This is the official sheet that host must fill in to record a podcast session for the JavaPosse Round-Up; here is mine on Functional Programming

The second session of the second day was a little vacant. So I thought I am going to propose an session. I grabbed a post-it note and felt tip pen and scribbled down: Functional Java. Well what did I know. People was interested, people including Bill Robertson, Dick Wall, Bruce Eckel and Daniel Hinojosa. I think from this talk that I was letting my Scala learnings slip away. I know why, because I have writing feverishly on the Java EE 7 book, which is the main priority. I will get back on the Scala horse sooner rather later. Bill Robertson had a great deal to say about Closure and ClojureScript schemas. It was an interesting session to say the least.

The title Engineering Management Techniques and Insights was the final and third session of Wednesday. This was hosted by Barry Hawkins and Guy and myself. The other people were the instigators of the session, in truth. As guy put it: how can we lead without managing? The session revealed that there were no easy answers to great engineering management. There topics dived into a performance reviews, 360 reviews, Agile retrospectives; management by walking around the office was controversially seen by some, not particularly myself, as a bad idea. I believe this is going to be an interesting podcast, because you may or may not agree with the points on the tape.

In the afternoon, I met up with Chris Phleps and we went to the Crested Butte, the weather was gorgeous, absolutely kind to us. Chris is a skier and I am a snowboarder; the funny thing is that we are both owners of GoPro helmet camera. Chris had a first generation and I had recently invested 200 quid in a third generation. It was a lot of fun riding together. If you want to see more about winter sports, please see my other related blog entry.

Day 2 of the Conference Wall.

In between the sessions on Wednesday, I went to the Camp Four coffee shop just around the corner from the church. I saw this outside painting with oil colours the Crested Butte mountain. This local painter is called Shaun Horne; his paintings are displayed at the Telluride Gallery in Colorado, http://shaunhorne.com

Camp 4 Coffe shop just around the corner. Hmmm Coffee+++

The mystery cosmopolitan cocktail from the Princess bar on Tuesday night: actually, it tasted delicious. A chance for developers to unwind and truly forget about professional work and the day job.

Uh oh. I have been caught red-handed with the cocktail in hand! To my right is Bruce Eckel, the co-organiser of the JavaPosse Round-Up open space conference, Bruce Eckel. This photo was taken by James Ward, who was working behing the Princess bar, helping out Joe Webber.

+PP+Tweet

JPR 2013: Thursday Introduction #2 from Peter Pilgrim on Vimeo.

JPR 2013: Going Downhill with Chris Phleps from Peter Pilgrim on Vimeo.

JPR 2013: Thursday Wipeout! Last Run of the Day! from Peter Pilgrim on Vimeo.

Thursday Best Solo Download Run from Peter Pilgrim on Vimeo.

JPR 2013: Wednesday Video 1 (GPR0011) from Peter Pilgrim on Vimeo.

+PP+

PS: On Friday, I completely messed up the GoPro footage: wrong camera alignment. There should have been no excuses, because the Hero, Third Generation has the GoPro mobile phone app, which allows a user to preview the camera! So I missed out on recording the whole group on Friday: DJ Hagberg, Chris Marks, Chris Phelps, Jeremy and myself. Bloody Hell!

PS PS: See you all next time at the JavaPosse RoundUp!

Tweet