JRoller

Thoughts on Django

Rick Hightower — Sun, 8 Nov 2009 16:41:13 -0500

Django will make hair grow on the top of your head, and chicks will dig you. Ok, not so much!

Django is yet another rapid development web framework. Where Django differs from many other rapid development web frameworks is quality and consistency. Django aims to greatly simplify developing web applications. Finally, web developers can assemble an application soup to nuts without managing 50 other dependencies.

"Django was designed to handle two challenges: the intensive deadlines of a newsroom and the stringent requirements of the experienced Web developers who wrote it. It lets you build high-performing, elegant Web applications quickly." (The Django project website)
"Django focuses on automating as much as possible and adhering to the DRY principle." (The Django project website)

Django is an end to end web development environment. It is everything you need to write a web application. And in true Python style, it is batteries included.

Batteries included means, you don't have to go and find and manage 50 other dependencies to write a full functioning web site.

Django provides the following:

easy-to-use object relational mapper,
admin console for easily mapping your model to an editable GUI
extensible forms management when you need a custom GUI
create and add custom widgets (with JavaScript delivery thought out),
create and add custom fields,
extensible request handling through "middleware" support
out of the box security system with Users, roles and permissions
and much more

Django is easy to extend and modularize. It is easy to turn an application into a module that decorates many other projects. Developing with Django is a pleasure. For the most part, everything just works like you would expect.

(Expect this to be edited a few times as I flush out some more ideas...)

Felicidades Diana y Angel (Posada Miranda)

Julian Rivera Pineda — Sun, 8 Nov 2009 14:51:40 -0500

Les deseamos toda la felicidad y bendiciones que Dios y la Virgencita les puedan dar, que en su hogar reinen la paz y el amor, son los deseos sinceros de la Familia Rivera Hernández (Xtecuan, Pocahontas y PrincessCuan)

Carbonado 1.2.1 released

Brian S O'Neill — Sun, 8 Nov 2009 13:21:56 -0500

Carbonado is another project I work on, providing relational abstraction for JDBC and Berkeley DB. I just released version 1.2.1, one year after the release of 1.2. I should have released sooner, but I've been busy with other projects.

Version 1.2.1 is mostly a bug fix release, but it adds a few new features as well. The most notable feature addition is value compression for Berkeley DB. It is enabled on the BDBRepositoryBuilder:

  BDBRepositoryBuilder builder = new BDBRepositoryBuilder();
  ...
  // Enable compression for a few types.
  builder.setCompressor(MyStorable.class.getName(), "GZIP");
  builder.setCompressor(AnotherStorable.class.getName(), "GZIP");

Compression is applied to all the data values of the Storable, but the key is never compressed. The design allows compressed and uncompressed records to co-exist, and changing the option does not alter existing records. An explicit update is required to force an existing record to be converted.

The design also allows additional compression algorithms to be supported, but only GZIP is provided for now. Records which have been compressed with different algorithms can also co-exist.

The compression feature should only be enabled for types which are relatively large. Otherwise, GZIP will add needless performance and storage overhead. GZIP needs to encode a little extra data even to indicate that the record wasn't compressed at all.

All the changes to Carbonado 1.2.1 are summarized in the release notes:

- Fixes in replicated repository resync when encountering corrupt records.
- Fixes for derived properties.
- Added option to allow derived properties to be copied to other Storables.
- Added record compression support for BDB repositories.
- Added hot backup capability for BDB repositories.
- Added nested rollback support for BDB-JE.
- More BDB options exposed.
- JDBC logging includes bind variable values.
- Fixes for JDBC CLOB updates.
- Fixes for excessive class generation.
- Query.after and Query.fetchAfter support subclass parameter.
- Support covariant storable properties when subclassed.

Swing Resources

Dhilshuk Reddy Jeeru — Sat, 7 Nov 2009 10:02:25 -0500

Best Sites for Java Swing

Please visit the sites below they are real helpful.I Might be missing many valid sites.
Please mention those list of site in comments so that I can add to existing list

Java Desktop

Java Desktop Community

Swing Depot

http://www.manageability.org/blog/stuff/

http://www.javootoo.com/ Java Look and Feel

General Purspose component Libraries
Components & Tools for Specialized Applications
Graphs,Diagrams and Charts
Date Choosers
Layouts
JTable & JList
- GlazedLists
Swing and CSS
- Introducing Java CSS
- Swing and CSS
- Graphical Data Displays with CSS
- Introducing Java CSS
Look and Feels
- Free Look And Feel
- Commerical Look and Feel
- Free & Skinnable Look And Feel
BallonTips
- BalloonTip
Java Visualization
- GINY
- JUNG
- HyperGraph
- JDigraph
- WilmaScope
- JGraphT
- TouchGraph
- GVF
- JGraphEd
- VGJ
- Zoomgraph
- Walrus
- Prefuse
- ZVTM
- InfoVis Toolkit
- InfoVis Cyberinfrastructure
- Large Graph Layout
- NV2D
- GUESS
- Otter
- Gravisto
- Linguine Maps
- Rox Graph Theory Framework
- JIGGLE
- Zest
- Relo
- SoNIA
Java WorkFlows
Java Expression Builders
- JEvaluate
- JEP
- JEL
- JEC
- JEXL
- Eval
- Jbcparser
- janino

Portlet life looks like EJB, but future sounds like GWT

Dominique De Vito — Sat, 7 Nov 2009 04:52:44 -0500

Portlets face different problems, but GWT looks like the solution.

Portlets are not small applications

Some documents, like this one, say "Portlets are small applications that are independently developed, deployed, managed, and displayed". I disagree. Because this vision is somewhat exaggerated.

If portlets were small applications, they should have OSGi-like features, like a unique id(entification), dependency definition and management, public part (API) and private part (implementation), etc. AFAIK, portlets don't have such features, and don't look really like small applications.

I believe the game portlet/OSGi may end up like a remake of the game between specialized supercomputers using ad hoc technologies (~ portlet) versus clusters of PC (~ OSGi) using COTS: at the end, OSGi will win as written in Portlets in the light of OSGi and other technologies, because OSGi modules are much more small applications than portlets. Well, there are already some JavaEE servers out there that enable to deploy OSGi modules authorized to call JavaEE API (see Spring or JOnAS servers). So, yes, one could imagine to implement portlets as OSGi modules.

Portals should have less priority than business solutions

Some architects layer the back-office while separating the presentation layer (think JSP) and the business layer on different machines. From the same point of view as above, I don't see portlets living into the business layer, but into the presentation layer, and then, accordingly, I don't see portlets as small applications.

So, again, IMHO, portlets are much more about (graphical) integration than business integration (even security, said to be specific to the portlet side, could be, in fact, set outside - think about servlet filter). One proof that portals are about integration, and that a good part of it is graphical integration, is one could see portal incarnations moving on client-side. Mozilla recently announced its latest project Raindrop, as said here: it's "an application intended to bring all of your online messages to one portal; That includes Email, tweets, RSS, and social network updates all in one place." Another portal incarnation from Mozilla is Snowl - see my post Snowl, a portal concurrent solution on the client-side.

Let's imagine one want to develop a portal with content presentation/management features. I am more keen to choose a CMS/ECM, and next, to choose a portal if it's really needed and brings added-value; then, IMHO business compliance should be first and presentation has only a second-level priority, as portals, then, do have.

I have already wrote here that many claim a portlet is sweet because it enables to include multiple views inside a single page. OK. But it's only about the entry page (in most cases). Then, on the technology side, using portlets for developing a portal web site is just like a house builder that would choose to build a wooden house just because one single reason, that is, because they have chosen first a wooden door! It just sounds like weird! Are you going to buy a business solution because your top priority is the look and feel of the entry page ? I don't think so.

So, portals are not silver bullets.

Portlets have had different lifes, like EJB

Portlets have had different lifes and a hard beginning.

I remember my first contact with a portlet solution. The vendor told me the following about its portal solution:
- Session clustering was not possible because some misalignment (?) between portlet session and http session.
- Presentation coding was not anymore possible within JSP: developers have had to code presentation into... servlets (oops).

Hard time, for sure.

Portlets have improved. And they have been kicked by AJAX arrival, so again they have given an old-style feeling. And they have improved again. So, portlets have surfed some times on top of the wave, at least, the buzz wave, and they have been, quite often too, at the bottom!

And it looks like portlets are going to be kicked again by other technologies (see below).

Portlets may have been simpler through more convention over configuration

I wonder if the problems portlets face may come from the fact that portlets are defined at the API-level, but also at other levels, like architecture design and at the framework level too, as the 2 following points of the specification suggest:
# two phases of action processing and rendering in order to support the Model-View-Controller pattern.
# portlet modes, enabling the portal to advise the portlet what task it should perform and what content it should generate

While portlet design is not enough on the server-side, it has impacts:
- if one choose first to develop portlets, the choice, next, of a web framework, and its integration with portlets, could be cumbersome (that's the reason bridges were proposed), and that's annoying because a web framework is more important than portlets that are here to design only the home page.
- another way is to choose a framework web (other that portlet's one) and next, to follow architecture design of the portlets (without using such a dedicated portlet framework); it's much more doable these days as, for example, existing JavaScript frameworks provide building bricks.

I wonder too if the portlet specification would not have been simpler following convention over configuration approach (for example, while defining a jportletid parameter, like there is a jsessionid at servlet level). May be portlet definition would have been simpler, without portlet framework, leaving developers to use their preferred web framework for implementing portlets.

Portal layout using GWT with Ext JS

GWT is going to be a portlet alternative: in fact, GWT sounds like portlet death

Well, I think the more important portlet's problem comes from it's a technology with ass between two chairs! This technology sits within the architecture/framework/API different levels, is both present into client/server side, is at the presentation/business interface... Then, portlets fight on all sides. So, they fight to compete with various solutions, and they are often shaken up by forward moves into connex/close domains (e.g. AJAX introduction). So, the portlet ROI is under pressure.

Portlets have had hard time where AJAX appeared, but now they are going to have much more hard time trying to compete with GWT.

GWT sounds like portlet death.

Why ? Portlets are cross-frontier technologies, being on the client-side and on the server-side too. And guess what !? GWT is a cross-frontier technology too, quite pretty good to mix easily JavaScript/Java and client/server sides.

GWT-based portal layout in action with Ext JS

So, if one would have to specify portlet again from scratch, IMHO a GWT-inspired approach would be at the center of such work.

Just take a look at Ext JS widget library for GWT. Ext JS library provides portal appearance: take a look at this demo.

So, on one side, you have portlet frameworks, with a lot of associated problems, like integration with your favorite web framework, and learning curve. On the other side, you have the quite popular, and powerful, GWT framework. GWT is evolving fast, and provides already most of the portlet features, except may be the hot deploy of a new kind of portlet on the server-side; but, still, GWT is evolving fast, and I prefer to bet on it instead of portlet technology.

While GWT could provide already most features portlets provide too, and much more, while GWT is a general purpose framework, and portlet frameworks are specialized ones, while GWT has a huge and growing community, the question is: are you going to bet on GWT or on portlets ? I don't believe in the second answer, over the first one.

Conclusion !?

I have written previously the term "portal" is, at least, rather confusing! But one thing is clearer now, portlets, as we know them today, have a dark future, close to no future at all; or to say things differently, I have difficulties to see portlet future. There could be specific use cases, but they seem a minority.

This being said, while this post is mainly discussing the portlet concept, it is not against portals, even if I don't forget to keep my eyes open when I read the word 'portal' as this is quite a fuzzy concept as I have written here.

Preview of the Portlets in Action book available

Jakub Holý (Holy) — Sat, 7 Nov 2009 04:40:24 -0500

The book Portlets in Action being written by Ashish Sarin, which I've already mentioned and which looks really promising, has been made available via the Manning Early Access Program. As of today there are two chapters available and you can read the first one "Introducing Portals and Portlets" for free.

What I find to be the most attractive features of this book is that it concentrates on the "new" JSR 286 also known as Portlets 2.0 and goes beyond teaching portlets with its intorduction of Spring MVC and Ajax libraries suitable for portlets.

Java Softwares and IDEs and Builders

Dhilshuk Reddy Jeeru — Sat, 7 Nov 2009 02:46:19 -0500

Java Softwares and IDEs and Builders:

Java SDK

Install the Latest Release of the Java SE Platform You can download the latest release of the JDK for free from here

IDE's

GUI Builders

Open Source & Free

Commerical

Links Matter For SEO

evans — Sat, 7 Nov 2009 00:44:00 -0500

Links are a source of traffic.

The more inbound links you have to your website, the higher the chance your potential customers could discover your products or services.

Relevant inbound links improve your organic search engine rankings.

Search engine robots crawl the web to analyze the vast amount of websites' contents and prioritize the relevancy of each website. Building relevant inbound links to your website with relevant anchor text is the most important thing you can do to improve your organic search engine ranking.

Your Website's Content And Inbound Links

Both content and inbound links are important for your website to rank high in search engine. Google controls 70% of search market and one of the most important factors in Google's search algorithm is inbound link popularity. The important factors when building inbound links are:

The number of inbound links to your website
The quality of sites that link to your website
The relevancy of sites that link to your website
The anchor text (Anchor Text) of links to your website

Ways To Get Inbound Links To Your Website

There are multiple ways to get inbound links to your website depending on your budget. Many small webmasters or bloggers often do manual link exchanges with other relevant sites. This process takes a long time and efforts, and two-way links are not as favorable as one-way links.

Larger companies often hire SEO professionals to do the arduous tasks of link building.

An effective method to build inbound links is to buy text links ads on relevant websites. Tips When Buying Links

Buy links from relevant websites some companies have search functionality that allows you search for keywords that exist in the publishers' websites.
Use various anchor text that reflects your Top Keywords. e.g: "Bad Credit", "Credit Card Loans", "Buy Home Insurance"
Buy links that point to an important page in your website e.g: You may want to link directly to the category page of your e-commerce website. If you sell electronics, then you may want to link directly to LCD TVs category ( http://www.mystore.com/widescreen-television )
Keep your links for at least three months to see results
Search engines also classify websites based on their IP addresses. The more unique websites with different IP addresses linking to your website, the better.
Track your results by looking at your web analytics (Google Analytics, Omniture, etc). Look at "organic" traffic referral for Keywords. e.g: "Bad Credit", "Personal Loans", "Buy Home Insurance".
Use Yahoo's Site Explorer to track the number of inbound links to your site. You can do query such as: "link:http://www.yoursite.com"

To sum up, inbound links are important factor for you to increase your traffic from search engines. The right content optimization and link building campaign can help make your website appear at the top of search engine's results for your keywords, and thus bring more customers to your website.

JBoss Best Practices & To Do List Before Production

Hakan ERDOGAN — Fri, 6 Nov 2009 17:09:29 -0500

[Important] Do not use all, default, and minimal nodes. Just copy whichever you need and rename it such as myapp_node, configure it, and keep this node in a SCM tool such as SVN or CVS.
[Important] Secure your JBoss consoles (jmx-console, web-console, jbossws), root application and change the context path of the root application. Here is the guide:
http://www.jroller.com/hakan/entry/jboss_security
[Important] Log4j Configuration. Use RollingFileAppender instead of DailyRollingFileAppender to prevent storage problem (use fixed size and number of log files).Here is the guide:
http://www.jroller.com/hakan/entry/defining_a_rolling_file_appender
[Important] Start your JBoss with -c and -b parameters (-c config_name -b bind_ip). It supports multiple JBoss instances running on a same server.
[Optional] Secure your EJB's, Webservices, public access resources, etc.
[Optional] Hsqldb to Mysql migration (for JMS, EJB3 Timers, etc.) Here is the guide:
http://www.jroller.com/hakan/entry/jboss_hsqldb_to_mysql_migration
[Optional] Change WSDL's ip address with pool ip, if you have a load balancer or a domain name for your webservices.Here is the guide:
http://www.jroller.com/hakan/entry/changing_wsdl_s_ip_address
[Optional] Enable Clustering and Session Replication (for Web App.)
[Optional] Set Login Modules Configuration & Auth Policies (LDAP, DB, File, etc)
[Optional] Set Transaction timeout, JVM Startup Arguments

JAVAWUG PROJAM #1 JQuantLib

Peter Pilgrim — Fri, 6 Nov 2009 10:28:15 -0500

JAVAWUG PROJAM 1 JQuantLib

PRO JAM #1
Tuesday, 17^th
November 2009
Skills Matter, London, UK

STOP PRESS: SKILLS MATTER HAVE MOVED OFFICES!

This wonderful London night time was by the talented and versatile imutto, Creative Commons license on Flickr.

Dear All

I would like to formally announce that the very first JAVAWUG PROFESSIONAL JAM (Java Web User Group) taking place at new location Skills Matter on Wednesday, 17^th

November 2009 from 18:30 until 20:30.

Talk

Title: JQuantLib and Financial applications

Your ProJam Organiser: Richard Gomes

Abstract

This talk demonstrates how open source software can be employed to implement enterprise financial applications in pure Java. There will be a mix of business concepts and technical details being presented in order to converge to a working demo application which could be deployed typically in a computing grid.

Target audience
This presentation is intended to experienced Java developers.
Previous experience on Eclipse IDE and Maven2 are desirable but not required to take advantage of the presentation.
Interest on financial applications in enterprise environments is desirable.

What you will learn
. Basic business knowledge about option price valuation;
. Support classes needed to represent those business concepts involved;
. Hello world: performing a simple option price valuation;
. Going further: the architecture of our distributed demo application;
. Implementing and executing a distributed application which valuates live market data.

What you will be doing
You can opt for bringing a laptop to this presentation, download the demo application and try it out yourself.
The location for download will be announced in the beginning of the presentation.
In theory, all you need installed in your PC, Mac or Linux laptop is:

. JDK6
. Ant 1.7.1 or more recent
. Maven 2.2.1or more recent
. Eclipse 3.5

More information
JQuantLib is a free, open-source, comprehensive framework for quantitative finance, written in pure Java.
It offers several mathematical and statistical tools needed for valuation of financial instruments like options, bonds, swaps, commodities and alike.
See also: http://www.jquantlib.org/

Devoxx Linked-Up

Peter Pilgrim is sadly at Devoxx 2009 when the projam takes place. It might be possible to arrange live link-up of 20 to 30 minutes between Antwerp and London. Although technical difficult to know before the head-scape in order to achieve this fantastic goal.

These are some introductory information related to Quantitative Finance:
http://en.wikipedia.org/wiki/Financial_markets
http://en.wikipedia.org/wiki/Derivative_(finance)
http://en.wikipedia.org/wiki/Quantitative_analyst

Biography

Richard Gomes is Brazilian and graduated at Unicamp, Brazil.

In the beginning of his career he developed numeric libraries in assembly language and a symbolic debugger whilst working for CPqD/Telebras R&D Center, which was a key player in the emerging Brazilian telecommunications scenario at that time.

He has several years of experience in IT and experience with full software life-cycle, working as developer, architect, production support analyst and configuration manager in complex projects with multiple development branches. He has experience with high volume, low latency, critical systems specially in banking and telecommunication sectors. He worked for companies such as: Groupe Bull, IBM, EDS and Unibanco.
His contact information and curriculum vitae can be obtained from http://tinyurl.com/frgomes

Richard is settled in London since September/2006.

Registration

Afterwards members can retire to the nearby pub/restaurant for more in depth discussion dinner, food and drink ...

The address is:

The Skills Matter eXchange
116-120 Goswell Road
London, EC1V 7DP
United Kingdom

Phone: 020 7107 2620
Fax: 020 7107 2621

If you would like to attend: NEED TO UPDATE THE REGISTRATION

Please REGISTER so that you can be added to the SECURITY DETAIL with Skills Matter Limited.
If you aren't a JAVAWUG member why not join the JAVAWUG at Google Groups?

For problems, send an email to peter (dot) pilgrim (at) gmail (dot) com and emin (dot) tatosian (at) gmail (dot) com.

Here is some relevant travel information:

By Underground: -
Barbican: Come out the station turn left to facing the o
Map: Google Maps Link to 116 Goswell Street

The venue has graciously been organised by Skills Matter team. We all appreciate this generous gift.

JAVAWUG.com

Media Copyright Notice

JAVAWUG BOF Events are video recorded, edited and published freely. We normally recorded the speaker and administrators. You can find examples on Vimeo and Google Video. On the rare occasion we record some audience members. Please let us know if you personally would rather not been filmed. We reserved the Copyright 2008 © JAVAWUG Video Productions, Creative Commons (Attributions: Attribution (by): Licensees may copy, distribute, display and perform the work and make derivative works based on it only if they give the author or licensor the credits in the manner specified by these.).

This is Peter Pilgrim, Out.

Roma SpringFramework Meeting Slides

Massimiliano Dessì — Fri, 6 Nov 2009 09:26:03 -0500

MongoDB and Scala

Scale your database and be happy

Spring JCR Extension

Spring Ldap

LambdaJ

Non conventional Android programming

The term "portal" is, at least, rather confusing!

Dominique De Vito — Fri, 6 Nov 2009 05:12:43 -0500

The Gartner's research note Magic Quadrant for Horizontal Portals (september 2009) gives the following market vision: "By 2011, Gartner expects at least 15% of new enterprise portal projects in Global 2000 firms to use open-source horizontal portal frameworks" and "By 2014, horizontal portal products based on portal containers will be used for no more than 60% of new enterprise portal projects." I have found this note into the French blog Le Touilleur Express.

Well, the term "portal" is, at least, rather confusing!

There is often confusion between integrated access (it's about interface) and collaborative suite (it's all about implementation(s)). And these categories could be even refined into portlet-based solutions or other solutions.

So, what about the future ?

I could imagine that most firms will be using, by 201X, a collaborative suite or parts of it. From that point of view, I could agree with Gartner. But I have hard time to believe that most firms will be using massively portlet containers (see JSR-168 and JSR-286 for example) as chosen portals.

Portals sound like communism: the idea is good, but the problem is on the implementation side

Well, this communist metaphor is a bit harsh for portals, because it could be applied to many other software solutions; but portals deserve that metaphor, may be more than some other solutions, because the term "portal" is related to various realities, particularly because the portal business is still evolving.

Some could say the confusion about portals is natural, but it's still quite irritating because this confusion is related to basic separation of concerns, that is, interace versus implementation! I suspect some vendors, or consulting firms, may even take benefit from that confusion and perspectives it draws.

So, the sentence "By 2014, horizontal portal products based on portal containers will be used for no more than 60% of new enterprise portal projects" says... little!

When It's OK Not to Write Unit Tests

Sebastian Kübeck — Fri, 6 Nov 2009 05:03:25 -0500

In a recent post, Uncle Bob is once again furiously defending Unit Testing against Cashto's post It's OK Not to Write Unit Tests. I personally don't thing there is anything to get outraged about Cashto's post. Cashto's claims are so superficial and weak that the whole post was probably just written to attract attention and it is likely that we see more such posts as long as Uncle Bob reacts this way.

However, there are cases where you don't need to write Unit Tests.

It's OK Not to Write Unit Tests if You Write Code That Doesn't Have to Work

Not every code needs to work (see If it doesn’t have to work, I can get it done a lot faster!). Let's assume you write a prototype or do some experimenting with technology. There's really no need to write Unit Tests for that. You might also sketch some initial drafts when you don't yet know how to implement functionality as I have illustrated in When Quality Really Does Not Matter.

It's OK Not to Write Unit Tests for Functionality that is Tied To Infrastructure

With infrastructure I mean databases, UI libraries etc. that don't provide a level of abstraction that lets you add non fragile tests (see Properties of Good Tests). In this case it's probaly better to write integration tests first. Then, you extract as much idempotent (see Limitations of Automated Testing) functionality as possible into components that can be easily Unit Tested.

It's OK Not to Write Unit Tests if The States They Should Cover Are Already Covered by a Different Verification Mechanism

Let's say you are using Design By Contract in addition to BDD as Greg Young explains in his talk TDD in a DbC World. In this case, you can limit your test writing efforts to what is necessary to drive your behaviour definition. The rest of the verification is covered by the contracts and verified by the theorem prover and/or a model checker.

Conclusion

Unit Testing isn't a religion. It's just a helpful method to improve the software development process in adding automated verification of functionality. For all folks that think like Cashto: I highly recommend that you start learning the craft instead of missing a chance to improve your programming skills. There are plenty of good books and training courses around. And believe me: it isn't rocket science. Most of the things you are doing in your current work are much more difficult than writing good Unit Tests!

Patent It

Russ Olsen — Thu, 5 Nov 2009 21:26:26 -0500

Here‘s another article I wrote for Art/Geek that I thought would be interesting here.

You can listen to the podcast over at www.artslashgeek.com

Techies and artists seem to have so little in common. We manipulate electrons and bits with manic precision. We dream of being the soul of a new machine. Their souls machine new dreams. We write code.They write koan. But when it comes right down to it, we are all in the business of creating ideas, concepts, schemes, designs. We do have some things in common though: Both artists and techies want to get our due for our work, if not in money than in credit.

The law provides for three ways to make sure that you get the benefit of your ideas, your work and your brand. In fact, if you live in the U.S., some of that protection is built right into the most basic law of the land, the Constitution. Article I, section 8 says that Congress has the power…

...To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries.

Lawyers generally interpret this to mean that Congress has two separate powers, one to grant patents and a second, different power to grant copyrights. A patent is the right of an inventor to prevent others from using his or her invention. A patent is a kind of a bargain: The inventor gets the exclusive right to profit from their invention for a limited term (20 years in the U.S). In return, society as a whole gets the invention when the patent runs out. The idea behind patents is that it will be better for inventors to get a patent than to try and keep their invention secret.

Getting a patent is not easy, nor should it be since a patent is a very broad monopoly. If you hold a patent on some invention, I can‘t use that invention without your permission, even if I thought of it independently. In order to get a patent, you need to come up with some invention that is really novel, novel in the sense that no one has thought of it before. Recently there has been a lot of criticism of the U.S. patent system for granting patents on fairly obvious ‘inventions‘. One of my particular favorites is U.S. patent 6,004,596 which provides a exclusive monopoly to the inventors of:

A sealed crustless sandwich for providing a convenient sandwich without an outer crust which can be stored for long periods of time without a central filling from leaking outwardly. The sandwich includes a lower bread portion, an upper bread portion, an upper filling and a lower filling …

Good thing this patent wasn‘t around when my mother used to stamp out sealed peanut butter and jelly sandwiches by making the regular kind and then using a coffee mug as a sort of cookie cutter.

In contrast to a patent, a copyright is easy to get. In fact it is automatic. Simply by typing these words as I am right now, I am magically creating a copyrighted work. Copyright protects the authors of any original work, be it prose, poetry, music, art and even computer programs. Copyright means that the author has exclusive ownership over the work for some period of time. For new works copyrighted in the United States (this article for example) the copyright lasts the whole life of the author, plus 70 years. Although copyrights are easy to get, they are a much weaker protection than a patent. A patent means that you more or less own the invention, the idea. A copyright means that you own your expression of the idea. Thus, while I own this article and could sue you if you turned the first few paragraphs into the lyrics of a Country and Western song, I would be helpless if you simply went out and wrote your own Country and Western song about patents and trademarks.

Although copyrights are automatic, you can register your copyright. Registration is mostly a processes of sending a copy or two of your work to the copyright office and paying a small fee. Even though you automatically own the copyright on whatever you create, registration strengthens your hand if you do need to sue over that Country and Western song.

Finally, there are trademarks. A trademark is a name or a symbol that you use to identify your business or product. The idea of a trademark is that you should be able to stake out some kind of symbol or name or catch phrase that uniquely identifies your product or business. Like copyrights, trademarks are pretty easy to get. In the U.S., you can register your trademark with the government in pretty much the same way that you register a copyright and for pretty much the same reasons: If someone violates your trademark, you are going to have an easier time in court if you took the trouble to register it first. Interestingly, different companies can use the same trademark, as long as they are in very different businesses. Sometimes this leads to interesting collisions, as when a small computer company founded in the 1970s took the same name, Apple, as an existing music publisher. After all, what could personal computers have to do with music? The rest is, as they say, legal train wreck history.

You can find all sort of further information at the web site of the patent and trademark office and also the copyright office.

Russ

"Twittering" TheServerSide Java Symposium

Daniel Lopez — Thu, 5 Nov 2009 15:49:10 -0500

I've had a couple of days to let things get back to normal after coming back from Prague, so I thought it was time to put write down about the overall experience of being at the TSSJS Europe 2009.

Overall, I think it was a great experience, as I met people I already knew, made some new friends, listened to interesting topics and visited, even if for a short while, a nice city like Prague. Compared to past experiences like JavaOne and the then-called JavaPolis (now Devoxx) conferences, the TSSJS was smaller in scale, but that also meant it was less crowded and you got to more easily come across the same people during the conference, so getting to know them was easier.

The first thing I felt was that the organization was very well done, as, for example, I received an email before going to the conference with straight to the point information about what I could expect from the conference. I was also able to browse quite complete summaries of the talks that were going to be presented, so I could decide if I was interested or not in the talk, and not feeling like you are being "cheated" to go to a session by a catchy titles and 2 sentences as it sometimes happens. The slides were also available before hand, but not many were up before the event, so that was a good point that didn't work 100%.

The Clarion hotel is nice and as Geertjan Wielenga pointed out in his report, the first thing to notice was that there were power cords almost everywhere in the rooms, so I was able to hook up my laptop and twitter about the sessions without worrying about the batteries running dry.

That was my first time twitting ever, so I was not sure what people was expecting or what they wanted, but I hope it helped people get an idea of what was going on in the conference. Twitter failed during the last part of one session, Rapid Application Development with Apache Wicket - Andrew Lombardi, and I could not get in again in time for the next one, Open Source Debugging Tools: The Top 10 App Boot Camp - Matthew McCullough, so that's one drawback of using "instant" report tools: if they fail, when they come back again you are already reporting something else so there's no time to go back and report what you missed except in a broad sense.

Some people also wondered about the lack of personal opinions on the feed, but I felt it was not "fair" for the audience and the presenter to introduce my own subjective thoughts about it. One the things you notice when you go to conferences and actually listen is that everybody has its own background and what might sound familiar and obvious to you, might be new and eye-opening to other people, and vice versus, so I'd rather let everyone for their own opinion.

Another interesting side effect of reporting through twitter is that you can see there are sessions that do not "translate" well to twitter language. Catchy phrases, short ideas in a structured order are easier to put into sentences of 140 words, whereas complex ideas with abstract elements that need a deep background explanation before getting to the point are difficult to put into such a concise way. That does not mean the session is not good, but it just makes it more difficult to broadcast it in short pieces of text and feels like watching a movie by just reading the subtitles.

In any case, I hope it helped those that were not able to make it but remember that, as Geertjan also pointed out, half of what you get from a conference is networking with people, and that cannot be twittered, so if you were unable to make it due to being in the "wrong" J/K side of the Atlantic ocean, you have another chance in the upcoming TheServerSide Java Symposium to be held in Las Vegas in March 2010.

A final word on the conference itself: The idea of labeling lunch tables with topics so people with similar interests can get together and talk while having lunch is nice, but the truth is that there was no room for everybody in some tables, there was not much time for lunch & talk and it took quite a while for the last people to get their food... so everybody was busy enough finding a seat and then having lunch to worry much about the topic of the table or talk about it. I think the idea is interesting but it should be separated from the lunch time.

Disclaimer: I attended the conference with a press pass offered by the organization so we could report on the conference; if someone thinks that this circumstance had any effect on what I wrote about the conference, then they probably don't know me well enough ;), but I felt it was fair for the readers to know.

Cheers!

Learning ICEfaces through ICEfusion

Ted Goddard — Thu, 5 Nov 2009 12:17:19 -0500

Rainer Eschen's book ICEfaces 1.8: Next Generation Enterprise Web Development is now available in print. The book is as much about ICEfusion, a set of model classes and Facelet components built on ICEfaces, as it is about ICEfaces itself. Developing a framework that embodies the lessons of a book can be a very effective (and testable) way to communicate. Many thanks to Rainer for this great contribution to the ICEfaces community.

You may also want to check out Rainer's ICEfaces Mind Map.

Visualize metrics with Panopticode

Dhaval Dalal — Thu, 5 Nov 2009 11:03:55 -0500

I can't stop appreciating Neal Ford's book The Productive Programmer, full of good tips, one of which is "Generate Metrics with Panopticode". I finished integrating Panopticode (http://www.panopticode.org) on my current Java Project to visualize metrics. Panopticode uses Tree Maps to display the code complexity and coverage. On my earlier projects, I have used several tools, and all generated different types of metrics, which I would have to then collate and make sense out of it, especially to strategize and prioritize parts of code that need Refactoring.

With Panopticode, my collation and analysis task simply vanishes, instead now I can look at the Interactive Tree Map and the parts in RED refer to highly complex method. Click on the RED square and all the metrics, CC, NCSS etc... for a method appear on the side. Here is the Code Complexity Tree Map of my project.

Panopticode makes this novel use of Tree Maps for metrics visualization. To know more about Tree Maps, read this article by Ben Shneiderman. I also am not going to go into the details of how Panopticode works, but talking about it in a nutshell, it basically uses other tools like Checkstyle, JDepend, JavaNCSS, Cobertura/Emma, Simian etc... and aggregates data from these tools. The aggregated data is then rendered as Tree Maps. It produces two Tree Maps, one for Code Complexity and another for Code Coverage.

I have found another interesting use of Tree Map from my colleague and friend Anurag and they basically use the Tree Maps to radiate the current state of architecture of various components across the entire project. I don't know how they obtain the metrics, but the use of Tree Maps is noteworthy.

Component	Scalability	Maintainability	Security	Availability	Performance	Extensibility	Reliability	Manageability	Usability
Component 1
Component 2
Component 3
Component 4

If you are a products company, you can extend this to your product suites as well...

If you have made use of Tree Maps on your project and would like to share, please use the comments below....or send to me the link and i'll embed it in this blog post. Also, I don't know whether such a tool exists for the .NET world. If you happen to know, drop in a line ;-)

Spring 3.0 - Slides from a Seminar I Presented

Eyal Lupu — Thu, 5 Nov 2009 07:51:15 -0500

Hi all,
About two weeks ago I gave a 3 hours overview lecture on Spring 3.0 new features in a seminar organized by Interbit. For those of you who are interested you can find the slides here.

Mobile Twitter Client with JSF

Cagatay Civici — Thu, 5 Nov 2009 04:26:48 -0500

http://cagataycivici.wordpress.com/2009/11/05/mobile-twitter-client-with-jsf/

Scala and MongoDB

Massimiliano Dessì — Thu, 5 Nov 2009 04:17:25 -0500

MongoDB Scala Roma SpringFramework Meeting2009

View more documents from Massimiliano Dessì.

GridGain presents at Dallas Java User Group, November 11th

Nikita Ivanov — Thu, 5 Nov 2009 02:51:29 -0500

GridGain will be presenting at Dallas Java User Group at Dallas, TX, on Wednesday, November 11th. If you are around the Dallas area and interesting in Cloud Computing - come see our presentation. As usual we'll have lively discussion on clouds and grids and as always plenty of live coding that you won't see anywhere else - live cloud application in 10 minutes from scratch!. We also going to show some sneak peek preview of GridGain 3.0

Hope to you see there!

I Just Bought The Iomega Home Media Network Attached Storage

evans — Wed, 4 Nov 2009 18:24:03 -0500

I bought this product from PC world, after lots of exploration of the NAS drive market, for me it does the job nearly perfectly.

I wanted a drive that I could consolidate all my media to, and access it centrally when I needed to. It also had to be dimensionally small, look good (so as not to distress the missus) and run quietly. It ticks all boxes here then. It also had to fulfil my future requirement to stream media directly to the TV, which it will do no problems, via a suitable games console or any media streaming player.

It plugged straight into my BT home hub, and the drives were viewable instantly under the Network link. The drive is formatted so that there are 5-6 differing folders, these can be mapped straight to 'my computer' or you can install the software which will undertake this process for you.

Network speed is pretty good, an 800 megabyte file transferred in under 5 minutesish. Plenty quick enough to stream video from it. It does have back up software included, however I would always perform manual backups, so am not interested in this.

It is quite easy to configure, with multiple configurable options I don't understand, or require even the LED's brightness can be adjusted.

The only item that seems to be missing is a power saving function, as the drive does not seem to auto spin down when not in use. This does not bother me, as I have no problem with using the power switch on the unit. Hopefully this will be addressed in future firmware upgrades, unless I am a dimwit who has yet to learn how to utilise this function? If so please let me know.

Viewaframework 1.0.0 Swing framework

Mario García — Wed, 4 Nov 2009 11:02:23 -0500

After a couple of weeks I have published Viewa. Viewa is a java Swing application framework (nothing to do with the JSR296).

I made Viewa as an experimental project for making Swing applications in an easy way. It is still a very little project and it lacks a lot of things but I think it could be useful for somebody trying to do a Swing application from the beginning.

Viewa is hosted at sourceforge http://viewaframework.sourceforge.net . There is a getting started tutorial.

How to resolve conflict of DWR and jQuery

Serkan Guler — Wed, 4 Nov 2009 07:44:59 -0500

http://serkanguler.info/2009/11/04/how-to-resolve-conflicts-of-dwr-and-jquery/

Back to being a Java blog

Rick Hightower — Tue, 3 Nov 2009 19:12:40 -0500

I am working at a company where I get moved between Java web, Python daemons, Python web and C/GTK.

It seems I will be working on a classic Java/Spring MVC web application for the next few months (Dec 31st).

I am installing the VMWare developer instance now.... (with all the developer tools preconfigured for developing the Java/Spring MVC/JBoss stack).

Django and JQuery: Integrating Django and JQuery Autocomplete plugin

Rick Hightower — Tue, 3 Nov 2009 17:28:30 -0500

I have many what would otherwise be very large select input fields. I need some autocomplete support. I decided on JQuery (http://jquery.com) and its autocomplete plugin (http://docs.jquery.com/Plugins/Autocomplete).

This assumes that you have at least gone through the Django tutorial.

To support this I added the following to settings.py of the django instance.

JQUERY_ROOT = '/home/csmadmin/tools/jquery'

The above has where I downloaded and installed JQuery and the JQuery autocomplete plugin.

Then to serve up these files I add the following to urls.py

(r'^js/(?P<path>.*)$', 'django.views.static.serve',
        {'document_root': settings.JQUERY_ROOT}),

Then I adapted the example they gave in the JQuery autocomplete plugin to work with django.

I created two views to server up plain text and json respectively (for months and emails).

months_data = """
January
Feburary
March
April
May
June
July
August
September
October
November
December
"""
 
emails_data = [
    { "name": "Peter Pan", "to": "peter@pan.de" },
    { "name": "Molly", "to": "molly@yahoo.com" },
    { "name": "Forneria Marconi", "to": "live@japan.jp" },
    { "name": "Master <em>Sync</em>", "to": "205bw@samsung.com" },
    { "name": "Dr. <strong>Tech</strong> de Log", "to": "g15@logitech.com" },
    { "name": "Don Corleone", "to": "don@vegas.com" },
    { "name": "Mc Chick", "to": "info@donalds.org" },
    { "name": "Donnie Darko", "to": "dd@timeshift.info" },
    { "name": "Quake The Net", "to": "webmaster@quakenet.org" },
    { "name": "Dr. Write", "to": "write@writable.com" }
    ];
 
 
def months(request):
    query = request.GET.get("q", "")
 
    if len(query) == 0 or query[0] == " " :
        data = months_data
    else:
        month_list = months_data.split('\n')
        print "ml " + `month_list`
        print "qy " + `query`
        data = "\n".join([month for month in month_list
                          if month.lower().find(query) != -1])
        print "dt " + `data`
    return HttpResponse(data, mimetype = 'text/plain')
 
 
def emails(request):
    query = request.GET.get("q", "")
 
    if len(query) == 0 or query[0] == " " :
        json = simplejson.dumps(emails_data)
    else:
        json = simplejson.dumps([email for email in emails_data
                                 if email['name'].find(query) != -1
                                 or email['to'].find(query) != -1 ])
    return HttpResponse(json, mimetype = 'application/json')

Don't you just love list comprehensions. :) The above filters the lists based on the query string that was sent(with a HTTP get).

Then we just change the index.html to handle the JSON request. I changed the base.html template to support JavaScript as follows:

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"
            "http://www.w3.org/TR/REC-html40/loose.dtd">
 
{% autoescape on %}
<html>
 
<head>
    <link rel="stylesheet" href="style.css" />
</head>
 
<head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
    <title>{% block title %}My amazing site{% endblock %}</title>
 
    {% block css %}
    <link rel="stylesheet" type="text/css" href="/css/site.css" />
    {% endblock %}
 
 
    {% block javascript %}
    <script type="text/javascript" src="/js/jquery.js"></script>
    {% endblock %}
 
</head>
 
<body>
<h1>Submental Mutterings</h1>
<div id="sitenav">
    {% block sitenav %}
    <h4>our website</h4>
    <a href="mutter01.html">asdf</a>
    <a href="mutter02.html">humor</a>
    <a href="mutter03.html">asdf</a>
    <a href="mutter04.html">asdf</a>
    <a href="mutter05.html">asf</a>
    <a href="mutter06.html">asdf</a>
    <a href="mutter07.html">asdf</a>
    <a href="mutter08.html">asdf</a>
    {% endblock %}
 
</div>
 
<div id="sidebar">
    {% block sidebar %}
    <h4>Other Mutters</h4>
    <a href="mutter01.html">13 September 2001</a>
    <a href="mutter02.html">24 September 2001</a>
    <a href="mutter03.html">04 October 2001</a>
    <a href="mutter04.html">19 October 2001</a>
    <a href="mutter05.html">27 October 2001</a>
    <a href="mutter06.html">31 October 2001</a>
    <a href="mutter07.html">08 November 2001</a>
    <a href="mutter08.html">13 November 2001</a>
    {% endblock %}
</div>
 
<div id="content">
        {% block content %}
 
<h3>27 October 2001</h3>
<p>
I sit here in a lone island of light, tapping away, trying to stay awake.  I'm the only one who isn't trying to grab some sleep as we speed east.  Three hours knocked out of the night without even trying, thanks to the time zones we have to cross; everyone's hoping to spend as much of the four hours we'll be aloft as possible asleep.  Except me.  I'm doing my damndest to keep myself alert so that when I get home, I'll have been awake all night, just like my wife back in Cleveland.  So we can go to sleep together, be together for the first time in a week.
</p>
<p>
It helps that my body thinks it's only 2:14am, despite the fact that it's 5:14am where we're headed, where it's home.  I've been staying up late all week anyway.  Lying alone in my hotel room bed is like standing in a deserted concert hall, hollow and unused despite my presence.  There isn't enough of me to fill the space available.  The surface of the mattress is too flat because I'm the only one deforming it.  I kept lying there in my lone valley, trying to distract my hindbrain with late-night Comedy Central, where strangely enough they're showing <cite>Saturday Night Live</cite> reruns every night of the week.
</p>
<p>
But now the week is over, and I'm slipping eastward at 600 miles an hour and six miles aloft, arrowing through the sky in a small pool of light.  Power bleeds slowly out of my laptop battery, music pours through the headphones into my ears, and the thin air rushes past the skin of the plane.  I keep glancing out the window as though something will be there.
</p>
<p>
Soon I will see my wife again, will be home.  Through the exhaustion I can feel happiness, deep and solid.
</p>
    {% endblock %}
</div>
 
 
</body>
</html>
{% endautoescape %}

Notice that I have a javascript block and a css block.

Then we install the django autocomplete feature as follows:

{% extends "patientmgmt/base.html" %}
 
{% block title %}PM Index{% endblock %}
 
{% block javascript %}
    <script type="text/javascript" src="/js/jquery.js"></script>
 
<script type="text/javascript" src="/js/jquery.js"></script>
<script type='text/javascript' src='/js/jquery.bgiframe.min.js'></script>
<script type='text/javascript' src='/js/jquery.ajaxQueue.js'></script>
<script type='text/javascript' src='/js/thickbox-compressed.js'></script>
<script type='text/javascript' src='/js/plugins/jquery.autocomplete.js'></script>
<link rel="stylesheet" type="text/css" href="/js/plugins/jquery.autocomplete.css" />
<link rel="stylesheet" type="text/css" href="/js/jsthickbox.css" />
 
 
<script type='text/javascript'>
$().ready(function() {
 
	$("#month").autocomplete("json/months", {
		minChars: 0,
		max: 12,
		autoFill: true,
		mustMatch: true,
		matchContains: false,
		scrollHeight: 240,
	});
 
 
	function formatEmail(mail) {
        return mail.name + " &lt;" + mail.to + "&gt";
    }
	$("#email").autocomplete("json/emails", {
        dataType: "json",
        parse: function(data) {
            return $.map(data, function(row) {
                return {
                    data: row,
                    value: row.name,
                    result: row.name + " <" + row.to + ">"
                }
            });
        },
        formatItem: function(item) {
            return formatEmail(item);
        }
    });
 
 
 });
</script>
 
{% endblock %}
 
 
{% block content %}
 
 
<p>Here is some content
<input type="text" id="month" />
<input type="text" id="email" />
 
</p>
{% endblock %}

The above works. There might be better ways to combine Django and JQuery. In fact, you can almost count on it. I was able to get the above working fairly quickly and it is certainly one way to combine a JavaScript library with Django site.

Please feel free to comment on this and let me know where I can improve.

Hazelcast Meetings in New York

Talip Ozturk — Tue, 3 Nov 2009 12:55:05 -0500

It is very likely that it will be so cold in New York but I am sure the discussions in Hazelcast meetings will help!

If you or your friends in your company are interested in hearing more about Hazelcast, let me know (talip at hazelcast dot com). I will be in town from November 9th to 17th.

FEST-Assert 1.2: Fluent Interface for Assertions

Alex Ruiz — Tue, 3 Nov 2009 07:55:36 -0500

We are proud to announce that FEST-Assert 1.2 is out!

FEST-Assert is an "assertThat" library that provides a fluent interface for writing assertions. Its main goal is to improve test code readability and make maintenance of tests easier.

Example:

int removed = employees.removeFired();
assertThat(removed).isZero();
 
List<Employee> newEmployees = employees.hired(TODAY);
assertThat(newEmployees).hasSize(6)
                        .contains(frodo, sam);
 
String[] newHires = employees.newHiresNames();
assertThat(newHires).containsOnly("Gandalf", "Arwen", "Gimli"); 

assertThat(yoda).isInstanceOf(Jedi.class)
                .isEqualTo(foundJedi)
                .isNotEqualTo(foundSith);

Complete blog post can be found here.

Refactored GoogleWiki to Markdown to support MediaWiki

Rick Hightower — Tue, 3 Nov 2009 01:36:19 -0500

Refactored GoogleWiki to Markdown to support MediaWiki...

This code adds the ability to convert from Markdown text (the subset I am interested in) into Google Wiki format and now supports media wiki format.

FYI: "Markdown is a text-to-HTML conversion tool for web writers. Markdown allows you to write using an easy-to-read, easy-to-write plain text format, then convert it to structurally valid XHTML (or HTML)." See markdown and python-markdown for more details.

This python script coverts markdown text into Google Wiki text or media wiki format.

Using eval and polymorphism to easily support multiple Wiki's inputs.

JRoller exceptions out when I paste this in here so here is the original link.

http://code.google.com/p/courseconstructor/wiki/RefactoredGoogleWikiToMarkDownToSupportMediaWiki

What's new in JSF 2.0?

Ted Goddard — Mon, 2 Nov 2009 14:40:28 -0500

Kito Mann and I will be giving a webinar this Wednesday, Nov 4 on JavaServer Faces 2.0. WebMC has been ported to JSF 2.0, so I'll be using a few lessons from that easy porting exercise to explain some of the differences between JSF 1.2 and JSF 2.0. To prepare for the webinar, please download the ICEfaces 2.0 Alpha and JSF 2.0 and have your questions and comments ready.

Now that JSF 2.0 is ready, when do you plan on using it in production? As a start, we should really deploy some of the ICEfaces 2.0 demos on ICEfaces.org.

Have you tried Spring Roo?

Serkan Guler — Mon, 2 Nov 2009 07:01:00 -0500

http://serkanguler.info/2009/11/02/have-you-tried-spring-roo/

Dynamic Images with JSF

Cagatay Civici — Mon, 2 Nov 2009 04:34:35 -0500

http://cagataycivici.wordpress.com/2009/11/02/dynamic-images-with-jsf/

GWT and Jetpack are going to shake up the web-based (RIA) desktop application landscape

Dominique De Vito — Mon, 2 Nov 2009 01:31:09 -0500

Jetpack is a Mozilla project to leverage W3C technologies (HTML/CSS/AJAX), instead of XUL, to build much more easily Firefox add-ons. Indeed, the Guiding Principles for Jetpack tell Jetpack should be web-based:

""
* HTML, Javascript, and CSS will be the only tools required
* Have the full power of the open Web: AJAX & multimedia capabilities including <canvas>, <audio> and <video>
* Streamed & accessible via a URL, just like the rest of the Web
* Debugged in the browser without restarting using common Web development tools like Firebug
""
Jetpack should be also extensible: "Architecture will allow for the easy inclusion of reviewed & versioned third-party toolkits (e.g. jQuery, Dojo, etc.) and API libraries (e.g. Twitter, Delicious, Google Maps, etc.)".

That's funny because I wrote in July 2006, a post Revisiting XHTML as a base (?) for XUL-like programming that could be summarized like the following:
""
(1) XHTML is better than what you may think about,
(2) Sure, XHTML could be used on Java client side (and follows some uses you may be not aware of),
(3) XHTML could work, in fact, quite similar technologies are working and popular, but not standard,
(4) I think XHTML is not really about regular web page creation, but more about XML-based GUIs: that's why XHTML has not really taken off, new rich client W3C activities may be another good chance for XHTML. I hope it will.
Conclusion: (a) XML-based GUI technology is scattered into different ways/tools, XHTML may be a good choice for gathering effort and to promote such technology, (b) XHTML may simplify Java client development as XHTML programming model is common (similar to HTML) and because XHTML Java support and use are growing.
""

What are the relationships with GWT ?

GWT has popularized the idea of building web sites and downloading them on the client-side, into one single chunk (that may be split into different files), so that a browser is interacting with the server-side through GWT-RPC calls on http. So, GWT provides 2 interesting features:
* web applications are packaged into a single chunk to be downloaded once on the client-side.
* business service customers and providers are clearly separated (the former are on the client-side, with the GUI part, downloaded once, and the latter, the providers, are on the server-side).

Idea No 1: why not using GWT to develop Jetpack extension ?

Extension builders may consume GWT productions:
* GWT outputs on the client-side leverage technologies (HTML/CSS/AJAX), and Jetpack too.
* GWT client-side is packaged into a single chunk (downloaded once), Jetpack needs to be packaged similarly.

So, they fit quite well.

Let's face the different cases for developing a Jetpack extension:
(a) online extension: this is the ideal case, as the GWT client-side output may be quite almost ready to be packaged as a Jetpack extension.
(b) offline extension: it depends on the browser's services the extension uses or not.
- if the extension does not plan to use any service (e.g. an extension created to modify web pages), then this case looks like just as hard as (a).
- if the extension plans to use browser's services, then to program this extension with GWT, mock objects implementing these services should be available in Java to program the extension with Java (GWT does already that, providing a DOM class for manipulating DOM with Java programming). Then, during the compilation phase, those mock objects would be replaced with JS objects giving access of the browser's real services.
(c) online/offline extension:
- if no synchronization is involved, this case is (a) + (b).
- if synchronization is needed, technologies like Gears could help here, taking benefit from GWT architecture, more precisely, the clear separation between business service customers and providers.

Idea No 2: an application view is needed within HTML

As said before, a GWT application is available as a single chunk downloaded once on the client-side. But there is no HTML element to tell to the browser that, such a single chunk, while running into the browser, could be recorded/cached as packaged into a Jetpack add-on. Such an HTML element could be useful.

With Jetpack, GWT appears as a dual technology and the race is open between web applications and browser extensions

Before Jetpack, development options were web applications and Java-based applications (respectively, lightweight and heavyweight clients). With Jetpack, a third option appears: heavyweight clients, not Java-based, but browser-based.

With Jetpack, Mozilla provides an Adobe AIR competitor

Adobe AIR aims to build Rich Internet applications (RIA) using Flash, Flex, HTML, or Ajax, that can be deployed as a desktop application. With Jetpack, Mozilla provides to Firefox similar desktop application support. Add GWT, and then, developers get a development toolkit, that fits with Jetpack.

Conclusion !?

GWT and Jetpack are cross-frontier technologies (GWT compiles Java into JS, and Jetpack enables to leverage W3C technologies to create Firefox add-ons). They are great technologies. It would be even greater if, as expected, those technologies are used conjointly.

So, it's very worthwhile to keep an eye on both technologies within the next months (GWT 2.0 development is on-going with M1 recently released, and Jetpack - last version being 0.4 - is planned to be released with Firefox 4 at the end of 2010).

JRoller really doesn't work with Safari

Bob Foster — Sun, 1 Nov 2009 22:09:12 -0500

Three times now it has thrown away my post and refreshed the screen to nothing. I ignored the little alert that said it might not work - because I get the same alert with Google Chrome and it does work - but it's sadly true. Is JRoller basically obsolete?

JRoller really doesn't work with Safari

Bob Foster — Sun, 1 Nov 2009 22:09:12 -0500

JRoller really doesn't work with Safari

Bob Foster — Sun, 1 Nov 2009 22:08:52 -0500

JCP: At The Crossroads

robert burrell donkin — Sun, 1 Nov 2009 01:54:21 -0500

As drawn and analysed by Stephen Colebourne. It's clear now that the JCP must evolve or die. Be interesting to see how Oracle resolves this issue...

Headless image manipulation in Java

Carter Page — Sun, 1 Nov 2009 01:03:35 -0500

Java image manipulation generally revolves around libraries that exist for GUI applications. As such, if you are writing a server application on your desktop, and deploy it to a headless (no monitor) server, you may be unpleasantly surprised to encounter HeadlessException.

In general, there are two simple rules to keep in mind:

Set Java option java.awt.headless to true (ie, -Djava.awt.headless=true)
Use a JPanel instead of a JFrame. You get much of the same functionality, without the VM complaining about the display.

So, this stuff is always more fun with code. So I'll provide a sample for resizing an image, and another sample for cropping an image. Both are safe to run on headless servers.

The image objects used are ImageIcon, which are very simple to work with. For instance, you can construct the object with a image file name, or even a raw byte array for maximum flexibility.

Resize an Image

This resizes an image to an arbitrary size

public static ImageIcon resizeImage(ImageIcon imageIcon,
                                    int width, int height)
{
    ImageIcon resizedImageIcon;
    resizedImageIcon = new ImageIcon(imageIcon.getImage()
        .getScaledInstance(width, height, java.awt.Image.SCALE_SMOOTH));
    return resizedImageIcon;
}

Fit an Image

Most times an arbitrary resize isn't very practical since it skews your image without ratio calculations beforehand. Here I've created a method that packages those ratio calculations into the method. My standard use case is to fit an image of any size within a "frame" that contains an image (such as a div of set size on a webapp). This example resizes an ImageIcon object. You give it a frame with a maximum width and height, and it will shrink or maximize the image to fit within that frame.

public static ImageIcon resizeImageToFit(ImageIcon imageIcon,
                                         int maxWidth, int maxHeight)
{
    ImageIcon resizedImageIcon;
    float iconRatio = (float)imageIcon.getIconHeight()
        / (float)imageIcon.getIconWidth();
    float fitToRatio = (float)maxHeight / (float)maxWidth;
    if (iconRatio > fitToRatio)
        resizedImageIcon = new ImageIcon(imageIcon.getImage()
            .getScaledInstance(-1, maxHeight,
                java.awt.Image.SCALE_SMOOTH));
    else
        resizedImageIcon = new ImageIcon(imageIcon.getImage()
            .getScaledInstance(maxWidth, -1,
                java.awt.Image.SCALE_SMOOTH));
    return resizedImageIcon;
}

Crop an Image

Self explanatory.

public static ImageIcon cropImage(ImageIcon icon, int top, int left,
                                  int width, int height)
{
    JPanel jpanel = new JPanel();
    return new ImageIcon(jpanel.createImage(
        new FilteredImageSource(icon.getImage().getSource(),
            new CropImageFilter(left, top, width, height))));
}

Element of conception: name important external resources (or cross-frontier resources)

Dominique De Vito — Sat, 31 Oct 2009 12:38:26 -0400

Recently, I found into server logs that the Java EE server had closed itself a leaking connection. Unfortunately, no name was associated with that announcement, that is, no connection name was provided. So, I had no simple way to go up to the source of this leak.

One should be able to name a JDBC connection (and transactions too)

It's very unfortunate not being able to name a JDBC connection.

Well, the getConnection() methods of the DataSource class have no 'name' argument:

public interface DataSource extends CommonDataSource, Wrapper {

  /**
   * Attempts to establish a connection with the data source that
   * this DataSource object represents.
   */
  Connection getConnection() throws SQLException;
      
  /**
   * Attempts to establish a connection with the data source that
   * this DataSource object represents.
   */
  Connection getConnection(String username, String password) 
    throws SQLException;
}

Futhermore, the Connection class has no setName() method.

It smells like an anti-pattern.

It would be worthwhile, for example, if a Java EE server was naming used connections according to some annotation's value, or, by default, using the name of the called EJB business method using this connection.

Notes on Brion Vibber's recent Wikipedia Talk

Michael Levin — Sat, 31 Oct 2009 04:47:17 -0400

Please join this developing discussion on CodeTown for details about Brion's recent talk: http://www.codetown.us/group/orlandojug/forum/topics/brion-vibber-wikipedia-cto

ICEfaces 2.0 Alpha now available

Ted Goddard — Fri, 30 Oct 2009 19:41:41 -0400

ICEfaces 2.0 is now available for download. To get started, you can download either the source bundle (containing all necessary libraries, but only 10MB) or the binary (containing pre-built demo applications that you can deploy on Tomcat 6).

ICEfaces 2.0 provides two main components: "core" and "compat". The ICEfaces core provides Direct-to-DOM rendering functionality and Ajax Push for JSF 2.0. For some applications, the core is all you need (see the new "auction" demo made purely with the standard JSF components and no ajax-specific markup; also, no faces-config.xml, all beans are injected via annotations). Push features can be added with two lines of code via org.icefaces.application.PushRenderer (this is just a repackaging of the familiar SessionRenderer API).

Of course, most applications also need a rich set of components. We will be building a second-generation set of components (as Ken mentions), but at the moment we thought you would enjoy continuing to use the familiar ICEfaces components (such as, when porting your existing applications to JSF 2.0). This is provided by the "compat" libraries in ICEfaces 2.0. To see this at work, try out the "auctionMonitor" and "component-showcase" demos. The venerable auctionMonitor demo (from JavaOne 2005) only required a few hours of work to port to JSF 2.0 via the compat library. To build:

Unzip the source bundle
Invoke ant world in the icefaces directory (ant all builds only the core)
Deploy the .war files from samples/ and compat/samples/

You may also be interested in deploying ICEfaces 2.0 to a Servlet 3.0 application server (such as GlassFish v3). Servlet 3.0 is automatically detected, and gives you the scalability benefits of the new standard NIO-based asynchronous request processing.

(Remember, this is an alpha release, so you are bound to encounter a bug here and there; but we hope you have fun trying it out. Are there things missing from the compat APIs? Was your application easy to port? Let us know what you would like to see in the next alpha (or beta) release.)

Escaping HTML Entities in Python

evans — Fri, 30 Oct 2009 17:35:18 -0400

One of the cool things about working with Python is how easy it is to accomplish most tasks without having to roll your own utility module or program. I was working again on a slightly more advanced tool for one of my affiliate programmes last night, and needed to escape some HTML entities in my XML document before they were parsed by my XSLT program. I couldn't get it to do what I wanted because they were not properly escaped.

Unless I escaped the HTML Entities, the XML would not validate. To solve this problem, I had to resort to Python's escape() functions which are available both in CGI module or the Saxutils module. Here's an example of how to use them.

>>> import cgi
>>> cgi.escape('<&>')

>>>

>>> from xml.sax import saxutils
>>> saxutils.escape("&")

>>>

Depending on which one you use, they both do the same thing - making it easy to escape HTMLs. In fact, it also helps with URL encoding in instances where the '&' is confusing your application.

Sweb and styles (part 3)

Jean Lazarou — Fri, 30 Oct 2009 17:08:01 -0400

Following part 2, part 3 presents the implementation of the GUI part of the application.

The implementation uses the Sweb approach from Swiby. Sweb presents GUIs in a browser-like container. Swiby scripts loaded by Sweb run in the context of the container, they do not create it. Sweb is loading initially and creates the expected context.

The implementation of the GUI may look a bit odd because it uses the Game class implemented in part 2. The Game implementation was pretty simple.

Create the visual part

Let’s start by looking at the way the code builds the GUI layout, that looks like:

(Code explanation comes after.)

src/hidden_number_ui.rb

01 title "Hidden Number Game"
02 
03 width 640
04 height 380
05 
06 panel = content(:layout => :flow) {
07 
08   use_styles 'hidden_styles.rb'
09   
10   form {
11   
12     section "Game"
13       input "Your Guess", "", :name => :guess
14 
15       num_pad = [7, 8, 9] +
16                 [4, 5, 6] +
17                 [1, 2, 3] +
18                 [0      ]
19                         
20       grid(:columns => 3) {
21         num_pad.each { |key|
22           button key.to_s, :name => key.to_s.to_sym, :style_class => :numeric_pad
23         }
24 
25         column :span => 2
26           button 'Clear', :name => :clear, :style_class => :numeric_pad
27       }
28       
29       command "Propose", :name => :submit
30       
31     section "Control"
32       combo "Range", [], :name =>:number_of_digits
33       button 'Start', :name => :start
34       button "Restart", :name => :restart
35       button "Exit", :name => :exit_application
36 
37     next_row
38       label '', :name => :message
39       progress :horizontal, Game::MAX_GUESSES, :tries_left, :value => Game::MAX_GUESSES
40       
41   }
42 
43 }
44 
45 $context.start

Line 1 sets the title to appear in the title bar. Lines 3 and 4 force the initial window size.

The code between lines 6 and 43 defines the content displayed inside the Sweb container. It creates a top panel, a form, assigns th panel to the panel variable and sets the layout of the Sweb container to flow layout. The layout is responsible for placing the components. The flow layout is known to Java/Swing developers, it places every component from left to right, each one next to the other depending on their desired size.

Line 8 sets the styles to use to setup the fonts and colors, provided by a file named hidden_styles.rb (see part 1).

The code assigns a name to all the components see code like: :name => :guess.

Each section starts a new parent container. Each section contains all the following components, they are children of the section container. The next_row directive (line 37) forces to place (layout) the following components on a next row, Swiby creates a new section implicitly.

The code defines a form. Therefore, even if both button and command create button components, Swiby places the command buttons at the bottom of their container (section) where they appear.

Line 21 creates all the buttons with the digits, laid out like a numeric pad. The Ruby code format (lines 15-18) tries to make more visible what part of the final form it declares.

The column layout directive (line 25) is available because the surrounding container is a grid (line 20 starts a container with a grid layout – every component is taking place in a cell and all cells have the same size). The span attribute, like in HTML, says that next components consumes two-columns width.

Line 45 asks Sweb to display the container and somehow to start the sweb-application.

To make it run we need to require some files.

src/hidden_number_ui.rb

01 require 'swiby/layout/grid'
02 require 'swiby/layout/stacked'
03 
04 require 'swiby/mvc/text'
05 require 'swiby/mvc/combo'
06 require 'swiby/mvc/button'
07 require 'swiby/mvc/label'
08 require 'swiby/mvc/progress_bar'
09 
10 require 'hidden_number'

Running the script would result in displaying the window without any behavior (clicking the buttons or playing with the combo-box has no effect).

Give life to the window

As the require statements show, we use the (Swiby) MVC version of the components. We need to plug a controller that is going to give life to the view/form.

We start by implementing the behavior of the exit button.

src/hidden_number_ui.rb

01 class PlayerController
02   
03   def exit_application
04     exit
05   end
06   
07 end

The exit button name is exit_application, if the controller implements a method with the same name, Swiby calls the method when the user clicks the button.

Now, we need to link the controller to the view, stored in the panel variable.

src/hidden_number_ui.rb

01 ViewDefinition.bind_controller panel, PlayerController.new

We can use (Swiby) MVC in different ways. Here, Sweb is creating the form and we must make the binding explicitly between the view and the controller.

With the controller binding in place, the exit button is working.

Only the exit button is enabled. As we now plugged the controller, Swiby checks for every button if a handler method does exist. Swiby only enables the ones having a handler.

Let’s now go through implementing more functionality.

Step 1: the startup message

At startup the GUI should tell the player that the game is not started. We are going to use the label component named message .

src/hidden_number_ui.rb

01   def initialize
02     @message = 'Game not started'
03   end
04

As the field name is message, we just need to provide a getter and Swiby is going to use it.

src/hidden_number_ui.rb

01   attr_accessor :message

And the message appears.

Step 2: plugging the game object

In the initialize method we create a new instance of the Game class, ready to start the play.

src/hidden_number_ui.rb

01     @game = Game.new

Step 3: starting the game

Some tips, some tricks

interface?

The controller does not derive from any class, we could derive from the TextPlayer defined in the previous part. The benefit would have been that we could delay the writing of the callbacks and forgetting one of them would produce the output to the console. But the two classes are more like implementing the same interface than being of the same kind… See my entry: OO as a technique, OO as a concept.

naming convention

The naming convention presented, used by Swiby to find handlers in the controller, is not static. Swiby uses by default a name provider available through a property named Swiby.default_method_naming_provider, it defaults to an instance of the Swiby::MethodNamingProvider class. Change the property to plug a custom provider.

If you want to provide a custom version, don’t forget that Swiby uses the open-class character of Ruby to extend it.

To enable the start button a handler named start is enough but, when the game is running, it should be disabled. Therefore, we also add a may_start? method that Swiby uses to enable/disable the button.

src/hidden_number_ui.rb

01   def start
02     
03     @message = ' '
04 
05     @entry = ''
06     @started = true
07     @tries_left = Game::MAX_GUESSES
08 
09     Thread.new do
10       
11       begin
12         @game.start self
13       rescue => e
14         puts "Error in thread => #{e}"
15         trace = e.backtrace
16         puts trace.join("\n")
17       end
18 
19     end
20     
21   end
22   
23   def may_start?
24     not @started
25   end

Line 3 clears the message attribute to clear the message the player sees. Line 5 empties the entry attribute used to store the data the player enters. Line 6 sets the started flag used at line 24 to enable/disable the start button. The tries_left attributes stores the number of guesses made so far.

Previous part showed the game object implementation, it ran in a simple console. The game was an endless loop querying the player for actions. Reusing the same implementation here requires to launch the game in its own thread (line 9) and pass the controller as a player. The game object starts its endless loop at line 12.

In the initialize method we initialize the started flag to false.

src/hidden_number_ui.rb

01     @started = false

The player controller must provide the propose method.

src/hidden_number_ui.rb

01   def propose
02     
03     @command = nil
04     
05     sleep 0.1 until @command
06     
07     @command
08     
09   end

So, the game object runs on a different thread, when it needs input from the player it calls the propose method. The two threads (the GUI thread and the game) share the command attribute. Sleeping in the propose method prevents the game thread from waiting in an active loop.

Step 4: get game compliant

The Game class also expects other methods from the player.

src/hidden_number_ui.rb

01   def higher number_of_guesses
02     @tries_left = Game::MAX_GUESSES - number_of_guesses
03     self.message = "<html>Guess #{number_of_guesses} with value #{@guess} was too <b>low</b>"
04   end
05   def lower number_of_guesses
06     @tries_left = Game::MAX_GUESSES - number_of_guesses
07     self.message = "<html>Guess #{number_of_guesses} with value #{@guess} was too <b>high</b>"
08   end
09   def win number_of_guesses
10     @started = false
11     self.message = "<html>You win!"
12   end
13   def lose hidden_number, number_of_guesses
14     @started = false
15     @tries_left = 0
16     self.message = "<html>You lose, the hidden number was: #{hidden_number}"
17   end

All the methods create a HTML string used by the underlying JLabel class to format the display.

Step 5: roughly and quickly

Let’s have a quick look at the remainder of the controller implementation.

src/hidden_number_ui.rb

01   def guess
02     @entry
03   end
04   
05   def guess= value
06     @entry = value
07   end
08   
09   def tries_left
10     @tries_left
11   end
12   
13   def formated_tries_left
14     
15     if not @started
16       ''
17     elsif @tries_left == 1
18       'last try'
19     else
20      "#{@tries_left} tries left"
21     end
22 
23   end
24     
25   def clear
26     @entry = ''
27   end
28   
29   def may_clear?
30     @started
31   end
32         
33   def submit
34     @command = @entry.to_i
35     @guess = @entry
36     @entry = ''
37   end
38   
39   def may_submit?
40     not may_start?
41   end
42   
43   def restart
44     @message = ' '
45     @command = :restart
46     @tries_left = Game::MAX_GUESSES
47   end
48   
49   def may_restart?
50     not may_start?
51   end
52   
53   def list_of_number_of_digits
54     
55     @list_provided = true
56     
57     [(1..20), (1..30), (1..50), (20..80), (1..90), (1..100), (1..500)]
58     
59   end
60   
61   def list_of_number_of_digits_changed?
62     not @list_provided
63   end
64   
65   def number_of_digits
66     @game.number_of_digits
67   end
68   
69   def number_of_digits= x
70     @game.digits_range = x
71   end
72

Lines 1-7 connect the guess attribute to the input field (with the same name), using att_accessor also works.

Lines 9-23 connect the number of tries the player still has with the progress bar. The progress bar displays the text provided by formated_tries_left only if such a method is present (convention is the field name prefixed by formated_).

Lines 25-51 implement the handlers for the clear, propose (named submit) and restart buttons.

Lines 53-71 connect the controller to the combobox that the player uses to choose from several type of hidden numbers. Swiby makes a difference between the value of the combobox and the available values to choose from. The (selected) value follows the attribute getter/setter paradigm. For the data, that appear in the list, Swiby tries to find a method starting with list_of_ followed by the component’s name (line 53). It also checks if the list content changes using the method list_of_number_of_digits_changed?.

The code uses the list_provided attribute to ensure that the combobox gets filled only once.

Step 6: message, the last touch

At this point, running the game would not work as expected, because the methods lower and higher, called by the game object, change the message but Swiby is not aware of the changes. Swiby provides a way to help in such a case, the bindable directive declares attributes of the controller that Swiby should check for changes.

src/hidden_number_ui.rb

01   bindable :message

Step 7: the numeric pad

To implement the handlers for the numeric pad we use Ruby’s facility for dynamically defining methods in a class.

src/hidden_number_ui.rb

01   10.times do |i|
02     
03     name = i.to_s
04     
05     define_method(name.to_sym) {
06       @entry += name
07     }
08     
09     define_method("may_#{name}?".to_sym) {
10       @started
11     }
12     
13   end

Line 5 defines the getter for each numeric button and line 9 defines the enable/disable method.

Style reload at runtime

Using the reload button from the Sweb container has no effect because it does not know what reload does mean for the current window. We must provide it.

src/hidden_number_ui.rb

01 $context.reload_hook do
02   
03   s = load_styles("hidden_styles.rb")
04   $context.apply_styles s
05   
06 end

Line 3 loads the styles file and line 4 applies the styles.

Conclusion

The Swiby MVC does not pretend to follow some universal definition of the MVC pattern, if any. It tries to make easy to separate the view and the controller code. The basic idea is not to make it possible to reuse a view with different controllers or a controller with different views. It gives some framework to make GUI development easier, maybe simpler and make the code more testable.

Sweb and styles (part 2)

Jean Lazarou — Fri, 30 Oct 2009 17:07:48 -0400

Part 1 presented how to setup styles for a simple application, here we are going to present the implementation of the application. The application is a hidden number game. We split the implementation presentation in two parts to make them shorter.

Implementing such a game is rather simple as long as we don’t have to think about a graphical interface (GUI). This part starts by an implementation without GUI. Next part presents the GUI version based on the non-GUI version leading to a simpler work.

Introducing the game

The game is very basic.When it starts a random number is drawn, then the player makes a guess. If the guess equals the drawn number, the player wins, otherwise the player can make another guess. When the guess is wrong the player receives information about the hidden number: lower or higher than the last guess.

The game implementation

The game lives as a Game object. Once the game starts it enters an endless loop, when the player finds the number or consumes the maximum guesses the game ends.

Next sections present the specifications for the Game class.

Writing specifications is a test/development approach known as behavior-driven development. rspec is a framework providing behavior-driven development for Ruby and used here.

Step 1: stopping the game

The player should be able to stop the game at any moment.

src/hidden_number_spec.rb

01 require 'hidden_number'

02 
03 describe "Game" do
04   
05   before :each do  
06     @game = Game.new
07   end
08   
09   it "should allow user to stop the game" do
10     @game.start stopping_player
11   end
12 end

The above description requires creating a game object (line 6) before applying every specification.

The specification starts the game (line 10) passing a mock player created by the stopping_player helper method.

src/hidden_number_spec.rb

01   def stopping_player
02     
03     player = mock "Player"
04     
05     player.should_receive(:propose).and_return(:stop)
06     player.should_receive(:bye).with(0)
07     
08     player
09     
10   end

The mock object first expects to receive a propose message that would return stop (line 5), then expects to receive the bye message (line 6) with a number of guesses equal to zero as argument.

The Game class, hereafter, implements a basic endless loop (starting at line 11) and processes only the stop command (line 16). The game instance sends a propose message to the player to get some proposal at line 13. When the player stops the game, the game sends a bye message (line 17).

src/hidden_number.rb

01 class Game

02 
03   attr_reader :number_of_guesses
04   
05   def initialize
06     @number_of_guesses = 0
07   end
08   
09   def start player
10   
11     loop do
12       
13       proposal = player.propose
14       
15       case proposal
16         when :stop
17           player.bye @number_of_guesses
18           return
19       
20       end
21         
22     end
23     
24   end
25   
26   
27 end

Here is the result of the run

$>spec  hidden_number_spec.rb
.

Finished in 0.039 seconds

1 example, 0 failures

Step 2: drawing a number

When the game starts it should draw a number with a desired number of digits or in a specific range.

src/hidden_number_spec.rb

01   it "should draw a number with the given number of digits" do
02     
03     @game.number_of_digits = 1
04     
05     @game.start stopping_player
06     
07     @game.hidden_number.should >= 0
08     @game.hidden_number.should <= 9
09     
10   end
11   
12   it "should draw a number within the given range" do
13     
14     @game.digits_range = (78..80)
15     
16     @game.start stopping_player
17     
18     @game.hidden_number.should >= 78
19     @game.hidden_number.should <= 80
20     
21   end

The first specification describes the behavior for a hidden number with one digit and the second for a hidden number ranging from 78 to 80.

Changing the implementation of the Game class to conform to the above specification requires adding more attributes:

src/hidden_number.rb

01 class Game
02 
03   attr_reader :number_of_guesses
04   attr_reader :hidden_number
05   
06   attr_accessor :number_of_digits, :digits_range
07   
08   def initialize
09     @number_of_guesses = 0
10   end
11   
12   def start player
13   
14     loop do
15       
16       proposal = player.propose
17       
18       case proposal
19         when :stop
20           player.bye @number_of_guesses
21           return
22       
23       end
24         
25     end
26     
27   end
28   
29   
30 end

Call a draw_a_number method at the beginning of the loop

src/hidden_number.rb

01 class Game

02 
03   attr_reader :number_of_guesses
04   attr_reader :hidden_number
05   
06   attr_accessor :number_of_digits, :digits_range
07   
08   def initialize
09     @number_of_guesses = 0
10   end
11   
12   def start player
13   
14     draw_a_number
15 
16     loop do
17       
18       proposal = player.propose
19       
20       case proposal
21         when :stop
22           player.bye @number_of_guesses
23           return
24       
25       end
26         
27     end
28     
29   end
30   
31   
32 end

The draw method is using the standard random generator method. It also resets the guesses so far to zero.

src/hidden_number.rb

01 class Game

02 
03   attr_reader :number_of_guesses
04   attr_reader :hidden_number
05   
06   attr_accessor :number_of_digits, :digits_range
07   
08   def initialize
09     @number_of_guesses = 0
10   end
11   
12   def start player
13   
14     draw_a_number
15 
16     loop do
17       
18       proposal = player.propose
19       
20       case proposal
21         when :stop
22           player.bye @number_of_guesses
23           return
24       
25       end
26         
27     end
28     
29   end
30   
31   def draw_a_number
32     
33     if @digits_range
34       @hidden_number = rand(@digits_range.last - @digits_range.first + 1) + @digits_range.first
35     else
36       @hidden_number = rand(10 ** @number_of_digits)
37     end
38   
39     @number_of_guesses = 0
40     
41   end
42   
43 end

Of course, because of the random nature of the draw_a_number method, the above specification is not very reliable. For that reason, we used a one digit hidden number and, for the range specification, we limited the range of the values to 78, 79 and 80.

To make the use of the Game class easier, we can set a default value for the number of digits (2 digits) in the initialize method.

src/hidden_number.rb

01     @number_of_digits = 2

And here is the result of the run

$>spec  hidden_number_spec.rb
...

Finished in 0.04 seconds

3 examples, 0 failures

Step 3: making guesses

After drawing a number, the game needs input from the player. It asks input to the player object. If the value equals the hidden number, the game ends by sending a win message to the user. If the value does not match, the game sends the lower or higher message to the player. If the player fails to find the number after 10 guesses (a constant named MAX_GUESSES ), the game ends by sending a lose message to the user passing the hidden number.

src/hidden_number_spec.rb

01   it "should end if the proposal matches the hidden number" do
02     
03     @game.forced_hidden = 11
04     
05     @game.start winning_player
06     
07   end
08 
09   it "should end if the proposal matches the hidden number and give the right amount of guesses" do
10     
11     @game.forced_hidden = 11
12     
13     @game.start winning_after_3_guesses_player
14     
15   end
16   
17   it "should send lower if proposal is too high" do
18     
19     @game.forced_hidden = 11
20     
21     @game.start one_guess_player(20)
22     
23   end
24   
25   it "should send higher if proposal is too low" do
26     
27     @game.forced_hidden = 11
28     
29     @game.start one_guess_player(5)
30     
31   end
32 
33   it "should end if the player consumes maximum guesses" do
34     
35     @game.forced_hidden = 11
36     
37     @game.start losing_player
38     
39   end

To make them pass, the test code needs to force the hidden number, otherwise the hidden number is a random value (lines 3, 11, 19, 27 and 35). An easy way to force the hidden number is to replace the draw_a_number method, for the current Game instance only, which checks the forced_hidden attribute. Doing so prevents from breaking the previous specifications. Next code is added to the before part.

src/hidden_number_spec.rb

01     def @game.forced_hidden= number
02       @forced_hidden = number 
03     end
04     
05     def @game.draw_a_number
06     
07       super
08         
09       @hidden_number = @forced_hidden if @forced_hidden
10         
11     end

Four more helper methods create the mock players.

src/hidden_number_spec.rb

01   def winning_player
02     
03     player = mock "Player"
04     
05     player.should_receive(:propose).and_return(11)
06     player.should_receive(:win).with(1)
07     
08     player
09         
10   end
11   
12   def winning_after_3_guesses_player
13     
14     player = mock "Player"
15     
16     player.should_receive(:propose).and_return(1, 1, 11)
17     player.should_receive(:higher).exactly(2).times
18     player.should_receive(:win).with(3)
19     
20     player
21         
22   end
23   
24   def one_guess_player guess
25     
26     player = mock "Player"
27     
28     player.should_receive(:propose).and_return(guess, :stop)
29     
30     if guess > 11
31       player.should_receive(:lower)
32     else
33       player.should_receive(:higher)
34     end
35     
36     player.should_receive(:bye).with(1)
37     
38     player
39     
40   end
41   
42   def losing_player
43     
44     player = mock "Player"
45     
46     player.should_receive(:propose).and_return(1,1,1,1,1,1,1,1,1,1)
47     player.should_receive(:higher).exactly(9).times
48     player.should_receive(:lose).with(11, Game::MAX_GUESSES)
49     
50     player
51     
52   end

The mock for the winner is rather simple, it expects to receive the propose message that returns the right number (line 5) and then it receives the win message with the number of guesses (1 here) at line 6.

The mock, for a user that finds the number after some guesses, expects to receive the propose message that returns two invalid numbers then the right one (line 16). It must expect to receive the informative messages (higher at line 17) and the win message with the count of guesses equals to 3.

The mock for the player making one guess expects to receive the propose message twice, the first time it returns the value the specification wants, the second time it returns stop (line 28). It expects to receive the lower or higher message and then the bye message with 1 guess as argument.

Finally, the mock for the losing player expects to receive 10 times the propose message returning always 1 (too low compared to the forced hidden number, 11) at line 46. It also expects to receive 9 times the higher message (line 47), only nine times because at next guess the player loses as expressed at line 48. The lose message receives two arguments: the hidden number and the number of guesses.

And now implement the behavior in the endless loop, another branch in the case statement.

src/hidden_number.rb

01 class Game

02 
03   attr_reader :number_of_guesses
04   attr_reader :hidden_number
05   
06   attr_accessor :number_of_digits, :digits_range
07   
08   def initialize
09     @number_of_guesses = 0
10     @number_of_digits = 2
11   end
12   
13   def start player
14   
15     draw_a_number
16 
17     loop do
18       
19       proposal = player.propose
20       
21       case proposal
22         when :stop
23           player.bye @number_of_guesses
24           return
25 
26         when Integer
27         
28           @number_of_guesses += 1
29           
30           if proposal == @hidden_number
31             
32             player.win @number_of_guesses
33             
34             return
35             
36           else
37             
38             if @number_of_guesses >= MAX_GUESSES
39               
40               player.lose @hidden_number, @number_of_guesses
41               
42               return
43               
44             else  
45               player.higher @number_of_guesses if proposal < @hidden_number
46               player.lower @number_of_guesses if proposal > @hidden_number 
47             end
48           
49           end
50       
51       end
52         
53     end
54     
55   end
56   
57   def draw_a_number
58     
59     if @digits_range
60       @hidden_number = rand(@digits_range.last - @digits_range.first + 1) + @digits_range.first
61     else
62       @hidden_number = rand(10 ** @number_of_digits)
63     end
64   
65     @number_of_guesses = 0
66     
67   end
68   
69 end

Line 38 checks if the player didn’t consume the maximum of guesses, the constant is initialized as

src/hidden_number.rb

01   MAX_GUESSES = 10

Is it fine?

$>spec  hidden_number_spec.rb
........

Finished in 0.045 seconds

8 examples, 0 failures

Step 4: restarting the game

Say the player does not feel comfortable with the current number, he/she should be able to restart the game.

src/hidden_number_spec.rb

01   it "should restart the game on player's demand" do
02     
03     @game.forced_hidden = 11
04     
05     @game.start restart_and_win_player
06     
07   end

The helper method creating the mock, follows. It sets the expectation to receiving propose and return a wrong number, next a restart directive and ends by finding the number (line 5) but the number of guesses sent with the win message (line 7) must be 1, not 2.

src/hidden_number_spec.rb

01   def restart_and_win_player

02     
03     player = mock "Player"
04     
05     player.should_receive(:propose).and_return(1, :restart, 11)
06     player.should_receive(:higher)
07     player.should_receive(:win).with(1)
08     
09     player
10         
11   end

A new branch in the case statement handles the restart directive.

src/hidden_number.rb

01 class Game
02 
03   attr_reader :number_of_guesses
04   attr_reader :hidden_number
05   
06   attr_accessor :number_of_digits, :digits_range
07 
08   MAX_GUESSES = 10
09   
10   def initialize
11     @number_of_guesses = 0
12     @number_of_digits = 2
13   end
14   
15   def start player
16   
17     draw_a_number
18 
19     loop do
20       
21       proposal = player.propose
22       
23       case proposal
24         when :stop
25           player.bye @number_of_guesses
26           return
27 
28         when Integer
29         
30           @number_of_guesses += 1
31           
32           if proposal == @hidden_number
33             
34             player.win @number_of_guesses
35             
36             return
37             
38           else
39             
40             if @number_of_guesses >= MAX_GUESSES
41               
42               player.lose @hidden_number, @number_of_guesses
43               
44               return
45               
46             else  
47               player.higher @number_of_guesses if proposal < @hidden_number
48               player.lower @number_of_guesses if proposal > @hidden_number 
49             end
50           
51           end
52 
53         when :restart
54           draw_a_number
55       
56       end
57         
58     end
59     
60   end
61   
62   def draw_a_number
63     
64     if @digits_range
65       @hidden_number = rand(@digits_range.last - @digits_range.first + 1) + @digits_range.first
66     else
67       @hidden_number = rand(10 ** @number_of_digits)
68     end
69   
70     @number_of_guesses = 0
71     
72   end
73   
74 end

$>spec  hidden_number_spec.rb
.........

Finished in 0.044 seconds

9 examples, 0 failures

Now the Game class is ready.

Step 5: user interaction

Making the game usable with a terminal interface is simple with the presented Game class.

src/hidden_number.rb

01 class TextPlayer
02   
03   def propose
04     
05     loop do
06       
07       puts "Enter a guess or 'exit'"
08       print "Your guess: "
09       
10       input = gets.chomp
11       
12       case input 
13         when /^[\d]*$/
14           return input.to_i
15         when 'exit'
16           return :stop
17         else
18           unknown input
19       end
20 
21     end
22   
23   end
24   
25   def higher number_of_guesses
26     puts "Higher"
27   end
28   def lower number_of_guesses
29     puts "Lower"
30   end
31   def bye number_of_guesses
32     puts "Bye"
33   end
34   def win number_of_guesses
35     puts "You win"
36   end
37   def lose hidden_number, number_of_guesses
38     puts "You lose, the hidden number was: #{hidden_number}"
39   end
40 
41   def unknown command
42     puts "Unknown command: #{command}"
43   end
44   
45 end

The propose method is the longer method but very easy. It prints a prompt and waits for the user input. Depending on the type of the input value it returns:

an integer (line 13)
the stop directive (line 15), actually the user asks to exit the application

The method loops asking for user input as long as the user does not type any of the above options, providing an error message (line 18).

To play the game, run the hidden_number.rb script, but it needs a last thing: start the game.

src/hidden_number.rb

01 if $0 == __FILE__
02   Game.new.start TextPlayer.new
03 end

Here is an example of a game:

$>ruby  hidden_number.rb
Enter a guess or 'exit'
Your guess: 10
Higher
Enter a guess or 'exit'
Your guess: 60
Lower
Enter a guess or 'exit'
Your guess: 40
Lower
Enter a guess or 'exit'
Your guess: exit
Bye

Conclusion

The code presented is the final version after all refactorings, not presented here. The application started as an experimental development, as it covered some interesting features of Swiby, rather simple, it appeared that writing about it would be a good thing. Some of the experimental ideas should be covered in another article.

I don’t think that writing such specs is feasible, or done, on big projects, from that point of view this example is not realistic.

First OpenJDK commit!

Mario Torre — Fri, 30 Oct 2009 14:57:41 -0400

Finally my bug was accepted and I got the OK to push.

Formally, it‘s not the first patch I produce that goes into the JDK repository, but all the other patches, that were a result of the work on the Caciocavallo project, were always so heavily modified by Roman for inclusion in the OpenJDK forest and always committed by Roman that I never counted those as real contributions from my side.

So this is the first patch, and the first commit, all done by myself, and I‘m quite happy because of this!

I would like to say a big thanks to Roman, Andrew John Hughes, Jennifer Godinez, Phil Race, Gustavo Galimberti, Tim Bell and Jim Graham and everybody else at Sun (and outside Sun, especially Mark Wielaard and the Classpath people, because I know all I know thanks to them!) for the help and the reviews and all the great suggestions.

I wish I could share this joy with someone of you in particular (that I know is reading this blog post now), but I cannot, but in the end I‘ll drink some Port wine tonight to celebrate, and enjoy and be happy, and anybody around the city is of course welcome to join, no reason to call, just show up :)

Btw, I feel excited like a child!! I really committed to the java2d repository?? Wowowow :)

Python script to convert markdown text into Google WIKI format

Rick Hightower — Fri, 30 Oct 2009 11:44:07 -0400

This code adds the ability to convert from Markdown text (the subset I am interested in) into Google Wiki format.

This python script coverts markdown text into Google Wiki text.

from markdown import markdown
from xml.etree.cElementTree import ElementTree, tostring
from cStringIO import StringIO
from xml.dom.minidom import parseString

def paraHandler(element, output):
    para = parseString(tostring(element)).childNodes[0]
    output.write("\n")
    for node in para.childNodes:
        if node.nodeName == "strong":
            output.write("*{0}*".format(node.firstChild.nodeValue))
        elif node.nodeName == "em":
            output.write("_{0}_".format(node.firstChild.nodeValue))
        elif node.nodeName == "code":
            output.write("`{0}`".format(node.firstChild.nodeValue))
        else:
            output.write("{0}".format(node.nodeValue))
    output.write("\n")

def unorderedListHanlder(element, output):
    output.write("\n")
    for child in element.getchildren():
        output.write(" *  {0} \n".format(child.text))
    output.write("\n")


def orderedListHandler(element, output):
    output.write("\n")
    for child in element.getchildren():
        output.write(" #  {0} \n".format(child.text))
    output.write("\n")

def preHandler(element, output):
    code = element.getchildren()[0]
    output.write("""
{{{
%s
}}}
    """ % code.text)

def headerHandler(element, output, level):
    output.write(("=" * level) + element.text + ("=" * level) + "\n") 
    
handlers = {
    "pre" : preHandler,
    "p" : paraHandler,
    "ul" : unorderedListHanlder,
    "ol" : orderedListHandler,
}


def convert_to_wiki(text):
    output = StringIO()    
    html = markdown(text, ["fenced_code"])
    html = "<div>\n{0}\n</div>".format(html)
    root = ElementTree().parse(StringIO(html))
      
    for child in root.getchildren():
        if child.tag[0] == "h":
            try:
                level = int(child.tag[1])
            except:
                continue
            headerHandler(child, output, level)
        elif child.tag in handlers:
            handlers[child.tag](child,output)
        else:
            print "Unknown tag {0}".format(child.tag)
    return output.getvalue()

To exercise this, I wrote a unit test (not a very good unit test btw) as follows:

import unittest
from markdown import markdown
from util.markup_util import convert_to_wiki

class MarkupTestCase(unittest.TestCase):
    def setUp(self):
        self.markup_text = """
# H1
## H2
### H3
# Header 1
## Header 2
### Hi how are you
----------
1. Hi
2. How
3. Are
4. You
---------

-----------
- this
- is
- anohter
- list
----------
This is some text. 

This is some more text.

I am a paragraph with special sauce *emphasis* and **strong asdfasdf** `printf (void)`.

~~~~{.python}
def foo():
    pass
~~~~
        """
    
    def testMarkup(self):
        
        html = markdown(self.markup_text, ["fenced_code"])
        print html
#        self.assertEquals(1, html.count("<h1>"))
#        self.assertEquals(1, html.count("H1"))
#        self.assertEquals(1, html.count("<ol>"))
#        self.assertEquals(4, html.count("<li>"))
#        self.assertEquals(1, html.count("</ol>"))
#        self.assertEquals(1, html.count("<code"))
#        self.assertEquals(1, html.count("</code>"))
        
    def testConvertToWiki(self):
        print "WIKI \n" + convert_to_wiki(self.markup_text)
        
    def ztestMarkupToDOM(self):
        from cStringIO import StringIO
        
        html = markdown(self.markup_text, ["fenced_code"])
        html = "<div>\n{0}\n</div>".format(html)
        print "\n"
        print "#" + html + "#"
        myfile = StringIO(html)
        
        from xml.etree.ElementTree import ElementTree
        tree = ElementTree()
        root = tree.parse(myfile)
        print `root`
      
        for child in root.getchildren():
            if child.tag == "pre":
                child = child.getchildren()[0]
            print "tag " + `child.tag`
            print "text " + `child.text`
            print "attrib " + `child.attrib`
                
        #http://docs.python.org/library/xml.etree.elementtree.html
        
if __name__ == '__main__':
    unittest.main()

I tried posting the whole thing here and ummm... Jroller exceptioned out.

The rest of this entry can be read here:

http://code.google.com/p/courseconstructor/wiki/MarkdownToGoogleWiki

Why You Should Never Talk to the Police

Sebastian Kübeck — Fri, 30 Oct 2009 08:11:54 -0400

This is fascinating as it is frightening: The two guys tell you why you can make your case only worse when you talk to the police as they will trick you into more trouble once you are making an attempt to open your mouth. The end result will be that the only people talking to the cops will be innocent people or "lightweight" crooks and those will be the only ones who are most likely to be punished whereas the "heavy weights" will defend themselves effectively and go free. I personally don't like where this is heading...

One note to the police in Italy and other countries: The claim that the police is worse everywhere else in the world compared to the US is simply a lie. It's used to excuse police practices in the states. If you don't believe me, read the Amnesty International Report on the US:

There were reports of ill-treatment by police and prison officers on the US mainland, often involving cruel use of restraints, or electro-shock weapons.
In October, former police officer John Burge was arrested and charged with perjury in a civil case in which he had denied knowledge of the abuse and torture of suspects. John Burge had been in charge of the Area 2 police station in Chicago where scores of black suspects had allegedly been tortured in the 1970s and 1980s. Although indisputable evidence of torture came to light through a subsequent inquiry, no officer had been prosecuted and John Burge was the first person to be charged indirectly in connection with the abuse.

That said, that doesn't excuse what the police does in Italy and other countries.

Fix gateway: Time synchronization/accuracy problem

Serkan Guler — Fri, 30 Oct 2009 07:50:29 -0400

http://serkanguler.info/2009/10/30/fix-gateway-time-synchronizationaccuracy-problem/

JSF RI 2.0.1 StackOverflowError problem and solution

Hazem Ahmed Saleh — Fri, 30 Oct 2009 06:06:51 -0400

I found the following error when I configured a JSF 2.0 application from scratch (JSF RI 2.0.1-b01):

java.lang.StackOverflowError
	javax.servlet.http.HttpServletRequestWrapper.getSession(HttpServletRequestWrapper.java:216)
	javax.servlet.http.HttpServletRequestWrapper.getSession(HttpServletRequestWrapper.java:216)
	javax.servlet.http.HttpServletRequestWrapper.getSession(HttpServletRequestWrapper.java:216)
	javax.servlet.http.HttpServletRequestWrapper.getSession(HttpServletRequestWrapper.java:216)
	javax.servlet.http.HttpServletRequestWrapper.getSession(HttpServletRequestWrapper.java:216)
...

The JSF 2.0 seems to enter in an infinite loop ...

After digging in the problem, I found this problem happened because the faces-config tag version in the (faces-config.xml) file was not "2.0" (A silly mistake done by me and can be done by others).

So please always make sure that your JSF 2.0 faces-config is:

<faces-config xmlns="http://java.sun.com/xml/ns/javaee"
              xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
              xsi:schemaLocation="http://java.sun.com/xml/ns/javaee http://java.sun.com/xml/ns/javaee/web-facesconfig_2_0.xsd"
              version="2.0">

...
</faces-config>

About similarities or constants in software and hardware businesses

Dominique De Vito — Fri, 30 Oct 2009 03:22:28 -0400

Having read Design Patterns: Elements of Reusable Object-Oriented Software few years ago, I was keen to read the interview Design Patterns 15 Years Later: An Interview with Erich Gamma, Richard Helm, and Ralph Johnson.

Ralph Johnson has risen an interesting point: "Writing one-line programs is not usually what we mean by design. Software has improved a lot over the years, and a lot of systems that used to require careful design can now be built by reusing other software. But there are a lot of systems that we can't build that way, and writing 100K lines of new code isn't that much easier now than it was 15 years ago. It will does a lot more, but costs the same."

That's a funny point because costs look like (always) the same, more or less, over time, both for software and hardware. And over time, what one get does a lot more, more and more.

Coincidence or not, sometimes software (business) micmics hardware (one), or may be, it's the opposite. Or it could be just some business constants.

It reminds me that Fred Brooks wrote in The Mythical Man-Month: Essays on Software Engineering that an average programmer writes about 15 000 lines of code in a year. It looks like another constant, too, because he wrote he noticed this ratio of lines whatever programming language is used. So, his advice was to prefer (more and more) high-level languages.

JRoller

Thoughts on Django

Felicidades Diana y Angel (Posada Miranda)

Carbonado 1.2.1 released

Swing Resources

Best Sites for Java Swing

General Purspose component Libraries

Components & Tools for Specialized Applications

Graphs,Diagrams and Charts

Date Choosers

Layouts

JTable & JList

Swing and CSS

Look and Feels

BallonTips

Java Visualization

Java WorkFlows

Java Expression Builders

Portlet life looks like EJB, but future sounds like GWT

Preview of the Portlets in Action book available

Java Softwares and IDEs and Builders

Java Softwares and IDEs and Builders:

Java SDK

Install the Latest Release of the Java SE Platform You can download the latest release of the JDK for free from here

IDE's

GUI Builders

Links Matter For SEO

Links are a source of traffic.

Ways To Get Inbound Links To Your Website

JBoss Best Practices & To Do List Before Production

JAVAWUG PROJAM #1 JQuantLib

JAVAWUG PROJAM 1 JQuantLib

Talk

Devoxx Linked-Up

Biography

Registration

Media Copyright Notice

Roma SpringFramework Meeting Slides

The term "portal" is, at least, rather confusing!

When It's OK Not to Write Unit Tests

It's OK Not to Write Unit Tests if You Write Code That Doesn't Have to Work

It's OK Not to Write Unit Tests for Functionality that is Tied To Infrastructure

It's OK Not to Write Unit Tests if The States They Should Cover Are Already Covered by a Different Verification Mechanism

Conclusion

Patent It

"Twittering" TheServerSide Java Symposium

Learning ICEfaces through ICEfusion

Visualize metrics with Panopticode

Spring 3.0 - Slides from a Seminar I Presented

Mobile Twitter Client with JSF

Scala and MongoDB

GridGain presents at Dallas Java User Group, November 11th

I Just Bought The Iomega Home Media Network Attached Storage

Viewaframework 1.0.0 Swing framework

How to resolve conflict of DWR and jQuery

Back to being a Java blog

Django and JQuery: Integrating Django and JQuery Autocomplete plugin

Hazelcast Meetings in New York

FEST-Assert 1.2: Fluent Interface for Assertions

Refactored GoogleWiki to Markdown to support MediaWiki

What's new in JSF 2.0?

Have you tried Spring Roo?

Dynamic Images with JSF

GWT and Jetpack are going to shake up the web-based (RIA) desktop application landscape

JRoller really doesn't work with Safari

JRoller really doesn't work with Safari

JRoller really doesn't work with Safari

JCP: At The Crossroads

Headless image manipulation in Java

Resize an Image

Fit an Image

Crop an Image

Element of conception: name important external resources (or cross-frontier resources)

Notes on Brion Vibber's recent Wikipedia Talk

ICEfaces 2.0 Alpha now available

Escaping HTML Entities in Python

Sweb and styles (part 3)

Create the visual part

Give life to the window

Step 1: the startup message