Open Source

IBM files patent- Automatic blogging during media viewing

2009-08-31T07:21:00.001-07:00

Yet another micriblogging news. IBM has filed a patent according to which people will be able to blog while watching TV by using a remote button.
No mobile phone just remote button and your status (what are you doing means watching tv) will be posted to blog.
How this autoblogging will affect our life can be experinced only when people whill have this device.
You can see more information about this patent at :-
http://tinyurl.com/m6hgxe

OpenXava 3.1.4 released

2009-08-31T06:54:00.001-07:00

OpenXava is Ajax based Java framework for rapid application development. Its new version 3.1.4 is released. This new version has more than 50 new features and fixes over previous version.

Among other things this new version has :
Translation to Serbian and Swedish.
This will also support Liferay 5.2.3 with Tomcat 6.
@Formula annotation from Hibernate supported for detail and list mode

You can see full change log on its site:
http://www.gestion400.com/web/guest/home

With OpenXava you can build application rapidly. You only need to write your model, controllers are reused and Ajaxified UI is automaticaly generated by framework for you. You need to create your JPA classed only.

Some of the features of OpenXava are:

Fine tunning of front-end can be done.
AJAX support.
Can be adapted to work with legacy database schemas.
Application server: Tomcat, JBoss, WebSphere, etc.
Supports JSR-168: All OpenXava modules are standard portlets too.
Persistence engine: EJB3 JPA, Hibernate or EJB2 CMP.
IPortals suppoted: JetSpeed 2, WebSphere Portal, Liferay and Stringbeans.
Integration of reports made with JasperReports.
Licensed under LGPL.

Twitter look a like requires minimum 1400 chars.

2009-08-30T03:30:00.000-07:00

Social networking fever is high. People are innovating new ideas to utilize this fever in their benefit. A twitter look a like site woofer requires minimum 1400 characters where twitter allows maximum up to 140 correctors. Woofer wont ask you to register but it asks you to put your twitter user name. I see one security issues here. I can post by using any twitter ID and that can be a serious issue. I am not sure why they did it and how will they respond to this challenge. I have posted a woof by just putting my twitter ID and it allowed me to do so.

They call it macro-blogging site. You can see it here.

Is govt trying to controll internet ?

2009-08-30T02:37:00.000-07:00

Yesterday almost all news agencies reported about Senate bill which would allow president to shutdown online traffic by disconnecting the private sector computers from the internet. If this bill will be enacted president in some cases can announce cyber emergency and can seize private networks. Full story can be seen at cnet news and foxnews.com.

Now issue I have with this idea of controlling the cyberspace is govt interference in freedom and right to speech. First of all cyber emergency is not clearly defined and infact this updated version they will submit with ambiguous language because earlier there were issues raised about it and they have redrafted it. Misuse of it can not be ruled out. Won't it be like countries where govt's have banned social networking sites, news sites explaining it as a measure to secure their country and people. All know what those govt's are doing. Today one can say that is is not possible in US for govt to misuse it but this belief is because it was not allowed and not promoted in past. There was freedom to speak, US citizens never tolerated infringement of their freedom and right to speak. Once govt starts taking control of things in the name of security, nations pride and other things and later some people (who become dictators ) misuse those laws against their citizens.

Still we will have to see ultimately what shape this bill is going to be finalized or even not finalized.

Nokia is going for its first Linux phone

2009-08-28T14:44:00.000-07:00

Good news for Lunux lovers. Nokia is going for its first Linux OS based phone. It is confirmed. According to its website :- "this is a phone with RAW computer power. Operating software is Maemo 5 software on Linux. It will have 32 GB of built in storage and can be extended further 18 GB with external microSD card. 800X480 pixel resolution with 3.5 inch touch sensitive widescreen display. Up to 1 GB of memory (256 MB RAM, 768 MB virtual memory). 5 megapixel camera with CMOS sensor, Carl Zeiss optics, Tessar lens. 3X digital zoom".

There are lots of applications like calender, PDF reader, RSS reader etc along with application download manager. Most of the popular audio formats are supported and there is FM transmitter also coming with it.

For full detail please visit its website here. Must also see the image gallery there, you will love it. This product is not yet launched and is coming soon.

Poll Results- Will Google Chrome OS beat Microsoft Windows?

2009-08-28T01:37:00.001-07:00

Recently I posted a poll on this blog and it is closed now. Here are the results.

41% only said Google Chrome OS will not beat Microsoft Windows. This is very important. Although total number of votes is too low but still it gives us an idea what people think. Only half of the people were in favour of Microsoft Windows. People are serious about Google Chrome OS. 41% said yes Google chrome will beat Microsoft Windows. This is a great number for a new comer when it is compared with an existing giant in desktop computing.

Google has been a trend changer for a long time and if it will be able to fulfil expectations of the people they are willing to welcome it.

Future of Open Source

2009-08-26T09:57:00.000-07:00

Lots of people have been talking (in favor of big commercial product vendors) that open source products are not mature, they can not be considered seriously because there is no one who owns and take responsibility etc etc...

There are few facts with respect to open source which needs to be considered while adopting open source but the fear which people were creating and some were living in was not a reality. Lots of open source projects are better and more mature than their commercial counterparts. The question of ownership has various aspects and one of them is: "Even if some one is taking ownership after selling you something, what penalty clause they have is something is buggy in their product?". Do you go to judiciary often to sue them? Do they resolve product bug withing a day or two? If it is only phone support and you have got it in past you must be knowing what happens. Their bugs become features, your functionality becomes not feasible and you have to compromise functionality till the time they release bug fix.

On the other hand there is nothing to hide in case of open source. Thousands of people work on them, lots of people do testing along with thousands submit bugs while they use the products and solutions too if they find. This collaborative culture improves the product regularly. But you should have technical staff in your team is you are using open source not clerks. There is a breed in IT and that breed wants to work like a clerk to just put stamp on all dispatching letters. These kinds of people find it difficult to use open source. Even if a solution to a problem is available they can not or don't want to look at it because they are not made for this. They are made to put a stamp of a letter. It is not difficulty which stops them using open source, it is their nature, attitude (they are not made for analytical work) which stops them finding solution.

It is true that some open source products have limited documentation or not so good documentation but same is true about commercial products too. All the commercial products don't have very excelent documentation. Some have, some don't have.

It is true that solution to a problem in open source may take some time ( it is not guaranteed) but who among commercial vendors gives you guarantee about fixing a bug in the product within 1 day or one week? They will respond to your email or call within a fixed time period but will they resolve it?

It is also true that every open source product is not very mature and some have bugs but can you tell me name of any commercial product in which there was no bug in any release? How funny. Let me further explain: All release versions, fix packs, patches, so what these patches, fixes provide you? Bug fixes, right? One of the commercial operating system during 1997 was crashed during its launch in public you all know. How stable that product can be considered?

It is true that you need a good team of technical people who can work on open source stuff, you should be careful while selecting an open source product and should select one which has enough maturity and supported by a big community. You should also have an open source policy in place and if paid support is available for a product that will give you assurance of support too.

Now I would like to share some stories which are floating around for some time related to open source and these stories will prove what I am saying is not incorrect (future of open source is quite bright and even commercial vendors will depend on them):-

Gartner reported some time back in its website titling "Predicts 2009: The Evolving Open-Source Software Model" very interesting story about open source future. According to this prediction of Gartner:- "Through 2013, 90% of market-leading, cloud-computing providers will depend on OSS to deliver products and services". You can read this report at Gartner website here.

ComputerWorldUK recently reported "Microsoft deepens open source embrace". According to CW : "Microsoft's Developer and Platform Evangelism Interoperability team is launching an open source project to bridge PHP and Microsoft's .Net programming model.". See full story here.

According to InformationWeek : "The Defense Information Systems Agency is making its internally developed suite of administrative and workflow applications available to other government agencies as open source software". See full story here.

PHP Frameworks Comparison - Part 1

2009-08-26T00:43:00.000-07:00

Frameworks provide lots of things which you need to create if you are not using framework. They also provide a standard infrastructure using which you can create applications faster and with some degree of discipline. There are so many frameworks in PHP and they are growing in number as well as becoming more mature with time. I am trying to list down some of them here and will try to compare them with features. Later will try to identify top 10 or even 5 PHP frameworks. Although there is no single solution to everyone and every requirement may have its own expectations which a particular framework may or may not be able to match but still it is worth looking which frameworks are more popular among people.

There are two aspects when we compare features. One is support of feature and another is how well that feature is implemented. So even if two frameworks support a feature, one may look better to some one because it is easy to use. Some one else may find this "easy to use" feature not so important. Instead of this feature some one may be interested in other features like robustness, scalability, extendability or stability..

I started study of some of the PHP frameworks but study itself can not be done fairly on all the PHP frameworks because they are large in number. Here I am not considering Drupal kind of framework in this discussion because they are very popular feature rich CMS and I will discuss about them later. There are more then 30 PHP frameworks listed in Wikipedia website. List maintained in Wikipedia may be older but it gives good understanding of basic features of various frameworks. The list compares frameworks based on features like support for:- form validation, security, ORM, Ajax, i18n, caching, templating, database migration etc.

Here I am taking 10 out of so many PHP frameworks for my study and analysis. Lots of work has already been done so I will first try to consider what others have found and told and later I will try to get feedback from community (you) on their choice of best PHP framework. My focus is web frameworks mainly MVC frameworks but obviously I am looking for support of features like security, ORM etc too.

Things to consider while selecting a framework

Performance
Security
Flexibility
Documentation
Simplicity

Generally I keep above mentioned features in that order of importance or priority. Other features like PHP versions supported, ORM, validation, caching etc are also important but almost all or at-least top 5 support all these things so than how will you decide which one is better? Above mentioned 5 features can differentiate various frameworks even if they support same features.

My first pick

CodeIgniter
Zend
Yii
CakePHP
Symfony
PHPDevShell
Prado
Akelos
ZooP
QPHP Now PHPWork
Kohana
Fuse
P4A
Horde
DooPHP

Out of these 15 I have identified top 5 (1 through 5) very interesting because there are stories, benchmarks and discussion available on the internet but I would like to get feedback, inputs and popularity vote from all of you to to decide which are the top 5 PHP frameworks.

Why did I found top 5 very interesting?

Popularity
I saw a voting result at http://www.phpframeworks.com/top-10-php-frameworks/ (although it may be based on older versions and right now I am unable to see this page or even this site. This site is not working for some time. Number of votes were also too low to decide anything but it is interesting to see if you are starting.)

Another poll result I found at http://www.opensourceuniverse.com/component/poll/17-which-php-framework-holds-a-promise-for-the-future (still very few votes but yet gives a hint.)

Feature base comparison

I found interesting comparison of Yii, Cohana and CodeIgniter at http://www.beyondcoding.com/2009/03/02/choosing-a-php-framework-round-2-yii-vs-kohana-vs-codeigniter/
It is based on various features and is newer one.

Another comparison chart is available at http://www.phpit.net/demo/framework%20comparison/chart.php
But it looks like older comparison. It compares 10 framewroks including Zend, Cake, Symfony, CodeIgniter, Prado etc.

One more comparison which is feature based comparing Cake, Zend and CodeIgniter can be seen here - http://www.tablefy.com/compare/2710/php-frameworks-comparison

Performance benchmarks

I found three interesting performance benchmarks

Paul M Jones (comparing Cake, Solar, Symphony and Zend) at http://paul-m-jones.com/?p=315
Denial Carrera (comparing Cake, CodeIgniter and Yii) at http://daniel.carrera.bz/2009/01/
Aventis Lab (comparing Cake, CodeIgniter, and Zend) http://avnetlabs.com/php/php-framework-comparison-benchmarks

What I concluded (although these tests are done on older versions of these frameworks) that CodeIgniter, Zend, Symphony and Cake are performing in the order I have written them here (i.e. CodeIgniter being the best performer, Zend second, Symphony third and Cake fourth).
I didn't find any performance results for other framewroks.

In next part of this series (PHP framework comparison) I will come up with a matrix comparing various PHP frameworks based on some features (whether they support or not and how well they support these features) but I would like to limit this comparion to 5 frameworks so I request you to please first vote for your best PHP framework so that we can shorten the list. Things you can consider for choosing best framework can be : Performance, Security, Flexibility, Documentation, Simplicity, ORM Support, Caching Support, Validation, Support for Unit Testing etc.

So please vote on this blog:- which one is the best PHP framework?

Free download - SOA Design Patterns Poster

2009-08-21T09:16:00.000-07:00

SOA design patters poster is available for free download. This is my small contribution for community. This poster will help you having single view of all the 85 SOA Design Patters described in famous book: SOA Design Patterns (by Thomas Erl, Prentice Hall, Copyright 2008). This poster is available in two sizes:-
1) A3 Size 2 pages (good for deveopers to take printout and fix on their desk)
2) A2 Size 1 page (Bigger poster so that all the patters with more detail can be viewed at one place)

Here is the URL for download:-
http://www.soapatterns.org/reference_posters.asp

Open Source Java Profiler - OKTECH

2009-08-20T04:09:00.001-07:00

This is an open source project recently anounced and available at google code.

"It can be executed in a local JVM or it can connect to a remote JMX server" according to project site at google code.

This profiler is a sampling profiler which periodically samples the JVM. Profiling information is stored in a file. It is a lighweight profiler.
Major drawback as compared to other commercial and some of open source projects is that this profiler doesn't provide graphical reports, html or any other form. Report is only available in text form but I hope project team and community will work on this area to generate good reports our of profiled data.

Maturity of the project for profiling big code base is not yet reported by community and if it proves itself on this then other things cam be improved. How stable this relase is is not yer known to me because I have not run it on a production application or large code base.

For more information please visit project page: http://code.google.com/p/oktech-profiler/

Best Open Source Framewrok related to Java

2009-08-15T13:39:00.000-07:00

Some time back I started discussion on Open Source Frameworks. I posted few times in this blog and was also collecting responses / inputs from community in my professional network. As I promised I will come up with more open source frameworks and my previous attempts were focused on :
First: Full stack framework in Java.
Second: Web application frameworks in Java

Regarding other frameworks (PHP etc) I am working on them and will come up with whatever I will have.

Now coming again to Java related frameworks, there was a confusion when I started poll and survey on it. Some people thought I am going to compare and choose best of all framework from previous polls but that was not my intent. My focus was Java related open source. If there was any confusion on my part I apologize for the same. One thing I would like to clarify here is that Wicket was included in this poll along with others intentionally to know feedback from people whether they think other (web / MVC) frameworks like wicket should be compared with frameworks like Seam, Grails, Rails etc. People responded correctly that Wicket should not be put here. It can be used with Seam with which it is competing here. It can be used in place of JSF in Seam. I included Rails along with others (Java/JVM based frameworks for comparison only) obviously Rails is not originally Java based (before JRuby came).

I thought I would put Spring MVC with web frameworks in another poll but some people were interested to see Spring here in this list.

Note: Polls discussed here were conducted in the month of April/May 2009 and some were continued till June also. Till today popularity of any of the framework might have changed.

Full stack framework in Java
My personal network of professionals

Here are final results which I collected from my personal network:-
When I asked people to vote for best application framework and choices were - JBoss Seam, Grails, Rails, Tapestry, Wicket following was the response from my personal network of professionals-

JBoss Seam - 26% Votes
Grails - 33% Votes
Rails - 15% Votes
Tapestry - 5% Votes
Wicket - 19% Votes

Total vote size was 1017

So it clear that popularity vote gone in favor of Grails and we can understand it. Its simple and easy to use feature made it popular.

JBoss Seam got second highest votes and it was much ahead of Rails.

I never expected such a good response for Wicket and I had to accept my mistake after seeing its vote percentage. This was one of the very interesting candidate because population of Wicket community is low as compared to others in my personal network so it's vote percentage is more important.

I am surprised with vote percentage of Rails because it has got very good community and it is in use for a long time as compared to few other in the race.

By company size 47% of the voters were from small companies, 22% from enterprise, large and medium 15% and 16% respectively.
By Job Function 74% voters were engineers, consultants 13% and rest were others.

Poll from this blog

I also posted poll on this blog in which competitors were JBoss Seam, Grails, Rails and Spring and following were the responses:
Grails - 27% votes
Rails - 7% votes
Seam - 25% votes
Spring - 54% votes

Total vote size was 79 which can not be considered huge but still it was voice of core professionals and can not be ignored.

If you leave Spring for a moment you can again see Grails and Seam in first and second position which was the case in my poll results posted above (professional network).
As I mentioned above people from my professional network were asking why not Spring and it is reflected here when I included Spring with Grails, Rails, Seam. Spring got first position with 54% votes. This is a significant difference.

Conclusion

When I discussed with people, saw their feedback and comments I came to conclusion (along with this data which I collected) that JBoss Seam and Grails were gaining popularity but Spring is in use for a long time and lot many people are using it. Grails and JBoss Seam will have a long journey before they can see same popularity as Spring.

Web application frameworks in Java

Spring MVC - 33% Votes
JSF - 22% Votes
Struts - 24% Votes
GWT - 15% Votes
Wicket - 4% Votes

Total vote size 45

Spring MVC got best vote % here too. Struts was second and JSF third. GWT has got popularity in the recent past it is reflected in its % which is not too behind Struts or JSF.

Poll on this blog

I also put a poll on this blog to compare web application framework in which I included GWT, Spring MVC, Wicket, JSF, Struts, Tapestry

GWT - 21% Votes
JSF - 30% Votes
Struts2 - 14%
Spring MVC - 39%
Wicket - 23%
Tapestry - 3%

Spring MVC is clear winner but JSF is also very close to it. Wicket here too got very good percentage of votes competing at third place.

Conclusion

Spring MVC has a large user base and people are using it. Wicket and GWT are becoming popular in short period of time.

As a final comment - Nothing in this world is constant. New promising frameworks have gained popularity in very short period of time as compared to veterans which have history of 5, 8 years.
New frameworks can beat olders anytime within next couple of years (may be next year) but it has not happened yet.

You can see some of the poll results at following URL's
http://polls.linkedin.com/poll-results/33616/pikzw
http://polls.linkedin.com/poll-results/34665/vzaom

Google Chrome for Software Developers- Possibilities

2009-08-10T10:20:00.000-07:00

When I saw posting on Google's official blog about a new operating system "Google Chrome OS" one thing came to my mind: Can software developers be able to write their code and compile it using this OS? Reason behind this question was the fact that Google always focuses on web based applications. Now this time it is OS so it should also be influenced by web technology.

I can see lots of opportunities for software developers and software development companies through this OS. This OS can realize dream of software as a service and no organization other than Google can make it possible.

When I am saying development it means developers can work on this OS using their IDE's and can write code, compile it. It can be a great value proposition for software development organizations too. These organizations fight for IDE's licenses, upgrates, setup and issue resolution. There are very good commercial software too which are too costly to buy but they can be used on pay per uses basis. Small to medium companies can afford these software as a services on pay per uses basis. These software as a service can be made available through Google Chrome (which may not be the case with currently what Google to going to launch but it is my thought)

Similar to other services offered by Google to businesses, like office suite, email service etc, software development environment can also be offered as a service on web through Chrome OS. This software development environment may include famous IDE's like Eclipse, Netbeans etc or may be Google's own web based IDE (which can be built on top of eclipse). Another services for software development can be version control system, build and test environment. All can be offered as service through this OS on web/internet. This can be a great example of SaS. Other vendors can also offer their software as a service on per uses basis like IBM Rational suite.

What is Google Chrome OS ?

2009-08-10T10:03:00.000-07:00

Last month Sundar Pichai, VP Product Management and Linus Upson, Engineering Director posted on google's official blog about Google Chrome OS. As per this blog entry:-
Google will open source code of Chrome OS, later this year and Google is targeting to launch it in the second half of year 2010 for Notebook. Initially it will be targeted at Notebook and will run on Linux kernel. Most of it will be web based as google always tried to set new trends through web technology.

According to official google blog post: "Google Chrome OS is being created for people who spend most of their time on the web, and is being designed to power computers ranging from small netbooks to full-size desktop systems."

Google is focusing on 3 s : Speed, simplicity and security as key aspects of this OS. They are designing this OS from ground up to prevent viruses & malware. Google knows very well that key to success in this area is speed, simplicity and security and we hope google will be able to meet the expectations of new generation.

Will Google Chrome OS beat Microsoft Windows?

2009-08-10T08:47:00.000-07:00

Google has announced its plan to launch Chrome Operating system which will run on very low end machines. As google has already launched office suite (word processing, spread sheet, calender) as online web based solutions for office users which will specifically attract small businesses this new initiative (Chrome OS) may prove to be another trend setter. Low cost user machines equipped with browser based applications, software may become choice for small to medium businesses.

On the other hand Microsoft Windows and its Office Suite was unbeatable till now even when other OS and Office suites came and launched as open source products.

There is no reason (technically) why big organizations will not consider this proposition (google OS) seriously. Still it needs to be considered from acceptability point of view whether this will also be accepted by organizations as other google applications. Although google apps are very cool and very well accepted among users, netizens but big organization are different. They have cultural issues, people taste and valuable data (sensitive for business) which may (and was) cause delay in acceptance of google apps including google OS.
Benefits of OS suggested by google may be unlimited but still we will have to see the acceptance.
I request you all to vote on poll posted on this blog: "whether Google Chrome will change desktop landscape or not?"

Open Source Full-Stack Frameworks Comparison - JBoss Seam, Grails & Rails - Part III

2009-04-27T07:33:00.000-07:00

Advantages & Limitations of Seam, Grails & Rails
[ Note:- This post is part of a series on open source frameworks. Please see my previous posts (Open Source Application Frameworks, Open Source Full-Stack Frameworks Comparison - JBoss Seam, Grails & Rails - Part I , Open Source Full-Stack Frameworks Comparison - JBoss Seam, Grails & Rails - Part II )
In my previous posts and in this post too I am focusing on Java based frameworks taking future development in mind. ]

Seam
Advantages
Very simple and intuitive. Has support/APIs to cover other layers of application rather than just web interface. Full-Stack framework with CRUD generation. Conversations with jBPM. Supported by JBoss/Red Hat. Based on JEE so suitable for enterprise class applications.
Makes it easy to implement multi-window operations, back button support, and multiple workspaces in a single window, stateful navigation.
Limitations
Works best on JBoss Application Server. Designed for EJB 3. Things like ICEFaces, Quartz Scheduler are still hard to use (whether it is due to lack of sample example or any other issue it is not a good thing for a developer).

Grails
Advantages
High productivity. Less lines in code. Groovy is easy to learn for Java Developers. Using Spring and Hibernate at its foundation. It has all the support of Java Platform but with the charming of Groovy, with the power of Spring, Hibernate & Sitemesh, comes with excellent ideas like GORM and GSP.
Limitations
Not as performant as using the raw frameworks.
Can be hard to sell to stakeholders who like Java
Enterprise grade maturity for very large projects may not match raw frameworks.

Rails
Advantages
Already mature framework, big community support, books available, high productivity.
Limitations
Performance may be a concern for those who are very much keen about it.
Scaling issues (with growing users) - throw more CPU's for more users.
Enterprise grade maturity for very large projects may not match raw frameworks.

What types of projects they are suitable for and where they are not suitable for?
Seam
Suitable for
The solution for people adopting Java EE 5 standards (JSF, EJB3).
The solution for applications that support multitasking (multi-window/multi-workspace).
Projects where BPM is required.
Projects where rich user interface is a must have requirement.
Not suitable for
Not for splashing data from a table onto a web page. Use PHP or Ruby on Rails for that
Not for people using JDK 1.4.
Not for people who have already invested in large teams on Struts and are totally happy with Struts.

Grails
Suitable for
Small to medium projects where very high productivity is required.
Projects where time to market is very short.
People who are from Java background and are looking to adopt a high productive framework.
Not suitable for
Projects where very high performance is a must and highest priority requirement. Although running on JVM will give it advantage over Rails in terms of performance but it can not beat raw frameworks.

Rails
Suitable for
Small to medium projects where very high productivity is required.
Projects where time to market is very short (product is required in very shot period of time).
Not suitable for
Projects where very high performance is a must and is a highest priority requirement. With growing users it may not scale well.

Conclusion
Seam requires some time (and time will prove whether it will beat Rails/Grails or not) to get maturity/perfectness of what ever it claims or suports. Since it is based on JEE and supports all those high productivity features required for a full stack framework its future is bright for enterprise grade applications. Grails is good for small to medium applications where high productivity and time to market is very important. Rails was already enjoying the place where Grails was trying to place itself and in fact Grails was able to prove itself as an alternative to Rails. If community of Grails as well as Seam will grow in near future an people behind them will try to fix issues in these frameworks (at the same time if they will not try to make them complex and heavy weight) then you will see only two names in the Java/JVM related full stack framework market and they may be Grails and Seam. It is more important for Seam how they make it more perfect and working, simple to use, developer friendly. Seam has potential to beat Grails, Rails and any other full stack framework.

Interesting information
Grails & Rails compared on maturity, development, documentation, production, potential:-
http://blog.zmok.net/articles/2008/03/18/rails-and-grails-comparison

Grails & Rails comparison on common features, differentiators:-
http://www.jtict.com/blog/grails-vs-rails/

Interview with a twitter developer - One of the answer explaining scaling issues with Rails:-
http://www.radicalbehavior.com/5-question-interview-with-twitter-developer-alex-payne/

Open Source Full-Stack Frameworks Comparison - JBoss Seam, Grails & Rails - Part II

2009-04-23T20:54:00.000-07:00

[ Note:- This post is part of a series on open source frameworks. Please see my previous posts (Open Source Application Frameworks, Open Source Full-Stack Frameworks Comparison - JBoss Seam, Grails & Rails - Part I )
In my previous posts and in this post too I am focusing on Java based frameworks taking future development in mind. ]

Now it is decided that in this discussion we will focus on open source full
stack frameworks (mainly Java based and one very close competitor to Java
frameworks). There are few criteria based on which we should compare these
frameworks. Here I have listed few expectations of a business from a solution
(here framework in question):-

Scalability, performance, manageability, maintainability
Rich User Interface (Ajax support)
IDE Support
Productivity
Developer's choice
Application size/footprint
Flows
Platform
Delivery date

Whether these features are sufficient or few are not required may be debatable
but whatever criteria people choose, it should suggest based on their
experience that which framework they are comfortable with and what are the good
things for a business in that framework.

I conducted a poll on linkedin and also sent some information to people who are
in software development for last one decade or more (specially in Java based
development). Few of the above mentioned criteria I put for deciding best
framework and few criteria are sugested by people who participated in the poll.
Till this writting more then 850 people casted their vote, out of which
following is the score:-

Please visit this link to see the detailed results:- http://polls.linkedin.com/poll-results/33616/pikzw

Grails

Grails gives you all the benefits of Rails but much better performance (runs in JVM) and removes the hassle of adding the support for Ruby, etc. You can call Java/J2EE components. Productivty is high. Grails gives you the speed of development of RoR with the maturity of the Java platform. Grails uses the power of Spring and Hibernate under the hood and combines that with SiteMesh/GSP and the strength of a JVM based dynamic language as Groovy.

Seam

JBoss Seam has the best JSF implementation around, coupled with an extensive EJB layer makes it suitable framework for enterprise solution. Red Hat support is also available if you wish to step up from the "free" status. CRUD screens generation very easy.

Rails

A lightweight web framework written in Ruby scripting programming language. It contains its own application server. Big community, books are available. Wide variety of libraries and plugins.

Each of these frameworks claim similar features but here are some facts and figures:-

Time to market and adaptability

People at AlterThought conducted a study. A medium complexity application with
approximately 20 medium-to-high complexity features (use cases or user stories)* was considered and they came out with following data:-

(Source:- http://alterlabs.com/technologies/java/revenue-on-ruby-on-rails/)
Although this information doesn't say anything about RoR vs Grails or Seam. Not an issue, look ahead.

Productivity

People at AlterThoughts also published following report:-

(Source:- http://alterlabs.com/technologies/java/grails-vs-rails-the-thrilla-in-manilla-a-study-on-grails-productivity/)

Performance

A benchmarking study for Seam is published at -

http://www.dell.com/downloads/global/power/jbossworld_2006_june_jaffe.pdf

Following benchmarking compares Grails vs Rails peroformance-

http://docs.codehaus.org/display/GRAILS/Grails+vs+Rails+Benchmark
shows Grails out performed Rails. They conducted tests for creation, updation, query etc.
a) Read operations

b) Create Operations

c) Qeerying

In my next post (part III ) I will compare these frameworks with some more data and information (facts and figures :) )

Interesting information

Seam/ JSF vs Wicket - You can find out performance comparison of using Seam with JSF VS Seam with Wicket

Open Source Full-Stack Frameworks Comparison - JBoss Seam, Grails & Rails - Part I

2009-04-22T06:53:00.000-07:00

Here we will focus our study on three open source frameworks :- Seam, Grails and Rails (Why only these three? Please see my previous post by clicking here).

What companies are looking for

We have taken these graphs from indeed.com to see how many job postings are posted for a framework. Following graph shows percentage of job posted is highest for Rails and Grails, Seam are at second and third position. Important to see here is that difference between percentage of jobs for Rails and others to huge. It means more companies are working on Rails, more applications are under development using Rails as compared to Seam and Grails.

What is the future trend?

Now see the graph bellow. It shows percentage growth (increase in job postings) of job market for three frameworks. After looking at previous graph as well as graph bellow we can say that although more existing applications are using rails but growth of Grails and Seam is more than Rails. Rate at which Grails jobs and Seam jobs are growing is much higher than Rails. Grails is the leader here. So more new applications are considering Grails and Seam.

What about other frameworks?

Here I tried comparing job postings for other frameworks along with Rails, Grails and Seam. You can see Grails is leader in percentage growth (increasing demand). Other leaders behind Grails are Wicket, Seam and Rails.

Where is Spring?

Following graph shows existing job requirements for Spring are highest among other frameworks. Struts comes after Spring and then comes Rails.

Now considering Spring as complementary to other frameworks (as I mentioned previously) if we don't compare Spring with other and we also remove Struts because it focused on application layer not a framework for all layers of an application so comparing it with Rails, Grails and Seam is not good. Now look at graph bellow. Rails is leader.

Again following graph shows growth rate for Grails, Wicket and Seam is better than others.

Conclusion
From industry demand perspective it looks that Spring, Struts were past and are present but Grails, Seam are future.

Discussion on other aspects to compare in next post

Interesting resources
http://developers.sun.com/learning/javaoneonline/2008/pdf/TS-6457.pdf (2008)
http://static.raibledesigns.com/repository/presentations/ComparingJavaWebFrameworks-ApacheConEU2007.pdf (2007)
https://equinox.dev.java.net/framework-comparison/WebFrameworks.pdf (old one-2006)

Open Source Application Frameworks

2009-04-21T11:46:00.000-07:00

When you talk about framework and search the internet, look at wiki posts or blogs you see lots of frameworks (they are 100+ in numbers). I bet you must not have heard about all of them. Even if you have heard about few of them you don't know much about all of them. In such a scenario it is very difficult to choose one among so many. There are few people who tried comparing few frameworks but those comparisons don't give very useful information based on which a decision can be taken. Issues with those comparisons are:-

1. Those comparisons were published in 2006, 2007 and few frameworks were emerging at that time so what ever happened during last 2 years is not available so those comparisons can not give you correct picture.
2. Even if you look at 2 years old comparison, they are web framework comparisons. Their focus is not on full fledged application framework. They take Struts, JSF, Spring MVC etc and points of features are Ajax support, Page Decoration etc.
3. Even if you try to look at them considering web frameworks only (and older too) you see they are also comparing frameworks like Rails, Grails and some times Seam with Struts, JSF.

I will try to focus separately on web frameworks as well as complete application frameworks.

Before talking about any framework one should think about features and their priorities. Features selected for comparisons should also include what a business require rather then technical features which are fancy, great inventions and are important for networking sites (social or other networking :D ) . Here I am trying to list very few things from business point of view (what is more important for a business and business user):-

1. Long life of business applications (requires framework with future vision & features).
2. Businesses can not change their applications frequently (Stability in framework is required. If direction of framework is going to change it will not suite the business).
3. Outage, unavailability of application may result in loss of millions of dollar or loss of customers, customer's interest and faith worth millions of dollar (Framework should be robust with better availability).
4. Data is very important for them and so stability and security are of great importance.
5. Support is very critical and they can not rely on ad-hoc support (only community support). There should be onus on some reputed organization having expertise in it

Now if you consider yourself as a Bank Manager or CEO of a car company looking for an application to solve some of your business problems:-would you consider whether something is written in Java or any other language or something is using EJB 3 that is why you can not use it.

Don't take these big examples and consider buying something for yourself (car, washing machine, furniture) don't you put "after sale service" (good service) on top. Even if a company is giving 7 years warranty but is known for not giving good service or not known at all at one hand and another company giving 4 years warranty but is a trusted brand, known for giving good "after sale service". Trusted brand with good "after sale support" is more important then cheaper products without a trusted company/organization. Company or organization in case of software products may be an organization working on open source/open standards or may a commercial product vendor.

Crux of the above: - trusted, guaranteed support is a must.

Almost all the frameworks claim all the features support but which one really works is important. Another point to consider is how quickly you/your team can learn and use a particular framework.

I have not discussed technical features here because almost all the modern frameworks claim to support all those features still I will discuss about those features later.

List of frameworks:-
Rails, JBoss Seam, Grails, Tapestry, Wicket, Spring, Appfuse, Struts, GWT, SpringMVC etc

Structure of an application (typically web based enterprise application)-

It was very difficult to integrate and configure all of these parts together. Release of a framework every day added more salt to the injury. Glue was required to work with various frameworks for various aspects of an application. Apart from integration and configuration issue there were productivity concerns too. If database entities are know, why to code for making maintenance screens for them? It should be auto generated. There were various aspects of an application for which plug-in or solutions came later but again gluing them with whole eco system was a challenge and was time consuming.

Framework like Seam, Rails, and Grails tried to solve above mentioned issues.

Elimination

First of all Appfuse is eliminated because it is not a full-stack application framework but a project skeleton creation tool. What we are focusing here and expecting from a framework is next generation framework providing glue between various frameworks and layers as well as provides lots of services and productivity enhancements.

Struts, GWT and SpringMVC, Wicket are web-frameworks for application layer in another word they are focusing on problems of presentation, controller so we will not compare them with others like Rails, JBoss Seam, Grails and Spring.

Spring was an alternate to EJB at time when EJB was too complex to write (lots of interfaces, configuration and heavy weight container) but now with annotations and today’s EJB specifications EJB’s are just like a POJO. Another aspect of new changed scenario is that JEE containers can host POJO’s too. So advantage of Spring over J2EE in this context is no more after J2EE became JEE. Another angle here is that Spring doesn't compete with frameworks like Seam, Grails but these framework use Spring and Hibernate so comparing Seam, Grails, Rails with Spring and Hibernate is not correct.

Now main players left in full-stack framework race are: - Rails, Seam and Grails.

I will compare these three in my next post (although I will try to discuss Spring also along with them but will avoid comparing Spring with them because it complements them not compete with them)

Privacy Policy

2009-03-22T13:11:00.000-07:00

This Privacy Policy is intended to describe how your personal information is processed and used, and this blog (completeopensource.blogspot.com) will make every effort to ensure that its activities keep within the spirit of this Privacy Policy.

Please note that by visiting and using this blog, you are accepting the practices described in this Privacy Policy.

Third Party Advertising

We use Google AdSense and other third-party advertising companies to serve ads when you visit our Web site. These companies may use information (not including your name, address, email address or telephone number) about your visits to this and other Web sites in order to provide advertisements on this site and other sites about goods and services that may be of interest to you.

Google, as a third party vendor, uses cookies to serve ads on our site.
Google's use of the DART cookie enables it to serve ads to our users based on their visit to our sites and other sites on the Internet.
Users may opt out of the use of the DART cookie by visiting the Google ad and content network privacy policy.

Java Code Review Tools

2008-04-01T17:29:00.000-07:00

One of the reason behind failing projects is the fact that people don't do lot of reviews and when they find bugs in system testing phase just jump to code review. At this time it is not possible to fix all those issues in code.

People may think that what is impact of comments or naming conventions but poor code clarity, no comments are the reasons why at later stage even people who wrote the code can not understand code written by themselves and if some one else is working on the code written by others then it is a serious issue.

Advantages of code review tools
Tools not only review basic standards and conventions but they can also validate your design decisions whether implemented in code or not. Issues important for performance and memory can be caught through tools.
Most important is that full coverage of code review can only be possible in a big project and it saves your money too. If you have 800 Java files in your project you can understand manual review even if done for month can not cover 100% code. But tool takes hardly 10 to 15 minutes to review 700 Java files.

Tools available
Although many tools are there in the market for Java Code Review but I will mention only few which either I evaluated or got feedback from my teams, other folks.
Following is the list of Java Code Review Tools:-

Hammurapi
PMD
Check Style
Jalopy (It is a source code beautifier not a code review tool.)
I would like to mention here that selection of tool suitable for you is important but more important is use of tool and implementation of code review process.
Different tools have different advantages and are more suitable for specific use.

Hammurapi
1. Best thing about hammurapi is no dependency on any IDE and can be run as Ant task from any IDE.
2. Another very useful feature from project management point of view is that results of review are stored in a database and published through web application. This way whole project team and project manager can see the results and multiple runs of the project can be compared.
3. There is no requirement to compile it for review. To review code you need source files only not all the dependency libraries. It makes reviewer's life simple.
4. Time taken in a review is also not too high. With hammurapi version run on P-IV machine 800 it takes 11 minutes for review. This time may vary depending upon machine configuration. As per my experience its performance is good.
5. It provides review reports in two ways:- 1. Summary of the violations. 2. Individual code file with all violations on that code file. Here with hammurapi a developer can see all the violations on a source file written by him and by clicking on violation can directly go to the section of the html report (having javasource embedded) to see which piece of code is having problem.
6. In later versions of Hammurapi eclipse plugin is available only for development purposese but for commercial you it requires subscription (fee).

PMD
1. PMD is also a powerful tool and much talked about tool for Java code review.
2. Eclipse plugin is available so tool can be used from within IDE. It may help developers better by allowing them to see the violations on IDE and going directly to actual source like where problem is.
3. Web report is not at all available through a database store so you can not have results stored in a database to be compared later on.
4. You can generate off line html page of review report but source is not linked to it. You will have to use some tool to convert your java source files to html. Again here is a problem:- review report generated by PMD will have hyper links to the html of source but you can not navigate to source html by clicking on report likes because generated html from source will have extension .java.html
Solution:- You can use glean with Java2HTML to generate PMD report along with generation of source HTML. This way above mentioned problem can be avoided.
5. In PMD you can write your own rules either in Java or using XPath expression.
Both the tools provide similar default rules and new rules can be created.

With my experience hammurapi has one great advantage and that is:- results are stored in a database. Another good thing is it comes bundled with everything, only you need to do is run the installation. One thing that may not suite you in longer run is that its eclipse plugin for commerical use is licenced and you will have to pay for it. For some new review rules also they can charge but it was a plan only till last update I had.

PMD can better help developers in IDE because eclipse and other IDE plugins are freely available. If you are using gleen along with Java2HTML pages and you can take some pain to configure glean properties (feedback) file then PMD is most rich in terms of more rules available, ease of rules creation. Even results can be deployed in tomcat or any other java web container using an Ant task. See my post to see how it all can be done.

Check Style
Historically it was a code layout checker but with new versions it can check many aspects of Java code. There are standard and optional checks available. Plugins written by third parties for Eclipse/WSAD, NetBeans, Borland JBuilder, IntelliJ Idea and few other IDE's are available.

Jalopy
It is a source code beautifier not a code review tool. Even if it is not a code review tool I have listed it here to clarify this fact. Plugins for Eclipse, JDeveloper, JEdit are available.
Its commericial version is also available with more features.

Latest versons of Java code review tools
Hammurapi 5.7.0
PMD 4.2.5
Check Style 4.4
Jalopy 0.3.1 (not commercial)

Strategy for code review
Clearcut strategy and its implementation is most importnat for code reviews otherwise doing them is waste of time. Following things should be kept in mind:-
1. Plan for reviews. Identify reviews as a task in your project plan.
2. Keep practical time duration and effort for review and issue closure activity.
3. Setup code review tool early in the project life cycle. Before start of coding tool should be setup.
4. Let your developers know what are the things to keep in mind, means rules to take care of so that they should minimize injection of more violations. Also give them training of tool.
5. Ask developers to run the tool frequently (e.g. 2 times a day). Also run tool on complete code and publish it on central machine. For it you should setup a project website where along with code review results other important documents can be stored (e.g. html version of project plan, published design model etc). In most of the companies people don't have modeling tool license for every developer so publishing model as html is a good option.
6. Most important is closer of violations and it should be done on daily basis only then people will stop injecting more bugs in future.

Java Frameworks- Feature List

2008-03-26T08:43:00.000-07:00

Some time back I was trying to create a catalog of Java frameworks which can have some desired features for comparing various frameworks available. I had to stop that activity and unfortunately I was not able to pick it for a long time. Now I thought posting it in its current form is better then not posting it at all. Here I want to share idea on which I was thinking to work but didn't fill all the features for all the frameworks. What ever I put long back may also require validation from people who are in touch with specific frameworks. I invite all of you to let me know any information which can help filling empty cell related to individual frameworks.
This matrix is based on some features support or availability with the framework like :- built in support for caching, back button handling etc. Some description of individual cell is also available if you mouse over to any of the cell. This matrix is made in such a way that you can compare a feature on all the frameworks by looking at a feature column and this will help comparing frameworks for specific features.
This matrix may become very useful with your support so I request you all to let me know about blank cells information(means feature information) if you have for any framework.

Framework

Apache Click

Struts 2

Wicket

Aranea

ItsNat

IT Mill Toolkit

Seam

OpenXava

RIFE

Spring

Stripes

Tapestry

WebObjects

ztemplates

GWT

Ajile*

Wavemaker

Ajax

MVC

Push /Pull

i18n /l10n

ORM

Async /Batch Proc

Transaction

Testing

DB migr

Security

Templating

Back Button

Continuatio n

Caching

Validation

Releases

CRUD Gen

IDE

Portal

Scheduler

BPM

App Serv

EJB3

Reporting

Latest JDK

Act Coomu

Docu

Succ
stories

Jobs

%
Jobs growth

%
Jobs

Pull

1.4

Both

1.5

184

12000

0.0175

Pull

>=1.4

122

2750

0.0065

Pull

Push

2M*

>=1.4

Pull

>=1.4

Pull

>=1.5

133

1750

0.0075

>=1.5

Both

>=1.4

6152

2500

0.375

Push

150

0.0075

Pull

>=1.4

157

100

0.015

Both

Push

391

17500

0.025

Both

Push

Open Source - Innovations and a bitter truth

2007-05-14T16:24:00.000-07:00

Aim behind open source initiatives and community was to make standards and develop software on open standards. Benefit in open standard software was portability, re usability and elimination of vendor dependency.

Lot many open source initiatives and projects came in last decade. Most of them were really great initiatives. These projects set an example how community can drive the technology and what are the benefits of open source.

Dark side

Previously maturity of open source tools was a problem. Support, proper documentation and trained workforce were also some issues raised by people but these are non issues now a days with many open source software tools available.
Issues now a days are- it is becoming incompatible, non reusable, non portable, proprietary and most importantly expensive.

Reason

When open source was started large number of vendors were also involved in it and were promoting it, but they were promoting it to fight with another proprietary non open source house (you can easily guess who). At that time these vendors (supporter of open source) were interested in selling hardware. Now when open source became mature, great and worth selling, these vendors are trying to sell open source means they want to earn money from it.

Vendors have made "Java" language itself incompatible on various platforms because every one implemented JVM according to his own wish. Idea behind letting others implement JVM was to promote better quality of JVM but strict implementation of specification was expected so that anything developed on Java can be ported on any platform and JVM.
Now they have made their own JVM and if you have written some application on a specific vendors JVM sometimes you will feel that now it is risky to move to other. Few years back it happened but JVM is not an issue now a days. Issue is what happened next.....

J2EE specification was a great promise to the business community that once an application is written it can be ported to any J2EE compliant application server.
Now every vendor has written J2EE application server according to his proprietary libraries and they only promote those proprietary things.

What is the disadvantage

Most promising thing about open source (Java) is diluted that is "write once run anywhere".

New tools are coming every day and developers have to learn tools every day. Organizations ask them:- "did you work on this (xyz Java Application Server)) server?" How funny about open standards?
So you don't have large number of common work force but have specific folks and then they become costlier not only because they ask for more money but because you don't have sufficient numbers of developers at right time.

Now next dangerous step vendors are moving towards is:- making hit open source software proprietary by imposing worst licensing. Either vendors have acquired these open source organizations or these organizations themselves became money friendly vendors.

Solution

Community should stand firmly to protect work done by community and save open source.

Vendors should realize that things should be on standards not on proprietary specifications. In manufacturing industry manufacturers don't use different specifications or standards and in fact various parts are standardized. You can use same tyres on so many models. Manufacturing or any other industry (either mechanical or electrical) are surviving and making life better by creating difference with quality, real value and benefit.

At least industry should finalize common specifications of core things. There is a huge scope of making the difference so basic things should not be used to make a difference.

If industry and these vendors will not think about it we will never have a mature industry like manufacturing industry and we will be fighting forever for compatibility and other issues.

Disclaimer

2007-05-01T14:23:00.000-07:00

All data and information provided on this site is for informational purposes only. Owner or authors of this blog make no representations as to accuracy, completeness, currentness, suitability, or validity of any information on this site and will not be liable for any errors, omissions, or delays in this information or any losses, injuries, or damages arising from its display or use. All information is provided on an as-is basis.

While every caution has been taken to provide readers with most accurate information and honest analysis, please use your discretion before taking any decisions based on the information in this blog. Author or Owner of this blog will not compensate you in any way whatsoever if you ever happen to suffer a loss/inconvenience/damage because of/while making use of information in this blog.

The opinions expressed here represent author's own and not those of their employer.

In addition, our (author's) thoughts and opinions change from time to time…We consider this a necessary consequence of having an open mind. This weblog is intended to provide a semi-permanent point in time snapshot and manifestation of the various memes running around individual's brain, and as such any thoughts and opinions expressed within out-of-date posts may not the same, nor even similar, to those we may hold today.

Author or owner of this blog may keep updating old posts on a regular basis and are not bound to explicitly state all corrections made.

The entire risk as to the results and the performance of the information is assumed by the user or reader, and in no event we will be liable for any consequential, incidental or direct damages suffered in the course of using the information in this blog.

Information on the blog is provided for general information only and should not be treated as a substitute for professional supervision or advice as circumstances will vary from project to project or person to person.

The use of Google, IBM, Sun, Oracle, JBoss are logos, trademarks or registered trademarks of their respective owners; in this site is for reference only and does not imply any connection or relationship between authors, owners of this blog and these companies.
All trademarks remain property of their respective holders, and are used only to directly describe the products or services being provided. Their use in no way indicates any relationship between authors, owner of this blog/site and the holders of said trademarks.