Use a spade to dig a hole – Use a bulldozer to dig a trench

In this interview, Andrey Grebnev tells us why he thinks AJAX is more hype than substance. He talks about the Java technologies that he is excited about and also gives us an overview of the IT scene in Russia.
We all read reports saying that Russia is changing rapidly and that it will be an IT force to reckon with in the near future. However when it comes to ground realities in Russia and the kind of software development and Java work happening in the country, most of us know very little.
We speak to Andrey about IT in Russia, how offshoring works in the country as well as his exploits in the Java world. Andrey also gives us an overview of the opensource AtLeap CMS and tells us when to go for a PHP CMS and when will a Java CMS be a better pick.

IndicThreads >> Hi Andrey! Thanks for being on IndicThreads. Could you introduce yourself to our readers?

Andrey Grebnev >> Hi, and thanks for inviting me. Before Russian custom software development company Blandware all my career was concerned with offshore IT outsourcing too: remote developer in big Russian company, Project Manager in small Russian company. Then for some time I was in Intel Corp. research and development center. Then I was General Manager of Russian production unit of Spanish company. Simultaneously during several years I give IT lectures in Udmurtia State University. As you can see, my career is so many-sided.

IndicThreads >> Can you tell us a little more about how IT and software development in Russia works? Do you think the scene in Russia is different from that in the other BRIC (Brazil, Russia, India and China) countries?

Andrey Grebnev >> I was at conference ?Teaching IT in Russia? one month ago. This conference musters representatives of big Russian IT companies and representatives of world-famous corporations (Intel, Microsoft, IBM, Apple) and leading Russian Universities. The conference was under the aegis of Information and Computer Industry Association (APKIT). Association unites more than 200 companies and they are hold up to 70% in a number of sectors of IT market. Sometimes I will give you some latest data from discussions and reports of this conference including report of Byahov O.V. (Ministry of IT and Communication).

The sales volume of our IT market is about 0.7% of world one (and 0.3% added cost). However the growth of internal Russian IT market is about 20-25% annually. Russia intends to invest around $650 million in the IT sector over the next five years to be among the world’s top 10 countries in IT by 2010. This year the government of Russia has taken a decision to create several IT techno-parks with tax breaks and reduced customs on imported equipment for companies operating from there.

“Russia intends to invest in the IT sector
to be one of the world’s top 10 countries in IT by 2010…

There is a very big percent of educated people in Russia (including high education). We are always proud of our high education. The number of institutes of higher education has been doubled last 10 years. For example our students are always in the top of ACM International Collegiate Programming Contest. I think it is very big index of Russian education to have first places among 1.5 thousands universities from about 70 countries.

Our education graduates 180 thousands of students a year which have deepest fundamental IT knowledge including perfect mathematical base. I mean applied mathematics, formal logic, theory of algorithms, discrete mathematics and so on. Their knowledge will not become antiquated after 5-6 years (it is IT cycle). However Russia does not have debugged system of ?two-years? colleges (as for example in India) which graduate students which have knowledge in one concrete technology or programming language and are ready to work in real projects. Russia has one of the highest percent engineers in the world.

“Russia graduates 180 thousand students a year,
each having deep fundamental IT knowledge.…

For example it is not accident that Intel Corp. has been working in Russia from 1991. Now Intel has research and development centers in 5 Russian?s cities with 1 thousand employees. The high level of Computer science in Russia allows Intel to carry out high-intellect research in most difficult areas: High Performance Computing, Network and Wireless technologies, Microprocessors and Compilers etc.

As for concrete technologies, for example Java, Sun Microsystems opened its first Center of High Technologies in Saint-Petersburg city last year. About 300 employees work at Sun Studio (Sun Studio Enterprise, Java Studio Creator), J2SE, J2ME, Jini technologies.

Other world-famous companies have their R&D centers in the country, e.g. Motorola, Dell, Siemens etc. Boeing has the largest development center outside of US in Russia.

For all who are interested in Russian IT-market I can advise non-profit portal Outsoursing-Russia.com dedicated to outsourcing of software development to Russia. In May Outsourcing-Russia.com has combined forces with largest IT association of Russia ? RUSSOFT Association. RUSSOFT unites 80+ companies with more than 6000 highly qualified, professional software engineers. The president of RUSSOFT Association is Valentin Makarov. I have discussed the problems of IT companies in non-central cities just some weeks ago at dinner with him.

IndicThreads >> Many believe that the future of all software development lies in the BRIC countries. Do you see this happening?

Andrey Grebnev >> As for about production of typical standard software I think yes, the BRIC will have the most pie slice of IT market (especially offshore IT outsourcing). The main reasons are big population, the perfect physico-mathematical education (or its remainders) and the difference between common weal of BRIC countries and, for example, European countries.

As for prominent open-source software the genius can be born in every country and almost all countries have IT infrastructure to allow showing yourself. As for world famous IT leading companies I do not think that the focus can be shifted from say USA and Europe to BRIC within the next few years.

“Every BRIC country tries to grab something to oneself
from USA and Europe companies.…

IndicThreads >> Some experts believe that the BRIC countries should instead of competing with each other, learn to specialize and collaborate. Have you explored such possibilities?

Andrey Grebnev >> The competition was and it will be. The competition is environment to provide normal evolution of every market including IT one. Every BRIC country tries to grab something to oneself from USA and Europe companies (I mean offshore). However now I know the examples of collaboration between Russian and Indian IT companies. The Russian analysts design software and Indian ?two-years? coders implement it. The Russia has professionals to design some algorithms and India has enough quantity of professionals to implement it quickly. Why not?

“Russian analysts design software and Indian ?two-years? coders implement it.…

IndicThreads >> Any major problems that you see with how offshore models work today? Are you worried by all the bad press that outsourcing gets these days?

Andrey Grebnev >> I do not think that offshore is evil which we should to exterminate. Consumers and suppliers have always been in the animate nature. There are countries which supply natural resources (e.g. oil, gas) and there are countries which consume them. Why can not we have countries which supply IT services? I think offshore is better than drain-away.

“I do not think that offshore is evil…Why can we not have countries that supply IT services?.…

IndicThreads >> Coming to Java, you have been actively developing a new Java CMS. Tell us a little more about it.

Andrey Grebnev >> Blandware AtLeap is a multilingual free Java CMS (Content Management System) with full-text search engine. Blandware AtLeap is a framework which allows you to rapidly start your own Web application.

“Blandware AtLeap is a multilingual
free Java CMS (Content Management System)
with full-text search engine.…

The idea of the AtLeap project is based on my many years? experience of management in the area of the site development. Blandware AtLeap is destined to creating both content-based sites and web applications with complicated business logic.

IndicThreads >> Why did you feel the need to develop a new open source CMS, instead of using and enriching existing open source CMS?

Andrey Grebnev >> The development process of Blandware AtLeap was begun in 2004. That time we needed a particular base to develop web applications for our customers. These applications had to combine both complicated business logic and original, exceptional graphical design. In addition to it they had to support multilingual content. On one hand these tasks required multilingual CMS which does not restrict graphical design (CMS like Nukes was rejected). On the other hand such CMS had to be enough extendable to allow us to implement specific functionality.

PHP CMS were rejected as we needed to implement SOAP, cryptography and transactions. We did not want to teach developers to proprietary technologies of specific CMS. We did not want to spend money for knowledge which cannot be used in some other areas. That?s why we did not choose well-known open source java CMS OpenCMS. Moreover OpenCMS did not have multilingual full text search. Besides the customer do not want to put much money to the project at one stroke. The web application should be easy portable from open-source DBMS to commercial one, from open-source servlet-container to commercial one.

“We understood that there are no
free open-source multilingual extendable portable java CMS
with full text search which satisfy our general requirements.…

As result we struck out all CMS candidates from our list. We understood that there are no free open-source multilingual extendable portable java CMS with full text search which satisfy our general requirements. That?s why we made a decision to develop our own CMS and we have done it.

IndicThreads >> How does AtLeap compare with the established Java CMS like JBoss Nukes or Apache Lenya?

Andrey Grebnev >> *Laughing* The good wine needs no bush! I only say about disadvantages. Now Blandware AtLeap does not have so big developers? community and noted name. But it is a matter of time.

First of all Blandware AtLeap is a framework which allows you to rapidly start your own Web application and on the back burner this framework includes CMS functionality. AtLeap is not ?out-of-box? for beginners. AtLeap is destined for professional developers to ease their work.

“AtLeap is not ?out-of-box? for beginners.
AtLeap is destined for professional developers to ease their work….

As for JBoss, it has more interesting project JBoss Portal. But now it is in its infancy. However AtLeap does not intend to compete with JBoss Portal. They lie in different niches. JBoss Portal will be able to run only under heavy J2EE application server JBoss in contrast AtLeap can be run on lightweight servlet-container e.g. Tomcat, Resin and also under the application servers e.g. JBoss.

IndicThreads >> PHP CMS seem to be more popular than Java CMS today. Why do you think is that the case?

Andrey Grebnev >> First of all PHP CMS have more long time period of evolution. PHP CMS is easier to setup, it requires lesser hardware resources, the PHP hosting is cheaper, the PHP developers are cheaper and at last PHP is easier to study.

For example if you find PHP CMS which satisfies all your requirements and you do not plan to extend its functionally greatly you can use it without hesitation. Why not?

IndicThreads >> In your comment to the article Can Java CMS match the PHP ones?, you say that ?PHP CMS is suitable for simple graphical web sites (usually for projects less than 100 man/hours). Java CMS is suitable for web applications with complicated business logic (usually for projects between 100 and 1000 man/hours). Can you give us your experiences that made you come to the conclusion?

Andrey Grebnev >> When I said about 100 and 1000 man/hours I meant time which you should spend to develop functionality which is not provided by the CMS and its plugins. The difference between PHP CMS and Java CMS is not directly in programming language. The both programming languages allow to write bad and good software. The difference is in human factor.

“Both Java and PHP allow you to write bad and good software…

The reason is concealed in difference of skill levels of developers. For example if you pass PHP certification tests you can have no knowledge about design patterns. But if you pass J2EE certification test you will have perfect notions about the development process of difficult enterprise systems. As a result usually the certificated Java developer is more skilled than PHP one. Java implies complexity of multi-tier enterprise systems. Java encourages developers to write accurate code, use design patterns and so on. I teach students several years I know it.

“Usually the certificated Java developer is more skilled than the PHP one…

If you hire experienced certificated Java developer you will be sure that he can produce software which will be more maintainable and reliable.

On the other hand why should I contrive multi-tiers in PHP from e.g. PEAR::DB_DataObject and Smarty if there are Hibernate, Spring, Struts and so on? Analogous situation is in PHP CMS and Java CMS area.

As result we understand that if we do not need multi-tiers and our application does not require something specialized we can boldly use PHP. Otherwise we are forced to use Java in spite of high price of Java developers. But we know that these charges will be recompensed during maintenance.

“Use a spade to dig a small hole – Use a bulldozer to dig a trench…”

Java is not a panacea. In order to dig a small hole it is easier and quicker to use small tool ? spade. If you need to dig a trench you are forced to use bulldozer.

IndicThreads >> Which Java technologies of today are you particularly excited about?

Andrey Grebnev >> I am waiting for appearing Spring 1.3 with WebFlow and implementation of EJB3.0 very excitedly.

“I am excited about Spring 1.3 with WebFlow and EJB3.0…”

The WebFlow is the thing which has been waited on open-source market for a long time. I think in complete with SpringIDE we will get the perfect tool to model business processes in web applications. It can really help to separate work of web analysts and coder.

As for EJB3.0, many people wait when EJB2.0 (2.1) with its absolutely not flexible EJB-QL will be displaced. It is the main reason why Blandware AtLeap does not use EJB.

There is one more thing to be substituted. It is a little bit ?crooked? XDoclet. The Java Community is waiting expansion of Java 5 annotations. It is future. After AtLeap 0.5 release we plan to begin development of version 1.0 based on technologies described above.

“Java annotations is the future…”

IndicThreads >> And which technologies do you think are more hype than substance?

Andrey Grebnev >> I think the example of such phenomenon can be AJAX. Ajax is shorthand for Asynchronous JavaScript + XML. Ajax is not technology; it is simple new name for old well-known technologies. In short by event the browser can dynamically download new data into some node of DOM. This way the page may not be reloaded as a whole in order to change some small part of page. This approach is suitable for web applications which may not support old or text browsers and the application may not be indexed by external search engines.

“AJAX is more hype than substance…”

The article of Jesse James Garrett touches off the big resonance in community. But technologies are used by Ajax is comparatively old. And the idea is not new too. The approaches of using hidden FRAME or hidden IFRAME in order to dynamically download new data are old as the world is. However this article impulse new open-source projects and new features in the existing projects and it is very well.

IndicThreads >> Thanks Andrey. It was good to know more about the Russian IT industry and your work in the Java world.

Andrey Grebnev >> You?re welcome!

IndicThreads >> Readers can stay in touch with Andrey by email: andrey dot grebnev at blandware dot com