About two years ago, the LAMP stack was just a hobby for enterprise level developers and Java was the language of choice. Oh, how things change. Today, Java is still prevalent, but LAMP is quickly building up momentum that it can attribute to its open source roots and a desire among developers to find a more flexible, low-cost environment in which to work. As developers drive adoption, the c-level decision makers are taking notice, too, says ActiveGrid founder and CEO Peter Yared.
Yared, who recently spoke with SearchEnterpriseLinux.com following a successful showing at last month's LinuxWorld, talks about the rising popularity of LAMP, the move toward service-oriented architecture (SOA) environments and how Darwinism has taken hold in the world of software.
As an enterprise LAMP vendor, what are your impressions of SQL Server Express, which some see as Microsoft's answer to the rising popularity of LAMP?
Peter Yared: Microsoft, in some ways, pushes LAMP more than anybody. They endorse scripting languages and have done so for years. They make it easy for developers to develop applications on them, and run them on x86 machines, while being very aggressive about pushing XML standards. They have definitely given the market a sense that people are moving to [LAMP], and their reflexive response to that is to lower prices. They did it with Oracle way back when Oracle wanted SQL Server in Windows Server. The symbiotic relationship is there;
In your presentations you mention a "LAMP Mission." What is the next direction for LAMP? What is it capable of, what is your company doing to continue to drive adoption?
Yared: We have a product today that is enterprise grade. When we started two years ago, enterprise customers didn't know what LAMP was, and even six months ago people were not sure, but now when we go in we have two guys dedicated to evaluating LAMP. The enterprise is really looking at it seriously; they understand the benefits of it. I think we're finally at a point where we are seeing a lot more LAMP applications established in the enterprise, and, of course, we're seeing a ton of them on the open Internet.
People have hit a point where Java -- while still great -- has become very fragmented. You could get into Hibernate, Struts -- what kind do you do? It's hard to build on Java; you need pretty educated people, so people start to look to alternatives.
The big thing is that a lot of enterprises are moving to SOA environments, and in that environment it doesn't matter what the endpoint is written in. They all operate at the service level very easily, and I think that's what is opening opportunities for new technologies in the enterprise environment.
You pitch LAMP as a benefit to all aspects of an enterprise: the developer, chief information officer and CEO alike. Why is it necessary to appeal to each of these segments of a business when promoting LAMP?
Yared: Nowadays we have to appeal to all of them to get something on board. The days of people doing what they want are over. Developers have to get technology they like, and it has got to be cost effective and meet business needs. Saying that today sounds obvious, but it was not the way people sold technology for many years; there was a lot of tail wagging the dog. There is a sense of a lot of maturity within these organizations today, and been very heartening.
What are some of the Web development pitfalls that you have sought to fix with this new release?
Yared: The development pitfalls with LAMP today are that it is a bunch of different pieces, and developers start with a blank screen. As a developer, you begin a new PHP file and then you have to start typing. There are lots of examples, it's so basic, and there are many avenues to get in with this technology. A developer can go to SpikeSource or Source Labs and download the whole stack, but then again you are on your own when coding.
What is powering Web 2.0? Why the slide away from Big Iron and J2EE [Java 2 Platform, Enterprise Edition]?
Yared: The software stack lacks enterprise features that enterprises are familiar with. With Java 10 years ago, the design goal for enterprises was to scale up, and that was the design goal for everybody. Now everyone wants to scale out. You can still scale out with Java, but it's hard. It is not what it was meant to do. In the meantime, Java took a long time to open source and even then it's still not really open source. Homegrown developers went off and did their own stuff and fine-tuned it.
It's 'software Darwinism,' and here the best things win out over time, especially with open source. With the old way of doing things, the big vendors would come together and compete on implementation.
Is the company size a problem in getting grid out to everyone who needs it?
Yared: One of things different with open source is that as long as a product has traction, the company itself doesn't really matter. We had very strong focus on standards and open source computing and providing open source additions. If you do those things right, you might not have the security of an IBM, but you can offer an alternative with the potential to attract capital.
This article originally appeared on SearchEnterpriseLinux.com.