News Stay informed about the latest enterprise technology news and product updates.

Maddog weighs in on the state of the Linux

Jon "Maddog" Hall is the executive director of Linux International, a nonprofit organization based in Amherst, N.H., that is comprised of developers and contributors worldwide.

Hall joined Linux International as a volunteer in 1991 and has been using Linux since 1994. He worked at Compaq Computer Corp. and Bell Laboratories and was a department head of computer science at Hartford State Technical College, where he received the nickname Maddog from his students.

In this interview, Hall spoke with about what he knows best -- Linux – and discussed integrated Linux applications in the enterprise, open source office suites and the future of enterprise Linux.

What's the current status and strength of major Linux-supporting vendors' enterprise applications stacks vs. the Microsoft stack? Can you offer some examples? Are Oracle and IBM going big in this area?
Various companies are going very strong in integrating their enterprise application suites onto Linux. A lot of this has come from the realization that they can not advocate to their customers the advantages of using F/OSS if they do not use it themselves. Therefore, internal drive has been given from companies like IBM, Oracle, Novell and others to move all of their employees onto F/OSS software. This also requires that a major portion, if not all, of the software that they produce also interact with F/OSS software.

Of course, this is possible since F/OSS software tends to follow standards, and when it deviates from standards, you have the source code to show you what it is doing. Therefore, you can integrate your own applications into F/OSS.

Can communications and e-mail applications running on Linux really make a dent in Microsoft's Exchange/Outlook reach?
We -- you and I -- live in a very strange world. We think that "computers are everywhere" -- but they are not. And a lot of people think that the desktop is all there is to computing.

But first of all, "sendmail" still drives most of the e-mail routing and delivery to most of the world. And a lot of the developing countries have not selected their operating system yet, or if they have, it is based on F/OSS, or the future direction is F/OSS.

There are a lot of countries that use "pirated" software. But pressure is being put on them to stop using pirated software and instead develop their own computer economy using F/OSS. In addition, they are looking out 10, 15, 50 years and saying that only F/OSS software gives them the security that the e-mail system they choose today will evolve, mature and still be around 100 years from now.

Read more

More maddog commentary: Linux piracy
Desktop apps are ripe for open source

Companies come, products go and even companies like the once-mighty Digital Equipment Corp. and Compaq are now dust in the wind … most of their products are only found on old advertising sheets.

What's the current status and strength of OSS stacks like Compiere? What other OSS stacks stand out?
What can I say? Compiere works across every reasonable operating system, is F/OSS and can be changed to meet the customer's needs. It may not have all of the things that people need today, but it is rapidly gaining both new features and new developers.

The closed-source model of development of software is dead for most mass-produced software. People will realize that when they start to try and remember the last time that they got any real service out of the company that they buy their software from. The ability of a top-down software engineering team to meet all the needs of all the customers is nil, and gets worse as the potential customer base grows.

How is open source development changing the enterprise application development world?
It is changing all of them by allowing the original focus of software -- service -- to be extended through an economical model of being able to pull down large chunks of code, then pay people to tailor it to their needs. Some customers will find that they can use the code they pull down without any additional tailoring – like OpenOffice, for example. But some people will find that if they change it just a little bit, they might be able to get more out of it. Those people will keep improving the software in an incremental way.

Sooner or later there may be another large OpenOffice project that will replace OpenOffice with another major step function in functionality. Customers using OpenOffice will have the choice of moving or maintaining the code as long as they want. Others may move to the new project. But it will be their choice, not forced on them by a manufacturer that is "retiring" the product.

What's the big story in enterprise Linux that's going to impact IT shops in the next year? Why?
Actually, I think the next big story is in the VoIP [voice over Internet Protocol] space, and it is called "Asterisk." Asterisk is a F/OSS project that allows a complete PBX system to be implemented on an inexpensive PC with a couple of peripheral cards inserted. It has most, if not all, of the features that large, incredibly expensive PBX systems have. But most importantly, it is F/OSS, which means that a company can integrate the rest of their software into the PBX. What's happened in enterprise Linux in the past year that surprised you? Why?
I continually am surprised by how fast it keeps moving forward and gaining momentum.

Proprietary companies are typically profit based. There are things that they do [or avoid] because it would cost them too much money, and they owe a profit to the stockholders. F/OSS people do not have those limitations. If a person has a strong enough itch, they can scratch it … and that makes for a stronger product.

What didn't happen in enterprise Linux in the past year that should have? Why?
Linus Torvalds was not elected president. Why? Because he is not a native born resident. But, you probably want me to be serious. I think that better acceptance of what the Free Standards Organization is trying to do, and support of that by more independent software vendors is something that should have happened. Even if people say that the current standard, V2.0, does not do everything it should, it is a good step forward, and people and companies should help them make it better. Anyone can complain; it takes an entity with vision to do something. What general trends are you seeing in the types of enterprise applications being developed for Linux platforms?
From the free and open source software (F/OSS) community, I see larger and larger projects being developed. Things like CRM [customer relationship management] and ERP [enterprise resource planning] systems are projects that are now being developed. Groupware projects are also being done in what has traditionally been called the ".ORG" space.

In addition, some of the distribution developers are developing extended frameworks to coexist with or to replace Microsoft Exchange.

Finally, some governments, like Brazil and Germany, are now showing interest in funding development of higher-level applications in the F/OSS space.

We -- you and I -- live in a very strange world. 
Why is it important to make strong, integrated application suites available for enterprise Linux environments? Can't IT shops just choose "best-of-breed" applications and use them as they wish in an enterprise?
Standards have long been a part of "tying things together." The standards of the IEEE and other standards organizations allow various companies to make products that can then interact. I can buy my resistors from any standard supplier and I know that they will work in my electrical circuit.

Interchangeable car parts allow automobile manufacturers to use any manufacturers' tires on their vehicles. And because the parts are interchangeable, we do not have to send our cars all the way back to the manufacturer just to get it fixed. I can also buy third-party add-on appliances for my car, things that the manufacturer of the car did not think were necessary, or did not think there was a large enough market for them to supply the part.

Does the fact that Microsoft has done a good job of tying its products together make it more difficult for enterprise IT shops to move off Microsoft and onto Linux? Don't integration problems ensue?
Microsoft does not do a "good job" of tying its products together. If they did, they would use open, published standards and interfaces to allow customers to use other components as desired.

Recently, I was in a shower in Korea, and the shower knobs were smooth and covered with soapy water. I could not turn on the cold water, and I was almost scalded. I suggested that the management go to the hardware store and buy new handles for the faucets, ones with rougher knobs. This would be possible due to the standards in the plumbing industry that set the size of pipes, screw pitch, etc., and allow you to put several different styles of knobs on the end of the pipes. If Microsoft had built the hotel, they would have to rip out the entire piping system, all the way back to the water supply on the street and replace the entire thing, just to allow me to use a different manufacturer's knobs.

To bring this back to software, the recent proclamations of several security groups that Internet Explorer is not safe from security attacks, enforced by the recognition that the F/OSS community is creating better browsers has driven some people to try to replace IE with Firefox. Many people have had problems in fitting Firefox into their Microsoft suite because of Microsoft's "integration."

The closed-source model of development of software is dead for most mass-produced software. 

Dig Deeper on Linux servers

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.