If you look at any list of open source content management systems, you'll find
Joomla right in top.  In this interview, we talk with Louis Landry, who
started as the
principal architect of Joomla 1.5 and the Joomla Framework and today acts
as the development coordinator.  If you're interested in where Joomla's at,
and where it's going, read on…

• Joomla and its relationship to website owners
• The difficulty of creating success metrics for free software
• The relationship between the core Joomla environment and extensions
• Cloud computing and platform independence
• Transitioning the web to new platforms and the semantic web
• The future of rich applications in a web browser

Scott Swigart: To start, could you introduce yourself and tell us a bit about your relationship to Joomla?

Louis Landry: Sure. I first used the predecessor to Joomla, Mambo, in 2001 or 2002. I was a computer science student at the time, and I was building a website for my friend. I sort of fell in love with the interface, and honestly, I was shocked to see some of the things that they had pulled off on the web.

Over time, I continued to build more and more sites and get more and more involved. At the time that Joomla became Joomla, I was in contact with several of the key members who started a fork to do some development patches and things into the core.

After I was asked to join the core team, I eventually served as project manager, communications team leader, foundation working group coordinator, development coordinator, and pretty much every aspect of the project. I am now development coordinator, working on future research and development, infrastructure, and things like that.

Scott: Talk a little bit about Joomla itself, the strengths of its architecture, and so on.

Louis: Joomla’s an open source content management system, designed to help you build and maintain your web presence. It doesn’t take the same approach as some of the older, larger-scaled content managers like SharePoint do.

It’s not designed to be a full-on enterprise thing. That doesn’t mean it can’t be used in enterprises, but it’s more of a presentation and content organization system than some of the more document-centric systems. Our main target market has been websites for small businesses and other small organizations.

The last reasonable numbers we saw showed maybe one or two million websites out there running Joomla. Most of those are individuals or small organizations putting up sites to share a voice related to some things they care about.

Joomla itself is really a platform with something like 4,000 extensions. It’s much more of a product based system than other applications like Drupal, where all of the extensions tend to be much more tightly integrated with the core system. With Joomla, you find a lot more third-party developers building product-based things. It’s much more packaged up, and you just sort of upload a Zip file and everything installs itself.

Scott: How is the Joomla organization structured? How do you make money, or do you?

Louis: The non-profit organization that holds assets for the Joomla project is called Open Source Matters. It’s incorporated in New York, and it holds the trademarks, copyrights, and those sorts of things. Most of their income comes from Google Ads, I believe.

It’s not really a money-making corporation. I am an owner in a company called JXtended, which sells third-party components, extensions to Joomla, and that sort of pays my bills, along with consulting work that I do around Joomla. That tends to be the trend with people around the Joomla project, if they’re making money here.

Scott: That makes sense. I would imagine that a lot of people are looking for certain functionality, and they may not even know that Joomla’s the solution. Then a consultant can come in and relatively quickly have a site up that meets their needs and just so happens to be based on Joomla.

Louis: That’s actually the case more often than not. People will hire a firm to put together a website for them, and they don’t really realize that Joomla is the platform on which it’s being built. A very large proportion of the people on our support forums and engaging in the Joomla project are small web design shops that are pumping out websites for different clients.

We’ve had more than one case of a client seeing the Joomla offline screen and thinking that we somehow hacked their site. Something went wrong with their Mind Skill server, and when the little screen comes up and says, “This site’s offline,” and it’s got the Joomla logo, they’ll email us and say, “Why did you hack my site?”

[laughter]

Then we have to explain to them, “No, really that’s not what happened.” So it’s funny, but it’s happened quite a few times. That’s indicative of the fact that a lot of people using Joomla don’t really think of it in those terms, since the site was put together for them, and Joomla is transparent to them.

Scott: That’s an interesting aspect of that space. There are obviously a lot of people who need a web presence but are not technical.

You’ve also mentioned the really rich ecosystem, for lack of a better term, around Joomla: 4,000 plug ins, consultants, developers, web design firms, and so on. One to two million sites must rank as a highly successful open source project.

Louis: It’s a pretty sizeable chunk. I think since July, we’ve had three million downloads of Joomla off of our sites. I’d have to double-check the numbers, but I think in the last two years, it’s in the range of 14 or 15 million. It’s been pretty astonishing to see the downloads crank up.

That’s also just the English version. There are plenty of versions that are distributed by our language partners or translation partners all over the world in something like 30 or 35 languages.

Scott: That’s an interesting difference between success metrics with proprietary software as opposed to open source. With proprietary solutions, you can point to an exact number of licenses sold, whereas with open source projects, you often have to rely on the number of downloads. That makes it somewhat more difficult to get an accurate picture of adoption.

Louis: That’s true. It’s difficult to gauge the ratio of deployed sites versus the number of download packages, so it’s really hard for us to know what the actual market share is.

There were some numbers at one point from a company that I believe was called Netcraft, and some people in our community have built crawlers to try to figure out market penetration. It’s been interesting to see these various numbers, and they vary widely.

Scott: It kind of goes against the ethos of open source, but open source projects could easily embed something in them that by default makes the thing phone home, when it’s up and running.

Louis: There’s been a lot of talk about that over the years. A few open source projects have tried it, but those efforts tend to backfire, because a lot of users don’t like it.

We’ve talked about the concept of anonymous usage statistics and those sorts of things several times, but it’s just never happened. I am actually not sure it would be a good idea, simply on the basis of the fact that I have no idea what sort of traffic that would generate for the server collecting the data.

Scott: Right; a little Google App Engine app could just shove the statistics into a big table data store.

Louis: Yeah, let Google suffer. [laughs]

It’s something that we’ve talked about, like having some sort of optional switch upon install, but even there you run into trouble, because one of the biggest ways Joomla is installed is via cPanel and Fantastico, which are server scripts to auto install web applications.

They don’t even run though the Joomla installer, so we wouldn’t see it. That’s another example of how our download statistics would be completely irrelevant, because that doesn’t get downloaded through us–it gets deployed through a completely separate system.

Scott: On the topic of the cPanel and Fantastico installers, talk a little bit about how Joomla updates. There’s always a tension between a really vibrant ecosystem of plug ins and the necessity of keeping the main project moving forward.

Louis: That’s been a real area of balance for us as well. Anyone who follows Joomla knows that we don’t release very often, as far as feature incremental releases, and that’s not always by design. Version 1.5 came out, I think, two years after 1.0, and 1.6 is set to come out probably toward the end of this year, which is about two years or so, as well.

It moves rather slowly, but the wealth of functionality created through the third-party extensions allows us to do that. And because it’s almost entirely volunteer driven, and we’ve been going through growing pains since this is still a relatively new project, we’re still trying to find our footing in a lot of different areas.

We don’t really have an update mechanism in the core quite yet, although that is something that is being actively worked on for Version 1.6. As far as extensions are concerned, there isn’t a real updating mechanism for that either.

Historically, it’s been either a case of uninstall and install the new version, or else install a new version on top of it. With PHP, it’s generally as simple as FTP-ing files up, if you want to do it manually. It has not been painless, but it has also not been a huge problem for us.

Scott: I imagine that the 4,000 or so plug ins are in various different states at any given time, from having been abandoned to having multiple people working on them full time, and everything in between. It’s kind of like when you upgrade WordPress, you just kind of cross you fingers and push the button, and hope all your plug ins work afterwards.

Louis: [laughs] That’s the thing with the slower release cycle. We put maintenance releases out to fix bugs and things like that pretty regularly, but those updates don’t change things in such a way that any of the third-party stuff would break, really.

There have been a couple of cases where we’ve had to change things, but we got out in front of it pretty well, and it didn’t really cause a whole lot of issues. Still, it’s true that when you’ve got a large third-party developer community building external extensions, there’s an awful lot of inertia.

If we go in and change a system in the core that’s going to have a cascade effect on how everybody uses the API, or the systems, or something like that, that’s a lot of ramp up. What we saw with our huge changes in 1.0 to 1.5 was that it took a while for the third-party development community to catch up.

In a lot of cases, they did abandon things, and other people picked it up. That’s part of the beauty of the open source community, right? Somebody says “Oh, well I’m not going to bother with this,” and somebody else comes behind them and says “Oh, I’ll do it.”

Scott: Or, in some cases, there are just a lot of options. When something gets abandoned, there may be four other projects that do something pretty similar.

Louis: That’s certainly true. For example, in just the category on our extensions directory for adding comments to articles, there are pages of different options. Some are commercial, some are not, they have different features, and so on. The richness of choice is really great.

Scott: That brings up an interesting point. I’ve talked to people in corporate environments who are considering using open source software about their methodology for evaluating the safety of a certain project.

They will look at all those stats to see if it’s being actively maintained. They will see if the contributions are just coming from a few people, or from a lot of people. They often have a specific set of criteria they use to judge whether they want to make a bet on it.

Do you have any advice for people as they look at those several pages of options for similar functionality? Obviously, the one that shows up at the top of the list isn’t always the safest choice.

Louis: That’s a great note. I find that this is one of the areas where the commercial side of open source really flourishes.

If I were going out to build somebody’s website, and I was presented with a commercial choice and a non-commercial one where they both met my needs, I would probably choose the commercial one and pay for it. That would give me higher confidence that somebody would be backing it.

That’s not to say that I don’t appreciate non-commercial open source software, because obviously I do. Still, in a mission-critical situation, that’s likely the path I would take. That said, I also tend to look at how active the people are who are running the project.

If I’m looking at those same two options, and the commercial people are nowhere to be found but the people running the non-commercial project are very active, then that tells me a lot. I feel better about their commitment to the platform, as well as their knowledge of how everything works.

I know that they’re more able to integrate certain things. They’d probably be better suited to finding answers to problems, if I ran across them, and that sort of thing.

I think there’s a balance to be made there. I really like to be able to know that whoever I’m getting my software from is active in the open-source community.

Scott: What’s the Windows story for Joomla? Is it common to run it on Windows? Is it one of those situations where some people do development on Windows but they pretty much always deploy on Linux?

Louis: I think that since the majority of Internet servers are Linux, it’s probably safe to say that the majority of Joomla installs are on Linux. And since the majority of end users use Windows, it’s probably safe to say that most people that test it and build up sites and all those sort of things are doing it on Windows.

There are certainly people using Joomla on Windows servers. We officially support Apache, and we don’t support IIS. We know it generally works on IIS, but there have been some issues. We don’t have access to those servers and the software on a reliable basis to catch all of the problems that come up.

We’re actually in talks with Microsoft now to figure out ways in which we can advance our Windows support and make sure that we have the tools that we need so that everybody can more officially assert that Joomla works in a Windows environment.

Scott: What’s in that for you? Do you feel like that will open up a significant new population to becoming Joomla users?

Louis: I’m sure it does, even though on a personal level, I don’t really care that much because I don’t use Windows. I’m not one of those hate Microsoft guys, but I just use Mac OS and Ubuntu.

Still, our mission is to provide a flexible platform for digital publishing online. And what matters there is that I would like it to be available for as many platforms, OSs, and databases as possible.

Whatever we can do to make it easier for people to deploy Joomla, within reason, is something that we will want to pursue.

Scott: Cloud computing is obviously very fashionable at the moment. What interest have you seen around running Joomla on a cloud provider? Is there any work or any thinking about enhancing Joomla to take advantage of some of the attributes of the cloud, like being able to dynamically scale up and scale down and that kind of stuff?

Louis: There are a lot of people in the Joomla sphere looking into playing with the cloud, and there are a few companies I know of in the Joomla community who are actively building things on the cloud and trying to provide integrations there.

The EC2 thing at Amazon is interesting to me, and I’m also very interested in the Eucalyptus project that Ubuntu is starting with, I think, this next version of Ubuntu server. I’ve been talking with some friends in the FreeBSD project about cloud solutions and what a path might be, as well as a bunch of other places.

I think one of the things that I will focus my research on for the next generation of Joomla will be ways we can restructure Joomla from an architectural standpoint to better live in that space and be able to really reap the benefits of the cloud.

Scott: You mentioned Eucalyptus. I’ve heard that name, but I don’t know much about it. Can you talk about it?

Louis: I don’t know an awful lot of the technical details behind it, but Eucalyptus is an open source project that Ubuntu is adopting as sort of their cloud platform. It’s got a bunch of functions, such as the ability to do node requisitioning and various things, and I believe that it’s the software that would tie the servers in the cloud together.

Scott: What I’ve heard, if it’s the same thing, is that it also exposes an API that’s compatible with Amazon’s API. So even though you’re not running on Amazon’s infrastructure, I could write code that could spin up and down instances over in Amazon, but I could point that same code at something Eucalyptus-based and it would do effectively the same thing, even though the whole underlying virtual infrastructure is completely different.

Louis: That’s certainly an aspect of it, and that’s one of the reasons why it was really interesting to me. I also am really excited about the stuff I’m hearing coming out of the Ubuntu server crowd.

Scott: Talk a little bit about that, because it’s not something I’m aware of. In the hosting space, I think of CentOS as being really popular.

Louis: It certainly is. CentOS and Red Hat are probably the standard bearers. Ubuntu’s appeal really stems from being what I chose for my desktop OS. I like the way the file system and the configuration stuff are structured, I like the toolkits, and I like a lot of the things that Ubuntu is doing across that space.

The fact that they’re starting to focus on the server is interesting to me, and I really like that they’re thinking about cloud and moving into that direction, because it looks like it’s one of the focuses that that whole community is taking with the server distribution.

Scott: One aspect of that space that really stands out to me, and that I can’t really figure out, is that one of the earliest cloud providers was Google App Engine, and when it came out, there was a lot of excitement around it. And then people were kind of turned off because they didn’t want to be limited to Python, and then they added support for Java.

It seems like the most obvious thing for Google to support would be PHP, and the second they came out with Java support, some guy ran off and started a project that I can’t think of the name of at the moment, but it lets you run PHP on top of Java.

So this guy said, ” I’m going to port WordPress to run on top of PHP, on top of Java, on top of Google App Engine, using their BigTable, as kind of a science experiment.” He ran off and did it, and it works, but obviously, it’s probably not the stack you really want to build on with something real.

When I look at that, I think that if they supported PHP relatively quickly, a lot of PHP projects–WordPress and maybe Joomla and Drupal–would probably be supporting some version that could use a BigTable data store and run on Google App Engine.

Louis: I’m actually not certain that that’s Google’s vision of what App Engine is about. They’re certainly not a PHP shop in any way. I’ve got several friends that work at Google, and I’ve never heard of them using PHP in anything, unlike Yahoo!, which is a “PHP shop.” They do a lot of PHP stuff.

Google is a Python shop. Yahoo! hired Rasmus, Google hired the Python guys, and that’s the direction they took. I think when they launched it, as is typical for Google, it was closed. You had to be invited to be able to get into the App Engine experience.

I don’t think they were certain about their ability to scale it out if adoption got too high, and the ubiquity of PHP and Drupal and Joomla and WordPress might actually have been a negative for them. I’m not certain that signing on to host all those apps would really be to their benefit.

I would almost think that part of their rationale there was wanting to raise awareness around Python and maybe get some clever people working on Python to see who those people are. Then they could collaborate with them, or hire them, or at least get a better feel for what’s going on in the Python community and expand upon it.

Scott: To turn a bit toward the future, you mentioned that cloud is an area that you guys are thinking about. How does Joomla fit in there? How does it play on this new computing paradigm? What else do you feel are the big opportunities on the roadmap for some deep thinking and important work?

Louis: The obvious thought is the semantic web, which everybody is thinking about. I feel that we in the open source CMS community can play an important role in terms of providing new ways by which information can be organized.

I think the semantic web is a place where a relatively small amount of work could have a very large impact. Still, that focus is difficult, in the sense that I’m really more focused within the Joomla community on the next generation, not necessary iterative versions that are coming out.

I’m trying to rethink the architecture. Joomla as it exists now is based on the same principal concepts as Mambo in 2000, namely being centered around software designed to run on a $6 a month hosting account on some offshore server.

The concept around one site, one instance is still a very serious bread and butter piece of the market for us, but I also want to be thinking about ways in which you can scale out multiple instances running the same software. That ability will position us to fall much more inline with cloud and enterprise deployment schemes.

I want to think about how we can make data more portable, so that you can transport things across instances or just have protocols to push in and pull out data, as well as incorporating the ideas of the semantic web.

I also think a lot about this concept of the rich Internet, like Adobe Air and various other technologies like the high-powered JavaScript engines that are starting to evolve thanks to Google.

It would be really incredible to bring out that rich Internet environment and rich interfaces to make it easier to build a more interactive experience. Those are the things that I have been thinking about.

Scott: This has been a great interview. Do you have any closing thoughts from your end?

Louis: I think we’re really just growing accustomed to this whole new phase of the web, and I’m looking forward to seeing where it evolves. I don’t think that in five years it’s going to look at all how it looks today, and I don’t think the experiences are going to be the same.

I’m really excited to be a part of how this whole thing evolves, including what Microsoft and Google are working on with their Web OS initiatives. It could really change the ways people think about their computers and connectivity. I find all of these things converging and I can’t wait to see what happens.

Scott: I remember one guy saying basically that, until you can run Quake in a web browser, Google is not done. [laughs]

Louis: Actually, you can. Google released an experimental thing called Native Client. It’s not for public consumption yet, but you can compile Windows apps or whatever into this native client code, and then it will run in a browser with a plug in.

So it’s kind of like a virtual machine. It’s interesting what they are doing with that, and I’m very curious to see things like the plug in that they have built enable OpenGL in the browser via JavaScript. I’m curious to see how they integrate that into future stuff.

Their example is quick. You can click in the browser…

Scott: It sounds almost like an ActiveX plug in.

Louis: I think it probably is very similar to that concept, and hopefully they can find a way to make it more secure than that.

Scott: That’s one approach: running a whole different something in the browser, which was kind of the Flash and the SilverLight approach.

Louis: We would much more like to see it work in JavaScript, though.

Scott: It seems like there is also an awful lot of work going that route. These are just things that are mainstream use cases for HTML and JavaScript, and they’re very fast. It’s just a whole new set of scenarios.

Louis: The ability to access the machine-level function-called things like that within the browser would be just intense, but also intensively dangerous. I’m very curious about how we are going to resolve that, because at some point it is going to have to get resolved.

You’re going to have to be able to access the file system and things like that, and I just don’t know how we are going to do it.

Scott: That seems like a good place to wrap up. Thanks for taking the time to talk today.

Louis: Thank you.

Bookmark this: