I recently contributed a story to Rodrigues & Urlocker’s Open Sources column on InfoWorld that explores the following question:

“The period from 2000 to 2002, when the economy was almost as terrifying as it is now, was the largest period of growth for open source in the enterprise. Is something similar happening again today?”

The conclusion is nuanced, sadly. There will be no chest beating and declarations of victory quite yet, but there are still good reasons to be optimistic. Read more over at InfoWorld.

We recently ran a poll on SourceForge.net, asking folks which languages they though would do well or start to die out in 2009. Interestingly enough, most of the more popular languages were named to both lists; survey takers were ambiguous about whether languages such as Slovensky, Ruby, Python, PHP, PERL, COBOL, Javascript, Java, C++, C#, C, and .NET would either rise or fall in popularity over the upcoming year.

AJAX, Object C, JavaFX and actionscript were cited as languages that might grow the most in 2009, and when asked which programming languages would fall out of favor, respondents predicted declines in asp, BASIC, Visual Basic, html, FORTRAN and, cheekily, “yours” (we certainly hope that isn’t the case).

Update: Someone just pointed out to me that AJAX isn’t really a programming language.

What do you think? What language do you expect will do well or start to die in 2009?

Things seem particularly contentious this election year, and everyone can only really agree on one thing: there’s an awful lot at stake. Are you nervous? We are too. To settle the debate, we polled over 250 random SourceForge.net and Slashdot users on a wide range of political issues…because, after all, you guys are not known for mincing your words. We figured that if anyone can get to the bottom of this, you can.

I guess what I’m trying to say is that there’s no need to wait until next week because the results are in! The open source community has spoken.

Our polls showed Obama with an almost 2:1 edge over McCain in the US, but internationally the ratio was a staggering 37:1.

But there’s more.

(more…)

It’s been a few months since we started rolling out our new UI, and we’ve completed the front page, the project summary page, the download file listing page (which we call the “showfiles page”), and the page that shows after the file starts downloading (which we call the “thank you” page.) In essence, we’ve redone the most highly visited pages, the ones for downloaders.

You may be wondering why. I think the most important goal is to provide a simpler user experience to the 95% of our audience that just want their damn files so they can get on with their lives. SourceForge.net is a complex site with lots of hidden nooks and crannies; downloaders are here with a clear purpose and need to be shown a clear path.

Naturally, there’s also a branding aspect to the new UI. We want people to know that they’re downloading from SourceForge.net, and that they, too, can create a project here. We wanted our message to be clear and strong, but not to get in people’s way.

The same is true for the messages from our advertisers. Yes, they pay the bills around here and, yes, we can’t exist without them, but we have to be a bit intelligent about it…an ad experience that pisses everyone off is useless to advertisers, too. Moving things around a bit allows us to show higher quality advertising at higher prices…and, therefore, to show less of it.

There are a lot of strong opinions in our feedback forum, and while I think that venue is probably disproportionally negative, that feedback is extremely valuable to us and we’re tweaking our design based on it. I’ve been keeping an eye for any mention of SourceForge.net on Twitter and it feels like about 40% love it, 40% don’t like it, and 20% aren’t really sure what their opinion is (even though they feel strongly enough to express it.)

What do you think?

Our good friend Joe Brockmeier, community manager for openSUSE, has just started blogging for ZDnet. In one of his inaugural posts, he ruminates over where a community manager belongs in corporate structure: engineering or marketing? His post was in response to Stormy Peters, who thinks the support team is a good place. As a fellow community manager, these posts are a fantastic opportunity for me to talk about a subject that’s near and dear to me: me.

I report to the VP of Marketing, and I won’t be shy about telling you that’s caused me a bit of stress. As an engineer, I always thought that marketing was just the art of yelling at the top of your lungs about some crappy product to people who don’t care, but that’s only true in organizations that don’t understand marketing. Marketing departments generally fall into two categories: ones that shout at the market and and ones that talk with it. The defining difference is whether marketing is an inspiration for the company’s strategy or merely its slave.

If my job were to coordinate bug fixes with upstream providers or negotiate solutions to interoperability issues (like it probably is for Joe Brockmeier from Novell), it’d probably make sense to think of it as an engineering discipline. If it were to find ways for developers to contribute to our application or govern the ones that already do (as it probably is for someone like Stormy at Gnome) it might be more appropriate under a product management team.

However, we’re dealing with a community of scale where our role is to provide the environment where people can produce - not to coordinate their actual production. So community management at SourceForge.net is about observing the community as a whole, engaging as much as we can, and tweaking the platform so that people can get more done. Since our marketing team is a “talking” team and not a “shouting” one, community management fits there nicely.

I was looking through Slashdot’s polls yesterday, and noticed their politically-charged poll “How many homes do you own?”, ostensibly regarding McCain’s recent difficulty remembering the answer to that very question. As funny as that is, the leaderboard advertisement that (presumably) Google decided to run was even funnier: a message from John McCain 2008.

I’m pretty doubtful that John McCain’s campaign has selectively targeted pages where users are being polled about how many houses they own. It’s probably more likely that he’s targeting Slashdot as a whole, or polls in general. Either way, I think Google has (completely by accident) made a point it probably wasn’t trying to.

Many of you have already read the news: Yesterday, the Court of Appeals for the Federal Circuit ruled that, yes, a software company has infringed upon an open source project’s copyright by failing to adhere to the terms of its license. Specifically, the Artistic License.

If you keep up with what’s been going on, you can skip this paragraph. KAM Industries sells software that powers model railroads, and used some technology from the SourceForge.net project JMRI in its Decoder Commander product. Last May, a judge ruled that KAM Industries did, in fact, violate the terms of the Artistic License…but since it was only a breach of contract it wasn’t a big enough deal to do anything about. Yesterday’s ruling found that it was also in violation of the original copyright, which is almost certainly enough to convince a judge to make them stop. I’m not a lawyer, but that sounds about right to me.

There was much rejoicing by Dana Blankenhorn at ZDNet and Matt Asay at CNet, who considered it a great victory for open source. I think it is too. PJ at Groklaw had a different spin, though. She believes that the open source community pretty much dodged a bullet with this judgement, and needs to be a bit more savvy about license creation in the future.

But let’s be fair about that. It’s kind of hard to be savvy about licenses when there are so many to choose from that developers are unlikely to read them all (and even more unlikely to understand them.) I think a lot of us dreamily reminisce about the days when we could just code and not worry about contracts, injunctions, or copyright laws. Will it ever be that way again?

No, I think the best we can hope for now is complete victory: a set of licenses, clearly defined and understandable, and an international legal system that is intimately familiar with all of them. But then we’d also have to deal with the postbellum mundanity that sucks the excitement out of every successful cool movement. If you don’t know what I’m talking about, you weren’t at LinuxWorld this year.

At its most fundamental level, open source is about legality: unless open source licenses can rely on our legal system to enforce the freedoms they’re trying to protect, open source doesn’t really exist. So let’s give a round of applause for the Court of Appeals for the Federal Circuit, and for JMRI for standing up for what’s right!

I have the pleasure of attending the Linux Foundation’s Collaboration Summit this week in Austin, TX. Immediately, the first session, “State of Linux Roundtable - Kernel Hackers”, was quite thought-provoking.

I use Linux every day of my life, through devices I carry around and a variety of services I use online. I talk daily with engineers who write software for it, and companies who make money by selling its associated services. I even have the penguin tattooed on my arm. My life is more intertwined with Linux and FOSS than most, but I’m not a kernel hacker…no matter how loosely you define it. I haven’t built a kernel since 2002.

Do I take the Linux kernel itself for granted? Yes. I don’t talk about it with colleagues, I don’t evangelize it amongst my peers, and I don’t think of ways I can contribute to its refinement. If the past two years of trade shows and conferences is any judge, though, I’m not alone.

Is this good? In some ways, it’s a sign of the increasing maturity of Linux distributions; after all, they exist so that people can use Linux without having to worry about its underlying complexity. I’m no longer concerned about the kernel, and that means that I’m happy with what the distributions provide. The success of the Linux kernel has made me think about it less. That’s good, right?

At the same time, though, it means I’m not contributing bug reports. I’m not trying new kernels on unusual hardware and providing feedback. I’m not helping to lobby hardware vendors to open their specs and stick to them. Without a community of users helping them in these three ways, they can’t build the kernel they want to build. In short: I’m not a part of the solution, I’m just using it.

What FOSS software are you taking for granted? Is it unavoidable that the more successful something becomes, the less willing people are to help?

(Incidentally, you can hear me talking a bit about this today at the summit here.)

O’Reilly’s Emerging Technology Conference is a four day meeting of geeks and inventors who share an interest in the technology we’ll take for granted five years from now. As I was traveling to sunny San Diego for the conference, I prepared myself to have an open mind about what our peers think the future holds.

I wasn’t prepared, however, to be hit with a stark revelation of the present from Mike Walsh during his talk about the evolution of media in Asia. Mike reported that half of the top-selling works of fiction in the first six months of 2007 in Japan were composed on mobile phones. Furthermore, there are over 72 million blogs in China, and 36% of them are active.

Similarly, during Dan Morrill’s presentation on Android, he mentioned that the mobile phone’s penetration in European markets is greater than 100%, meaning there’s more than one mobile phone for every person on the continent. So while you might consider mobile internet and blogging to be emerging technologies (compared to, say, the copy machine), you would be wrong. They’re evidently about as mainstream as the television, and maybe even more so.

What was “emerging” a mere five years ago has already become yesterday’s news. Talking about the rate of acceleration itself has even become somewhat mundane. So what’s next, according to the speakers at ETech? So far, here’s what has struck me:

• Bug Labs presented about BUG, their way to enable the long tail of gadgets by providing a variety of hardware modules that can be snapped together to create almost any device imaginable.
• Quinn Norton presented her thoughts about our bodies, the freedom to use technology to learn more about them, the freedom to modify and enhance them, and the dangerous social effects of this technology’s inevitable misuse.
• Kyle Machulis talked about something fascinating, hackable, and utterly utopian that you shouldn’t read about if you’re at work right now.
• Peter Norvig utterly amazed me with his talk about how relatively simple algorithms applied to massive amounts of data can yield interesting results.
• Michael Shiloh from OpenMoko gave us all a rundown of his company’s mobile Linux distribution and Neo handset…which hopes to prove that badass hardware + open development = the best mobile device ever.
• In his evening keynote, Lawrence Lessig strongly protested the extent of corruption in government, and suggested that the solution just might require the next great American heroes: geeks like us.

Naturally, we may find all of these things (and more) to be commonplace by the time ETech happens again next year. If you find yourself overwhelmed, Gina Trapani from Lifehacker spent part of her talk convincing us that leaving our laptops, sitting on a beach, and clearing our heads can make us more efficient in a world of constant technological entropy.

So what’s your take on what our future holds, and how does open source play a role? Oh, and if you think your favorite open source project’s got what it takes to cause the next worldwide social shift, make sure to nominate it for the “Most Likely to Change the World” category in this year’s Community Choice Awards when voting opens in late April!

Last week at FOSDEM, I attended Tux With Shades, Linux In Hollywood, a keynote by Robin Rowe and Gabrielle Pantera. Robin is the admin of the CinePaint project, which is used for frame-by-frame retouching of film - part of a long chain of technology needed for making movie magic, all of which typically runs on Linux.

This isn’t news to me, because I live mere miles from where a lot of this stuff goes on and my partner is a color and lighting technical director at Sony Pictures Imageworks. I get daily exposure to how movies get made. Still, I doubt it’s common knowledge that large movie studios like Digital Domain, ILM, Sony Pictures Imageworks, Rhythm and Hues, and Dreamworks use Linux extensively on their desktops and in their renderfarms. The way I figure, that makes Hollywood one of open source’s largest commercial adopters. However, CinePaint seems to be the exception to the rule - mostly, in Hollywood’s production pipelines, the applications themselves are proprietary (and, in many cases, internally developed.)

I can imagine there are several reasons for this: brutal competition and secrecy are commonplace in the industry, timelines are inflexible, and buckets of cash are usually at stake. Those things might make a company think twice about being transparent with their technology. This has created a bit of a challenge for Robin, though. CinePaint needs help to reach its full potential…and, quite simply put, Hollywood has legitimate barriers to contributing directly.

It’s probably not too much of a stretch to assume that Hollywood is not alone in this. Accepting Linux as a platform without embracing the core values and transparent processes of open source is becoming the new norm. Is this a sign of success for Linux, a setback for open source, or both?