Linux Today: Linux News On Internet Time.
Search Linux Today
Linux News Sections:  Developer -  High Performance -  Infrastructure -  IT Management -  Security -  Storage -
Linux Today Navigation
LT Home
Contribute
Contribute
Link to Us
Linux Jobs


More on LinuxToday


Looking for Mr. ISV

May 09, 2008, 22:30 (12 Talkback[s])
(Other stories by Brian Proffitt)

By Brian Proffitt
Managing Editor

One thing's for sure; with all of this talk about "open source billionaires" out in the ether-web, clearly people have some high expectations of open source's profitability. Or low, depending on the point of view. Honestly, unlike every other technology pundit this week, I'm not looking too hard for the open source billionaires. If there is indeed a dearth of people making beaucoup bucks from free and open source software, I think it's only a symptom of a larger issue.

The issue is a persistent one: the difficulty in attracting corporate developers to Linux.

Some would argue with me that this, like the billionaire hunt, is a non-issue; that there are plenty of ISVs interested in Linux (with more every day) and I am putting too much emphasis on a problem that's being solved. But here's the thing: in the last two weeks I have talked to four high-level execs from three different organizations actively involved in Linux, all mulling over the same problem: how do they get independent software vendors to their distros in particular and to Linux in general?

The answer to this question lies in what specifically is holding these ISVs back? Lack of interest? Trepidation about the whole open source universe? No money to invest in R&D? Confusion about licenses? An inability to see a fiscal return? Enslavement to the evil proprietary companies?

In a word, yes. These are logical (to them) reasons for ISVs to avoid the Linux platform. But these point to two broader problems that all Linux-reluctant companies share, no matter their stated reasons for not developing for Linux at this time.

The first problem is one we already know: a lack of information. They don't know enough about Linux, and they are basing their decisions on half-formed bits of data that they gleaned from various sources. It's been a challenge for Linux (and open source software) to face from Day 1. But, through various outlets, I really believe that this challenge is finally being overcome. A lot of people know what Linux is. And the GPL. And all the other things that used to look like boogymen in the dark, until the light was turned on and only a cute penguin is sitting there.

But the other problem is not one that is discussed too often. But it stuck me when I was watching, of all things, the movie American Gangster this week.

If you are not familiar with the film, this 2007 biopic follows the story of Frank Lucas, the drug lord who rose to power in Harlem in the 1960s and '70s. Lucas did what no other black man had done before; broke the drug monopoly of the Italian Mafia and corrupt NYPD to create is own empire that soon eclipsed that of his competitors.

The way Lucas did this was staggeringly simple: instead of trying to carve out a niche in the existing drug production and supply chain, he traveled to southeastern Asia and made a deal with a producer of heroin directly. By completely eliminating the middle man, Lucas was able to transport the pure product to the US (via military transports coming back from the Vietnam War), prep it for distribution, and sell a better product for less cost than the existing suppliers.

Now, just so we're all clear, this is not a comparison between software production and the evil, poisonous drug trade. Just say no, kids.

Luca's solution is instead a very dramatic example of lateral thinking. Faced with a seemingly insurmountable problem (the deadly hold on the practice by the Mafia and their corrupt police accomplices) Lucas sidestepped the problem entirely and came up with a brand-new solution.

Granted, "cut out the middleman" is not unique to any business plan, but Lucas' application of the principle, taking almost perfect advantage of the political and cultural situation of the era was certainly pretty smart. So smart that he was able to elude (honest) police investigation for quite some time because no one in the 1960s was able to initially believe that a black man would hold more sway than the Italian organized criminal organizations.

It seems to me that such lateral thinking, applied to the far more worthy and legal practice of software development for Linux, is something that is really needed. There is a wall, real or perceived, between Linux and ISVs. Someone on either side needs to come up with an idea that sidesteps all of the obstacles that are currently standing in the way.

One possible direction for a solution is this notion of a middleman. In the beginning, our bespectacled geek ancestors built the computers and the code that ran on them at the same time. But somewhere along the line, there was the bright idea of separating development between what talked to the computers all the time (the operating system) and the separate, temporary applications that were needed.

In effect, the operating system became the middleman. It was a good system, since application developers no long had to code to individual machines. If the app could run on a given OS, then it didn't matter what the hardware was. But then one operating system became too powerful, changing its mission from helpful platform to strict gatekeeper: if you didn't play ball by their rules, you had to take your ball and go home. (If they let you keep your ball.)

Today, it seems, ISVs are still approaching Linux as something like that bullying middleman. It's not Linux' fault--it's just that after so many years of doing things one way, the notion that there's another way to interact with an OS is completely alien. This, I believe, is where the need for lateral thinking comes in.

Linux as a community needs to dismantle the notion that ISVs have to deal with Linux just like Windows, OS X, or Unix. Yes, we can espouse till we're blue in the face that Linux is free and open. But what does that mean for ISVs on a fundamental, practical level?

If we can find that solution, I think the ISVs will come to Linux so fast, we'll wonder why they ever held back in the first place.

Related Stories: