BACKSTRIP

Open source: it’s not that hard

When I hear someone spouting nonsense about open source, I tend to let it slide. About ten years ago I’d kick against the pricks — and loudly, courtesy of a regular opinion column in a magazine — but it was a different time then. For starters, open source software was (at least in the eyes of detractors) still untested, and with billion dollar proprietary software stalwarts proudly FUDding up the issue, I felt I needed to add some balance.

But now the war’s over, open source is everywhere, and anyone who claims otherwise probably hasn’t used a computer since 1998. Every notable OS released in the past decade (except for Windows) nontrivially uses open source software, and most websites you’ll visit today are running open source code at some level, if not the whole stack. Alas, buying into the faux ‘open versus closed’ war still gets hits on specialist tech sites, so there’ll be people who still fan the flames. But it’s all a show. Open and closed source have their uses and I think most folks who actually look into both options will probably understand it.

And that’s why I spat coffee all over my keyboard when I read this. It’s a story about the fascinating case of a Goldman Sachs employee who apparently ‘leaked’ source code owned by the company — code that the employee claims was open source — but it could also be about bananas on Mars for all the sense it makes.

Coffee spray #1:

Many programmers have created and promoted the computer programming language known as “open source code”…

It’s grammatically correct, but that’s about it.

Coffee spray #2:

…to be shared on public sites at no cost, but licensing issues are murky.

No, it doesn’t have to be shared on public sites, nor does it have to be gratis, nor are the licensing issues murky if you’ve read the license. To the extent that they could actually be considered murky, they are no more so than the hundreds of pages on any proprietary end user license agreement.

Coffee spray #3:

Brent Cossrow, a lawyer who practices in the area of employee defection and trade secrets, but is not involved in the trial, said little case law exists around open source code licensing.

Let’s pull this apart. First, there is certainly case law around open source licenses. There might not be as much as you’d like, but it’s there.

Second, open source licenses are functionally the same as proprietary end-user license agreements in that a violation of the contract voids the license. What happens if an open source contract is broken? Good old copyright comes into play and, wow, look at that mountain of case law on copyright infringement.

(This is actually pretty ingenuous. Think about it: you believe someone’s using your open source code in their program, so you sue them for copyright infringement which forces the defendant to prove that they had a licence to infringe on the copyright. The end result is that open source licences get tested without the licence holder ever mentioning it.)

Coffee spray #4:

“There are hundreds of different open source licenses. They don’t all say the same thing,” said Cossrow

There are a hojillion end user licenses, contracts and distribution agreements. They don’t all say the same thing.

Coffee spray #5:

“There are some who believe on the one hand that when you drop open source code into this type of larger program, it dilutes the whole thing…”

What?

Coffee spray #6:

“…There are others who believe that it really doesn’t because it is one piece of a larger proprietary pie.”

No, really. What?

Who wrote this? This isn’t just me nitpicking or being more familiar with the topic, it doesn’t make sense. It’s like a satirical pastiche of late 1990’s anti-GPL crap, except that it’s 2010, it’s a very real and significant case, and Reuters — a proper news outfit — should know better.

What message would a typical Reuters reader take away from this?