February 23, 2006

On Good/Bad OSS, OSS 'Business Models', and Its Relationship to the REST/SOAP Religous Debate [Part 2 of a Mini-Series]

Of all the successful open-source projects that come to mind that I personally use with any consistency, not a single one of them actively pursues "donations" (Apache, Mono, Mozilla, Saxon/Saxon.NET, IKVM, more recently 4Suite/Amara (still gearing up, but I see it as a farely significant piece of my future, CherryPy (ditto), ICECAST, and a few other smaller projects) and instead have properly built a business model around licensing extended feature-sets, support, consulting, and other reputable forms of OSS financing such as corporate sponsorship.

The above projects are examples of OSS done right. There's a need. The community rallies around that need. That need is now taken care of by folks who know what they're doing.

What then represents the wrong way?

I'll stop short of insults (rarely do insults bring about anything but insults in return. Not a lot gets accomplished when you make insults the core message of a particular post) and instead provide a general rule of thumb:

'Guilt Marketing' is bullshit.

OSS projects that aggressively seek after donations, using terms such as "Have you donated yet?" in pop-up boxes that require you click a yes/no-type button contained in an alert box of some sort, or even those who believe that a good and reputable business can be built around a donation-based system are flat-out wrong. It doesn't work.

Accepting donations is one thing. If you are not aggressive about it, then fine, whatever... Not a big deal. If its a side gig that fills a needed niche, and people want to throw you a few bucks, and you make it easy for them to do that? You're not going to here me complain.

But using guilt marketing?

Like I said, bullshit.

OSS can be both a good thing and a bad thing. When you introduce half-a$$ed software solutions that do little more than confuse the market with mixed messages, diluting the perceived value of the good and reputable products in this space, more damage is done to the economy than can be fixed by the money "saved" by the folks using these solutions. When you attempt to disguise a product as "free" but insist on using guilt marketing tactics to make people feel like they are not doing their part unless they donate to the cause...

= The "Send me your weak, and your old, and I will give them REST!" Gospel Hour

The same can be said about folks who undermine the good reasons for using SOAP-based solutions by hiding behind the notion "if its difficult, its evil!"

No. Its just difficult for the average Mort. Thats not a bad thing. There are programmers, good programmers, great programmers, and mortal God's. When you are performing a banking transaction, do you want the software developed to perform this transaction written by a good programmer, a great programmer, or a mortal God?

If it take's a mortal God, or at very least a great programmer to ensure that my transaction completes successfully, and does so secured from end-to-end, and these same great programmers/mortal God's suggest they need something a little heavier than REST to get their job done...

Then shutup and let them get their job done.

Unless, of course, you don't mind if your bank only gets through the deduction portion of a transaction that contains a deposit as well. Or maybe you want someone to charge your credit card, but never ship the package, or ... (fill in the blank with anything that requires and all or nothing type atomic fulfillment)

Yeah, SOAP is *THAT* important. From the standpoint of posting content to a forum, or updating an Atom feed, its not. But to mission critical "It's all or nothing" transactions, it is! No religous movement is going to change that. Instead, it clouds the issues at hand and confuses the economy.

This needs to STOP! If you don't need SOAP, then don't use SOAP. If you do, do. But by making a religous war of things, you're doing a TON of damage, much of which is irreversible.

Seriously, you need to stop. No, I'm not kidding.

Stop.