The Half-Way Covenant: Companies aim for ‘partly’ Open Source

By Jack Bryar
NewsForge Columnist
Open Source business
This one might take a minute to explain. It’s the connection between the most important and least known event in
early American history and the reason that you should be unhappy that every business plan I’ve seen recently has the
words “Open Source” in it. If the connection holds up the Open Source movement could be in for a real challenge over the
next couple of years.

Here’s the issue:

Some readers know I do a little moonlighting with colleagues who specialize in doing market research and business
development planning for would-be start-ups. Recently I’ve been seeing a whole pile of proposals. Nearly all of them
employ the phrase “Open Source” several times in their executive summary. There’s only one problem. Most of the
time, those two words are preceded by the word “partly.” Can a program really be “partly” open? Is that enough
for most IT professionals? And if it is, does that become a real big problem for Open Source movement evangelists, not to
mention firms that have already made a commitment to the Open Source philosophy?

I’m either reading, or am supposed to read, about a half dozen business plans over the next week or so. For obvious
reasons I really cannot get into the details of these things. But I can tell you this: As I’ve been plowing through them I’m
struck by the variety of ways that new business ventures are trying to come to terms with the Open Source movement. I’ve
also got a lesson concerning just how difficult going “open” is for many ventures, for reasons both practical and
philosophical. I’ve begun to appreciate why corporate and independent systems integrators are going to find “partly
open” software difficult to resist.

I also know my history. It suggests that widespread acceptance of “partly open”
programs is both inevitable, and likely to drain much of the energy from the movement.

Here’s the history lesson:

Unless they’ve severely dumbed-down the curriculum since I left elementary school, every American kid knows a
little something about the Pilgrims and Puritans. They probably know those groups founded religious colonies in what became the
New England states. Few of those kids learn why those little theocracies became among the least overtly religious
places in what is now the United States. It’s due to something called The Half-Way Covenant.

The Puritans were among the leading intellectuals of their time. Many were amateur scientists. They were also political
and cultural revolutionaries. They got their nickname because they agitated for a purified, scientific kind of Calvinism back
in England. They hoped that they could use the colonies to develop into model “Communities of Saints,” places of and run
by the godly — all the better to help overthrow the established order back home.

The amazing thing was, they succeeded.

They had a huge, if temporary impact back home. And, after a rough start, they quickly developed into thriving
communities, which attracted boatloads of immigrants. Trouble was that many of the new arrivals weren’t exactly saints, a
development the founders never anticipated. The leaders of the colonies realized they had to choose between ideological
purity or growing their communities even while keeping all the new arrivals inside their churches.

Under a minister named Increase Mather, they came up with a compromise. They decided to let the less than saintly, the
partial believers, into their churches in the hope that these reprobates might eventually that they might eventually
“experience Full Grace.”

It didn’t work out that way. The long-term result was the dissolution of Puritan society. Congregational churches
grew for a while but they lost much of their religious
fervour. Their institutions have limped along for generations, but the New England landscape is now dotted with old,
white, Puritan-built, churches. Many of them are empty on weekends. And the Puritan passion has disappeared altogether.

Here’s why this matters:

The Open Source movement also started out in much the same way, as a collection of highly intelligent zealots.
The movement was hardly of the same scale as that one three and a half centuries ago. But the comparison is
a valid one, I think. And this new movement has likewise met with surprising success, perhaps too much of it. People want
to join the Church of Open Source without necessarily knowing what that entails.

Is this merely a movement to get something for nothing, for good stuff cheap, a notion Slashdot founder Rob Malda once scorned as free
code, like “free beer?” Or is it an attempt to collectively understand, and cooperatively use and improve systems for the
common benefit? And if not that, what type of Open Source will people settle for? And to what purpose?

Commercial developers and many current and would-be businesses want to know the answer.

Within the commercial vendor community, companies like Compaq, Sun, Imprise, and Apple have experimented with
“partial Open Source” for some time. In fact, “partly Open Source” products have been available in various forms for
years.

The reasons vendors give for not going fully open vary widely. Some newer product developers told me that their
product was “shippable,” but that its code wasn’t really in a form that users would find useful. I translated that to mean that
their first rev was so was so full of spaghetti code the developer would be embarrassed if anyone saw it.

In other cases,
vendors said their code is derived from a mix of proprietary code (which they can release) and third-party, licensed code,
which they could not. That is a valid problem. Often even the third-party code contains libraries or other elements, which
were licensed from yet other vendors. Tracking down and obtaining all the needed permissions becomes a non-starter.

In other cases, vendors really do want to hold back elements for competitive and commercial reasons. In many of these
cases, the “open elements” look to be little more than a glorified programmer’s interface, just code the developer hadn’t fully
baked. As one proposal put it, they want customers to “to have access to any part they may need to adapt,” but not all
elements, especially newly developed code or elements, which they considered “the core of our competitive
advantage.”

Whatever their motives, developers of such “partially ppen” solutions used to encounter skepticism, even hostility
from Open Source and Free Software puritans. Many activists I wrote to said they advocated boycotting any partially
closed program that relied on Open Source elements or ran on Open Source platforms. Several said they considered such
partially closed programs to be at least a spiritual violation of GNU licenses.

But that was before Open Source went mainstream. Will corporate system administrators and larger commercial
integrators be as picky? I’m still talking to various integrators I know, and many of them have said that they’re inclined to
meet vendors “half-way.” As one said, “What I really need is a clean, open API and source on the pieces of code I may
need to write to.” It’s an article of faith among Open Source Puritans that Open Source allows them to fix problems rather
than wait for vendor supplied assistance. But another system manager I spoke to characterized Open Source as little more
than a “security blanket.” He didn’t think having source was always that helpful.

The fact is, “fully Open Source” doesn’t
always mean terribly useful. I’ve seen open code where documentation is uneven or of such poor quality that it was of little
practical use, but the program ran well.

So will new, partially converted administrators and integrators welcome “partly open” programs? And if they do, has
something important been lost as a result?

NewsForge editors read and respond to comments posted on our discussion page