Open Source

Open source is not to blame for a lack of industry standards

One commentator blames the virtual data center mess on open source. Here's why that's wrong.

datacentervirtualizationistock92734913baks.jpg
Image: iStock/Baks

Carol Wilson wrings her hands over the "boring" nature of open source standardization, declaring that "Open source processes can take the fun out of everything, particularly technology wars." Putting aside for a minute the irony of expecting standards to ever be anything more than mind-numbingly dull, Wilson's larger argument misses the point.

The problem with open source standards aren't that they're boring; it's that they're largely the same as the proprietary standards that preceded them. In practice, this presents no problem at all.

The standards wasteland

I'm not really referring to "standards" in the ODF or MANO sense, i.e., hordes of HP and IBM employees sitting in dark convention rooms, haggling over dotted I's and crossed T's. Big enterprises fight over such standards, but eventually the market crowns a de facto standard without any apparent attention paid to the standards bodies.

SEE: International technical standards needed to accelerate IoT growth (TechRepublic)

Take ODF, for example, the industry's open alternative to Microsoft's Office Open XML (OOXML) standard. IBM and others raged for years against the Microsoft machine, arguing that OOXML wasn't a true standard (despite its blessing by the International Organization for Standardization), was too complex, and didn't go through the proper vetting process. Despite howls of protest (including my own), nobody cared. Nearly a decade later we're still happily using Microsoft with proprietary file formats, or proprietary cloud services (Google Docs), or none of the above.

"None of the above" is often the market's reaction to proprietary standards. Microsoft milking a desktop operating system monopoly? We'll move to the web and to mobile devices. Data tied up in proprietary databases? We'll introduce a new generation of NoSQL databases. And so on.

Open source "standards"

It's much the same in open source, except perhaps such shakeouts happen a lot faster. Wilson disparages the "Hey, we're open source, we all get along" idea as "nonsense," and she's right, though perhaps for the opposite reason than she imagines. Anyone that has spent time in the open source world knows that open source projects are as prone to fight against competing projects as are commercial concerns (in part because companies increasingly compete through the open source projects they espouse).

SEE: Lack of standards threatens to derail the big data innovation train (TechRepublic)

There is no magical kumbaya in open source.

Nor is it reasonable to assume there should be. Pointing to an argument made by Margaret Chiosi, former AT&T executive, Wilson writes, "If the telecom industry can't come to an agreement on a MANO approach sooner rather than later, its ability to compete against web-scale giants is at risk." This is a false dilemma, as the reality is that the web-scale giants are putting the telecom industry to the sword not because they can't powwow on a common fix, but precisely because the market is moving on to displace them.

It's not an open source problem, or a standards problem.

In open source, as in proprietary software, there are scads of developers that build yet another project to tackle this or that industry problem, proliferating "me-too" projects and frustrating standards. On that note, Wilson bemoans four different MANO projects trying to tackle management and orchestration for the virtual data center, but conveniently forgets this is always how things begin, whether in the proprietary or open source worlds.

How standards actually happen

In the proprietary software world, the proliferation is fed by venture capitalists; in open source, it's fueled by code. Either way, the market eventually determines a winner, and that winner is never picked by a standards committee.

For example, Linux, despite largely settling into Red Hat's comfortable dominance of the enterprise and Google's ownership of mobile, used to be a mishmash of warring factions (Turbolinux, Caldera, Conectiva, MandrakeSoft, Red Flag Linux, and SUSE, among others). There was an effort to ward off Red Hat's market leadership with United Linux, a "standard" of sorts, but the market crowned Red Hat Enterprise Linux as the standard instead.

Open source, in other words, acts largely like the proprietary software it displaces. Competing projects do just that: compete. Eventually a "standard" emerges, with that standard designated by market competition, not by committee.

Also see

About Matt Asay

Matt Asay is a veteran technology columnist who has written for CNET, ReadWrite, and other tech media. Asay has also held a variety of executive roles with leading mobile and big data software companies.

Editor's Picks

Free Newsletters, In your Inbox