March 13, 2006

Why we won’t be talking about Open Source in the future

Being a good analyst, I attended IBM’s Open Source and Linux analyst event last week. Although it was predictably on the east coast, which means a good west coast guy like me got to spend more time in airplanes than at the event itself, I nevertheless made the trek to hear from Big Blue. While some of the issues discussed and positions taken were predictable, not in a bad sense, just that they were consistent with past communications, I started having of sense of déjà vu, with the wayback machine pointing to 1996. After asking, Mr. Peabody reassured me that it was in fact a decade later, it all started to come together for me. So just what in the world am I talking about? Read on.

There was much talk about the specialness of Open Source, not necessarily so much from IBM, but from others in attendance. While there are religious devotees who believe that the most important role of Open Source is to bankrupt Microsoft, there are many who are not on the Redmond attack squad, however, that talk about Open Source as if it remains somehow discrete, or fundamentally different than other software. Then it struck me, well, OK it struck me about 2 hours into the overbooked flight home in seat 23C when I ran out of munchies that I pilfered from the snack table at IBM event. This bifurcation of Open Source from all other software was a very familiar behavior – it is the same one that ten years ago insisted that the Internet and related technologies were discrete from IT and the datacenter. A ha! As I am fond of saying, there is little new in the world, all things old are new again.

While people have the need to segment out what is new and/or different from what they know, the reality is that from a bits and bytes perspective, open source software is no different than any other. It is code that runs on the machine and hopefully solves a problem and delivers value to the end user. The development model and the pricing model vary, as do issues related to intellectual property and ownership, but at the end of the day it is just software. In 1996 we might have asked if a given piece of software of equipment was “Internet capable”, but today, the question is never asked as networking is just an assumed attribute. Some beg the question, is this software Open Source compatible? One the surface it may seem silly, but the real question is will this software be well behaved, standards compliant, and not break my IT infrastructure? (or get me fired.) Does this mean that all software will be open source in the future? This is not very likely, but the need to somehow distinguish Open Source from all others, I believe will diminish, if not vanish altogether.

Many years ago I postulated that Java was a priceless technology -- priceless in the sense that in order for it to be successful, it had to be ubiquitous, and in order for it to be ubiquitous, it had to have a price point of nearly zero. We can argue for quite some time as to whether Sun was successful in this regard, but when we look at where Open Source is thriving, it is in delivering the minimal level of functional foundation required on which discrete can then be added. Whether this is file or print services, database, basic management etc, these are all priceless technologies as well. In 1994 the TCP/IP stack and utility business for PCs was probably about a $700 million affair. In 1995, it nearly evaporated when TCP/IP begin to be natively supported in Windows 95. Thus the stack became discreetly priceless when it was bundled into the ubiquitous OS. In many ways, Open Source software is making the same demands on the marketplace – these technologies are priceless, therefore stop trying to make money them, but instead invest those same dollars in adding value on top of the priceless technology. As a result, freely distributable, standards based, basic technology will be a given, let’s learn innovate on top of it, where the real value, and may I add, margins, will be found.

From this vantage point, the future would be time where open source technology will be prevalent, found by itself, bundled into commercial solutions by business partners, and distributed and supported by major systems vendors and ISVs. It would provide key building blocks that are vetted by the large constituency and seek to ensure its widespread availability. High margin, differentiated, value added solutions would be built on top, that delivery the high value sought by the customer, combined with the support and backing of the commercial software industry. Heresy? No way. Software communism? Negatory. A smart way to focus an industry on value add and innovation? Absolutely.

So, let’s assume that the industry as a whole realizes the value in this approach. What would happen? The number of open source communities would increase. Contributions to open source communities would increase dramatically. Commercial vendors would decide to cede certain key technologies into the open source community. Business partners and solutions providers would weave together tailored solutions comprised of commercial as well as open source components for their customer bases. Notice that the end customer is probably not all that interested in modifying the code (perhaps they would take a peek inside) but rather Open Source would become simply another part of the broader ecosystem and lose that specialness that has been artificially adhered to open source technology. Suddenly, it would just be assumed to be present, and there would be no need to talk about it anymore, just like the TCP/IP stack, the Internet, or what Sun had envisioned for Java.

Not too surprisingly, when one looks beyond the hype of the moment, much of IBM’s position on Open Source seems to be following the scenario that I laid out. It is unlikely that Big Blue will suddenly cease to have any software revenue, in fact I would argue that it will ultimately rise, as the company expands it reach and boosts the market overall through its many open source contributions. While IBM is not the only vendor that appears to have seen the logic in this course of action, it is clearly the largest. With the Armonk Argonaut on board, it is likely that much of the industry will follow, and will eventually not need to talk about Open Source anymore than it does key foundational technologies.


Ludovic Windsor said...


This is an interesting post.
Would you say bandwidth is priceless too?


Clay Ryder said...

I would not consider bandwidth per se as priceless, as the speed of communications is generally less important than the persistence of the connection. Although we would all love instant massive bandwidth, the fact remains for many functions, simply knowing that connectivity is available is sufficient. Nevertheless, there are still many applications that function in a non-persistently connected environment, for example, many common laptop applications. I would tend to argue that at present neither bandwidth nor persistence is priceless, however, in the future I do think that persistence would increasingly become priceless, but not "free".

Avi said...

Excellent perspective.

I use to deliver exactly the same vision about Open Source in my presentations.