“The Network is the Computer”

Some of us Sun alumni have been chatting about CloudFlare adopting one of Sun’s earliest and most notable tag lines: “The Network is the Computer”. There’s a lot packed into this epigram, not the least of which is a ton of history.

Poster of a Sun-2 workstation monitor, keyboard, and mouse, with papers flying out of the sky shown in the monitor, directly towrd the viewr.  Caption: "Unleash the Power of a Sun"
This poster dates from just before “The Network is the Computer”, but still… pretty cool, right?

When Sun originated that tag line around 1985—the company just approaching its 3rd birthday—it was actually quite audacious. It was a stake in the ground: Computers should be networked, or they’re… not computers. Well, at least, you’re missing out on their potential by a country mile. They’re “islands of automation”, and you can do better than that. Join us!

Sun put a network interface in every computer they built, from day one. That was not even remotely the norm at the time. But the part people tend to overlook is that Sun didn’t just say “networks are good”: they wanted it to be open networking.

At that time, if you wanted to network your computers, you paid extra for proprietary, non-interoperable networks. SNA for your IBM mainframes, DECnet for your DEC minis, Novell Netware for your PCs (because Microsoft was just a tad slow on the uptake, here, so another company had to fill the void).

But Sun said, “Nah. Let’s all use Ethernet and TCP/IP. Those are open standards.” (And kept pushing the envelope throughout our history. We were a major contributor to taking Ethernet from 10 megabits to multi-gigabits.)

More importantly, we pioneered software standards that made that hardware actually useful. We developed open network directories (YP) and open network filesystems (NFS), built on top of a generic open distributed interoperability standard (RPC/XDR). The Sun-organized “Connectathons” of the ’80s and ’90s were legendary meetups, where vendors from across the industry would test the interoperability of their implementations of these open standards.

(This isn’t to say that other companies didn’t have some good work out there. DEC had arguably one of the best distributed computing environments ever. But they were also the ones that had T-shirts in the ’80s that said “The network is the network; the computer is the computer. Sorry for any confusion.” Oops.)

So all props to Cloudflare for recognizing a great tag line when they see one, but… “the CDN is the computer” is not quite as world-changing as what Sun did.

Update: This post was based on a response I made to someone on Facebook, who was looking for comments about Cloudflare trademarking this tagline, for an article he was writing.  Here’s the article!

Flashback: A reasonable discourse from the top of a moving train

I just ran across a conversation I had with InformationWeek in 2009, with the catchy title “OpenSolaris: No Standing Still On A Moving Train.”  This was not long before the Sun/Oracle acquisition, and yeah, the train sure felt like it was moving at breakneck speed.

You look at this train, and you know it's not going to be happy until it's going at least 150 mph.

(Credit: Wikimedia Commons)

I’m happy to see the interview caught the gist of Sun’s Solaris message of the time: what was really new wasn’t just about new features (although there were plenty of those), but in how we were delivering change to customers. We had gone all-in on a development/delivery cycle that’s very familiar today, but had almost zero mindshare in “traditional” IT of the time.  I think those of us who were at Sun should be very proud of how we executed on our vision, and I’m especially proud of the path we’d charted to do it without blowing up our customers’ collective vision.

The interviewer considered it to be an “unexpected” question that came to his mind as the conversation developed: How do you get people who use software measured in lifetimes of years and decades to move to software lifetimes of mere months? (“Mere months” was considered to be lightning speed in those ancient days of… a few years ago?) But that’s the question we’d all been wrestling with, and I believe the record shows our engineers came up with some solid ways to make that move predictable and logical.  My job, as a marketer, was to make the case for that—not always the easiest thing to do, given the amount of legacy angst that was out there.

Here’s a clue as to the challenges we faced and conquered: the interviewer concluded that Sun didn’t need to make our software open source to execute on this vision.  Why was that a big deal?  Because too many people heard “OpenSolaris” and thought that “open source” was the beginning and end of our value proposition.

The interviewer was right: “open source” by itself does not guarantee improvement.  Too often, it means a disinvestment, which was the opposite of what we were doing at Sun.  Open sourcing our operating system was very big news, but it still wasn’t the biggest news: Solaris innovation was now moving to our customers at warp speed, without sacrificing its legendary stability.

And of course, our messaging never underplayed the critical role of open source in our strategy.  It’s what fueled the growth of the Solaris ecosystem in the early 2000s, and the abandonment of open source was arguably what ultimately put the brakes on the moving train.

In the end, building a software strategy, and communicating that strategy, involves lots of moving parts, and it’s gratifying that this particular interview captured a critical moment in time in a concise snapshot.

Foreword(s) into the past!

Over the last few years, I’ve had the honor to be invited to write forewords for two books: Oracle Solaris 11 System Administration: The Complete Reference (2012), and Oracle Solaris 11 System Administration Handbook (2015).

Oracle Solaris System AdminstrationOracle Solaris 11 System Administration Handbook

(This is part of my secret plan to write all the forewords to all books ever, by first staking out Solaris administration books as my homeland turf. Once I’ve captured that hill, the rest is inevitable.)

It turns out that both forewords are visible on Amazon, in case you’d like to see what I wrote (see the links above).  Short summary:

  • The operating system matters
  • Unix is an amazing platform
  • Unix and its offspring are everywhere
  • Unix (and its offspring) turns out to be a fantastic springboard for agile microservices.  (Aka: we’ve been doing cloud since before cloud.)

Doesn’t seem too controversial too me; but, it’s a story I love to tell.  I’d like to expand on that topic in a future post.

Oracle blog posts

TEMP

It’s fun to see some of my old Oracle posts turn up in search results from time to time.  Here’s a representative sample of a few posts I particularly enjoyed writing.

My favorites were the ones where I could weave in a bit of outside arcana.  I’m grateful to have had an opportunity to cover such pressing issues as circus tent rental and meat printing.

But sometimes it was a little more about getting down to business, whether that business was related to operating systems, developers, or high availability.