Wednesday, September 27, 2006

Startup Fun Time

If you have never worked for a startup, you might not get it. Lots of risk and lots of potential reward. Many failures for every success. Daily swings of up and down that have forced many to mood enhancing (or stabilizing) drugs on a regular basis.

But, if you are lucky, you reach something I call Startup Fun Time. This is an amazing time that gets under your skin and becomes part of your psyche if you're lucky. It's amazing how many folks who have experienced Startup Fun Time never go back to a large company. A mentor of mine once told me that startups are like rollercoasters while large companies are like smooth sailing in a duck pond. Both can be rewarding and successful experiences, depending on your personal sense of adventure. Strangely, I've been involved in four startups and I really don't like real rollercoasters as they scare the broccoli out of me.

I have a vivid memory my first day at a real startup. I was told to go open the office doors for the first time with our only other employee who was going to be our office manager. We met at the office space, opened the door and realized that there was absolutely nothing inside. No cubes, no wastebaskets, no phones, nothing. Not a single pen or scrap of paper. The first thing I recall saying was, "Well, I guess we need to go to Office Depot...." Of course, we had nothing to make a shopping list on :)

Startup Fun Times had not begun although the logistics of building a company from scratch did has some intellectual stimulation. Being a one or two person company where the hours are long and the stress of being squashed nearly everyday is not Startup Fun Time. But, slowly, over time, with some success, we reached the world of Startup Fun Time. Then the company got big, went public, got bought, and got sold. Somewhere in that last sentence Startup Fun Time halted and the monotony of the duck pond took hold. Luckily, I left the company before the last few corporate moves and immediately went to join another startup looking for that elusive Startup Fun Time. I was hooked and the drug was not out of my system.

So, what are the components of Startup Fun Time? I think they are more or less this:

  • Intuitive products. You get it and your mom gets it. Customers buy without huge sales and marketing efforts. Folks really do get what you are doing in an elevator pitch and want to buy.
  • Small group of motivated people. Generally, the company has less than 200 people. You know everyone and what they do. You know the names of their wives and kids. No-one ever asks, "What do you do here?" I recall the day that someone asked me that question as the day I considered the beginning of the end of Startup Fun Time.
  • Everyone contributes. Engineers take sales calls, tech writers test code, sales people help spec features and functions. Everyone sees everyone helping out and silos are virtually non-existent. Wheat is quickly separated from the chaff, so to speak.
  • Spontaneous fun. People at the company have fun at work. Fun is not contrived with "corporate events or retreats" that feel more like work than a retreat. I've seen food fights, wiffle ball bats and radio controlled blimps in the office. No-one had to tell folks to go have fun, they just did on their own.

Yesterday, a fellow Vyattan told me that "this is the fun time" and he's right. Vyatta definitely appears to be entering Startup Fun Time and accelerating toward a successful future. As we grow, it'll be great to watch the fun expand. We'll just have to be careful that as we approach the duck pond we remember how we got there.

Thursday, September 21, 2006

The Open Prosthetics Project

I read a story today in Wired News about Iraq war veteran Jonathan Kuniholm using the open source development model for human prosthetics. The article talks about the Open Prosthetics Project that looks to be an extraordinary undertaking to help out folks that need artificial limbs. This is an amazing project and I sincerely wish them the best of success.

Reading this article I was again reminded how to the importance of the community to an open source project and business. The community can expand the market and reach of an open source project in a unique and rapid manner. So, while our business is clearly different than the Open Prosthetics Project, we share the same overarching open source beliefs, values and community motivation. Closed-source prosthetics companies, you are formally warned :)

Tuesday, September 19, 2006

The Money

In case you missed it last week, we announced our $7.5M series A funding. This is a great milestone in our company development and gives us plenty of cushion so we can continue to build the business!

The challenge for us now is to continue to build the community and work on expanding open source networking. If you've got a thought on how you want to see us expand and grow, please drop me a note.

Sunday, September 10, 2006

CoolTech Club meeting

If you're around Menlo Park, CA next Monday night (September 18th) I'll be speaking about Vyatta at the CoolTech Club at 7pm. Hope to see you there!


A bit of a post-mortem on the CoolTech Club meeting... I had a lot of fun presenting to this technical audience. There were definitely supporters and skeptics of our business in attendance - and both camps were not shy about their opinions :) I firmly believe in healthy debate as you have to challenge all business and technical assumptions to be successful. Thanks CoolTech Club - it was fun!

Friday, September 08, 2006

Open source beer

Lawrence Lessig writes in Wired about some Danes that are using the open source model for their beer - remarkably called Free Beer. Basically, they publish the recipe (the source code, so to speak) and allow people to brew their own. If you make and distribute changes to the code, err, the recipe, the license requires you to publish the changes to

So, you can make you own beer and it will taste just the real Free Beer. That is if you compile the code, err, make the recipe correctly. Truthfully, I'd probably rather pay someone to make the beer for me and serve it to me in frothy mug along side some wings and nachos. But, each to their own.

And there is the main point - open source licenses give the consumer a choice. You can choose to make your own router/firewall or your own Free Beer. The source is there for the taking. Or you can chose to have someone produce a router/firewall image and brew the beer for you for a reasonable fee. I'd contend that when you consume a sufficient number of router/firewall devices on your network or enough pints you're probably willing to pay for some support services as well - especially if you consume both at the same time :)

I know how the OFR performs and operates but I have no idea how Free Beer tastes. Anyone brew a pint on their own yet? Got some nachos handy?

Wednesday, September 06, 2006

Data and telephony worlds colliding!

Yes, Seinfeld fans, that titles is a blatant rip-off of George's worlds colliding in a classic episode :) I remember when there were separate data and telephony networks in most enterprises. Completely separate people, departments, management and technology. When someone was added, moved or changed you had to go talk to both the data networking folks and the telephony folks - they were rarely the same person or even in the same department. A phone system, a PBX, was proprietary software running on proprietary hardware (ATT, Rolm, Nortel, etc.) and data networks were running protocols that rarely see the light of day today outside of a historical textbook (Xerox XNS, Banyan Vines, Appletalk, etc.).

Today, I find it hard to imagine an enterprise IT department that does not handle both the data and telephony needs of an organization. And, in many cases, it may even be the same IT personnel handling both the data and telephony networks, especially with the acceleration of VOIP throughout the enterprise. So, the wall came down between the data network and telephony folks. Worlds colliding!

Today, companies like Sangoma and Digium are building hardware and supporting open source telephony software like Asterisk that runs on commodity x86 hardware. And the Vyatta OFR provides the data networking functionality for the enterprise on the same software. I've already written about how the two pieces of software could (should?) come together on the same hardware platform and this elicited some interesting responses. Some folks want to keep the wall up between the telephony and data worlds for reasons of not wanting a single point of failure and ease of administration. I can see that point in some situations.

The point I can't see is when folks approach me and say that the systems have to be separate because of compute concerns. With single-core, dual-core and quad-core x86 hardware available, the compute power of most modern laptops runs circles around the proprietary hardware compute power in the old PBX. Same holds true for what exists in most modern proprietary hardware routers - check inside your low-end and mid-range router and I'm willing to bet you find an x86 processor or a PowerPC chip (the same chip that used to run your Mac). If commodity hardware can run databases, desktops, webservers, middleware servers, grid computing and so much more, I find it very hard to believe that bringing together telephony and data networks is too much of a stretch. Let's see those worlds collide!