Monday, July 26, 2010

1000 Days of Uptime!

One of my main servers running a Debian distribution of Linux is having it's 1000th day of uptime today:

Thanks to for the nice stable data center. Wisconsin - no earthquakes, no hurricanes.

Another cool thing - one of the servers running on here, written in Java has been up and running since 2008 - not a lot of people can say they've written software that has run over 500 days without having to be restarted. The server has handled over ten million requests since being started (yes I know lots of sites handle that in a day) The java server is based on the core of FIT Issue Tracking Software - It's running on Java 1.5.0_06 VM.

Here's looking forward to another 1000 days, or jinxing it and having the server crash tomorrow…

Sunday, August 02, 2009

Top Nine Lessons Learned From Entrepreneur Dave Thomas

Top Lessons Learned from Dave Thomas talk at Lead to Win

Dave Thomas is an entrepreneur and former professor who started Object Technology International. To understand what kind of person Dave is, Dave and his wife owned 50% of the company each. When the company was bought by IBM [I couldn’t find financial details, but my understanding from friends of friends that this was on the order of tens of millions to low hundreds of millions], Dave gave his share to the OTI employees. Dave brings an incredible amount of integrity to everything he does, and the talk he gave at LTW was the best I heard (I took five pages of notes compared to one page of notes for the next best topic). This is my list of the best nine ideas that Dave presented.

1. If, as an entrepreneur, you’re not selling: yourself, your brand, your company, your products, and your team, then you’re not an entrepreneur, you’re just playing company.

2. Three key principles in his companies:
a. Principle#1: Sell
b. Principle#2: Build
c. Principle#3: Get along

3. Plan for an eight year exit – the half life of a software company is 5 years – not likely it will go ten… so plan on eight. Changes your perspective when planning.

4. If you want be unique, if someone gives you five minutes to make a presentation, only take two.

5. People’s behaviour is consistent over time, regardless of technology and tools. A 20 year study at BNR/Nortel saw that developers that were late to deliver, were always late, regardless of tools, technology, or team.

6. When people do you favours / act as advisors keep a list of IOUs – and when you’re successful, follow up with them.

7. Buy the best (in terms of tools, etc), build the rest.

8. As a startup entrepreneur, buy plane tickets and get in front of people, customers, thought leaders, etc.

9. For Canadian startups, leverage trade offices in major cities (DFAIT) in the US and around the world.

Wednesday, June 17, 2009

What I'm up to these days...


It's been a long long time since my last post. A quick update as to what I'm up to. I have enrolled in Carleton University's Technology Innovation Management (TIM) program as of January. I'm loving it Thanks to Ian Graham at The Code Factory for pointing me towards TIM, and thanks to my mentor Ross Blaine for pushing me to return to school to get a Masters. Studying entrepreneurship, innovation, ecosystems - things I love. People who know me ask why I would do this, when (they think) I could probably teach these topics based on my experience - but I've never really had the time to really study entrepreneurship - to learn the theory. I can talk about my A companies, and B projects - but there are studies out there based on thousands and thousands of companies and projects. I've learned more relevant information in the past 4 months, then I have ever in my life.

I consulting on the wireless front, and I'm getting ready to launch the next company.

Lots going on - Hoping to post more regularly as the new company gets going.

Monday, October 01, 2007

Moving on...

I have decided to move on from Alcea Technologies to pursue interests outside of the company. Going forward I will be making myself available to the Alcea & FIT team as needed.

I wish the best of luck to the Alcea team, as I look forward to working on solo projects and short term development projects as the opportunity presents itself.

Watch the space to see what I’m up to. I’ll be putting my resume up in the next couple of weeks, and if you’re looking for a rock solid software guy for contracts between one week to one month (I can travel for a couple of days / month at most), send me an email.

Please use for any contact going forward.


Justus Technologies Ltd.

Tuesday, June 26, 2007

Making your web app work on a BlackBerry (or other mobile device)

I had the good fortune to get invited to RIM's WES conference to speak about making our FIT issue tracking software work with a BlackBerry. I have taken the presentation I gave, and gone into a good deal more detail, and I am putting it up in an 8 part series on my product blog The first part went up yesterday, and there will be a new part posted every day.

Monday, May 14, 2007

Review: RIM's Wireless Enterprise Symposium 2007 and how it ties into our tool FIT

I was fortunate enough to get invited by RIM to speak at their annual conference, called the Wireless Enterprise Symposium (WES2007). I was discussing our experience as we begin to implement a mobile interface for our enterprise issue tracking software, FIT. It’s been a terrific four days, with RIM keeping everyone generally educated, fed and entertained. I’ve met and talked with dozens of people at various companies around the world – it has been quite eye opening to see how much mobile technologies are being used and deployed at so many different organizations. The most valuable part of the conference in my mind has been the networking sessions as well as the customer / user portions of the sessions that I attended (real users, real world stories – great stuff).

There were 2 tremendous keynote speakers at the conference as well. Both of these speakers spoke about topics that dealt with RIM, management, judgment, and innovation. Malcolm Gladwell. He’s a tremendous story teller, and actually went to University with Jim Balsillie (co-CEO of RIM), and Malcolm claims that he “knew Jim when he was poor”. The general gist of his talk dealt with the fact that an expert’s intuition (the split second judgment and decisions) is better than spending time analyzing a problem – because this can lead to information overload, and second guessing. One particular example he gave was how doctors assess whether a patient is having a heart attack when they have chest pain. He outlined how in the past – doctors accuracy rates had actually been quite low – but that they substantially improved when information was removed – the diagnosis was more accurate if they knew just their blood pressure, ECG (and two other factors). Things were much more accurate if the doctors ignored the patient’s age, family history of heart problems, smoker, etc. They were actually facing information overload, and unable to process all of the factors. This ties into what we are trying to do for leaders and managers running their organizations with FIT. Email and IM can be tools that allow leaders to micro-manage, collect too much information, and over-analyze. Our issue management tool FIT allows leaders to see the most relevant information, the most pressing issues, and filter out some of the noise of the day to day operations of the business. Malcolm’s talk was extremely relevant to our product.

The second keynote speaker was Clayton Christensen, a professor at Harvard. Professor Christensen described some of the analysis he had done to look at why successful companies fail. Why all of the minicomputer manufacturers more or less went out of business in the 80s and 90s (IBM succeeded by creating separate business units with mandates that allowed them to kill their core businesses: Mainframe to minicomputer, and minicomputer to PC), why North American car manufacturers are struggling, and how the iPod has succeeded where the iPhone will likely fail. Again, a great speaker, and once more, at the core of his talk, was information as it relates to FIT. In his talk (and his book, Seeing What’s Next, that RIM gave to all WES attendees), Christensen describes how there are disruptive opportunities when customers:
1) Are paying too much.
2) Require too much skill to use a product.
3) Have limited access to a product.
4) That is too time consuming to use a product.

All 4 of these factors are issues that FIT addresses when compared to existing “enterprise” systems. Systems that can take weeks to install, require consultants to configure, and that certainly are impossible to try before you buy – simply because of the complexity the products themselves. FIT deals with all of these needs including publishing pricing information on our site (no need to talk to a sales person first), and allowing everyone to try before they buy. Our goal is to have fully functioning systems up in minutes, and completely configured systems matching our customers requirements and needs (rather than requiring customers to change business processes to meet our software’s needs). We are aiming to bring information management to the masses – A perfect paradigm comparison to the shift from mainframes to PCs.

Watch this space over the next two weeks, as we begin to put detailed information (more detailed than in our presentation at WES), documenting how we have taken our issue tracking tool and made it mobile enabled, particularly for the BlackBerry but the same web-based interface is available to Windows Mobile and other mobile devices as well.

Friday, January 26, 2007

Mailinator Architecture... Great article...

Read a great article today about the architecture of mailinator. Very interesting reading. Paul Tyma's end product really lines up with my own core values:

  • To be ultra-efficient, where necessary, build it yourself.
  • Code for software survival.
  • Understand how your software is used to make things more efficient (ex: Mailinator compresses all inbound emails - 99% are never read, compress once, decompression is rare and cheap, and memory is finite).

We run hosted FIT servers on behalf of our customers and some of them have been running continuously with no restarts for over 200 days. This kind of robustness comes at the architecture level - creating a design that ensures memory is used efficiently, testing for memory leaks, and that a framework that attempts to continue running regardless of the conditions thrown at it. Expect the worst, and persevere through it. Great reading...