Restating the hypothesis of Brittle Compute and Organizations…
My Amazon author page!!!!

Great email yesterday pointing out that hind sight is 20/20. In looking back you can always spot the trends of yesterday easily while the trends of today and tomorrow remain shrouded in mist. My reply is below.

Agreed that Brittle Computing of the 1990’s has faded to a degree. Although I would in fact argue that many of the issues of Brittle Computing still impact many companies. The reality is Brittle Computing produced a number of solutions designed to solve the overall problem and that resulted in the reality of design for failure. Many companies still don’t fully embrace design for failure. While the heyday of Brittle Computing was the 1990’s, its still around just fading a little.

That said my primary argument is not that Brittle Computing is still the main reality, simply that in fact it produced the newer design for failure movement. The concepts of Brittle Computing however are still in the marketplace. The reality of Brittle Organizations aren’t going to go away any time soon.

Now that I am up to my neck in alligators let’s go back to the original job which was drain the swamp.

There was a time when you had to be very careful about how solutions were implemented. I call this time Brittle Computing. This style of computing drove the creation and expansion of a number of operations solutions that help us reach our computing goals without creating the Brittle solutions of the past. My original argument was that you have people who grew up in that age who look at, operate and evaluate technology differently than those who have grown up in the design for failure age of computing. So the original intent of this series was to point out to many design for failure companies the weakness they are creating.

Those who forget the past are doomed to repeat it.

From that initial stream I branched out to the broader concept of Brittle Organizations. While the two are related in that they are both born in the 1980’s and 1990’s they are somewhat different themes. The reason I brought them together is that I believe many of the realities of Brittle Computing have in the end created the realities of Brittle Organizations.

As my letter writer pointed out to me – I haven’t done the best job of linking the two together. So here is my linkage going forward.

  1. Brittle Computing created a culture of be careful don’t take big risks unless you know your backup worked.
  2. Out of that further evolved the great operational frameworks. (*they existed before the 80’s/90’s they just evolved quickly during that time period).
  3. As time passed the market matured. In the book “Why the mighty fail” by Jim Collins his team examined why many companies struggled with the rapid transitions that led to the Tech Bubble of 1999. The rapid absorption of technology in the late 1990’s and early 2000’s gave birth to the transfer. While the technology grew less and less Brittle, the Organization grew more and more Brittle. I call this the transition. Where what once was driving slow change (Brittle Compute) is now gone. What happens next is the absorption of the change culture into the very organizational dynamics of the company producing Brittle Organizations.

The difference between innovation in a large established company and innovation in small agile company is often quite large. In the large company you have layers of management you have to convince regarding your idea or concept. In the small company you lack the resources of a large company but you don’t have people saying that’s a bad idea.

History teaches us that in the end there are bad ideas. It also teaches us that cultures, left unchecked can become toxic over time. The reality of design for failure applies to organizations as well.

I fall back to one of my old favorites here. The base of a statue buried in the sand. A group of travelers come across that base and in clearing sand away from it are astonished to find text written on the base.


I met a traveller from an antique land
Who said: "Two vast and trunkless legs of stone
Stand in the desert. Near them on the sand,
Half sunk, a shattered visage lies, whose frown
And wrinkled lip and sneer of cold command
Tell that its sculptor well those passions read
Which yet survive, stamped on these lifeless things,
The hand that mocked them and the heart that fed.
And on the pedestal these words appear:
`My name is Ozymandias, King of Kings:
Look on my works, ye mighty, and despair!’
Nothing beside remains. Round the decay
Of that colossal wreck, boundless and bare,
The lone and level sands stretch far away".

Brittle Computing is fading quickly. The reality is that it will die. The unfortunate result however is that it appears to have moved into organizations. Brittle Organizations are everywhere. From the shark culture to the not invented here culture there is a Brittleness in many organizations.


Scott Andersen

IASA Fellow

The Brittle Organization Cultural Spectrum…
My Amazon author page!!!!

#BrittleCompute if you want to share something with me about the concept either of Brittle Computing or Brittle Organizations. (please DM me a link as my Twitter Feed scrolls about every 10 minutes I may miss your hashtag!).

There is, a spectrum within and around Brittle Computing and Brittle Organizations. As discussed yesterday within the broad strategic view of an organization (A, B, C, D) there becomes the edges. The closer to D the organization moves the more likely the edges both of the solution and the organization become hard and fast.

Interestingly organizations can become brittle by culture as well. A culture focused on becoming one thing will in the end miss the diversity needed to remain #NotBrittle.

I find many more companies pushing rapidly from A (young, agile) to B and C organizations. Things that should remain agile and flexible become bound the in the corporate culture. You see culture is as brittle as any inflexible org. Where inflexible orgs (in the D space) tend to lose young innovative talent (I can’t innovate here). A and B companies with tight cultures become inflexible for different but still dangerous reasons.

A D company will stifle an innovation and the inventive person leaves the company. The rise and fall of the brilliant ideas within Kodak are a great example of this. A and B companies that create a strict culture driven organization will also experience that except that in their case they won’t hire the diversity and will in the end create a shark culture. If everyone is like everyone else, eventually you are competing for success. That shark culture can take years to move out of a company. It never fails to destroy a company from within.

There is a spectrum and a number of warning signs. (Shark Cultures tend towards Stump the Chump interviews multiple people in the room bombarding the interviewee with questions, there is only one right answer and only the person asking the question knows it). Stump the chump interviews are a huge warning sign. Lack of prompt communication to people interviewing is another. If you see either beware, that is the birth of the Shark Culture.

The spectrum starts with the overall strategic nature of an organization A, B, C, or D.

A Culture:

A Cultures that move into the weaker end of the spectrum end up disappearing – often their technology consumed by a B culture company.

A Cultures that live in the power end of the spectrum can become very successful very fast.

  • Power: Open, innovative and agile solutions.
  • Weakness: Pre-Shark Culture, we are all the same. No diversity of talent in hiring.

B Culture:

B cultures that move to the weaker end of the spectrum signal this by justifying their hiring practices. We are looking for people who are better than the people we have.

B cultures that are at the power end of the spectrum become extremely flexible in integrating new solutions into existing solutions. They are able to remake themselves in the market rapidly.

  • Power: ability to embrace and absorb A companies. Integration of the A concepts allows the existing technologies or solutions of the B company to become better. Move from Niche to market leader.
  • Weakness: A full bore Shark Culture. Every person for themselves. We openly compare employees and people interviewing to people already working in the company. We hire people that act like we do.

C Culture:

C Culture companies that move to the weak end of the pool can become extremely Brittle extremely fast. They often slide quickly from C to D and then out of business.

C Culture Companies that are in the power end begin to move slowly back to the more agile B culture. A B culture company that is embroiled in a Shark Culture can slide into the more sedate C culture, shed the sharks and return successfully to a B culture again.

  • Power: Steady market and ability to move further down the BCG experience curve reducing costs and increasing profitability.
  • Weakness: Full on Shark Culture that in the end continues to replicate itself until the company moves to D and then out of business.

D Culture:a

D Culture Companies that move to the weak end of the pool are in trouble. Reorgs and restructuring will not solve the problem. They are in a morass of cost cutting and price wars. In the end if they remains in this end of the spectrum they will fade out of business.

D Culture Companies that move to the power end of the spectrum can begin to control their market. They won’t be creating innovative solutions but their continued drive toward improvement will create a culture of inclusion. Effective companies in this space are able to contains pockets of A and B companies within their corporate boundaries without crushing them.

  • Power: Pockets of innovation are protected and managed effectively allowing innovative ideas to flow into the business and be leveraged effectively. The company has an established market and is able to protect its market share overall.
  • Weakness: Two ends, one is that the culture has become so Brittle all innovators leave. The other is the toxic reality of an unchecked Shark Culture. The second goes out of business very quickly.

More to come!


Scott Andersen

IASA Fellow

More on Brittle Computing…
My Amazon author page!!!!

The inherent weakness in any Brittle Organization is always the link between the past and the present in that organization. The exceptional book the Blue Ocean Strategy points to finding a way to get away from competition to break out of the Red Ocean and move the organization forward. That is very easy to say but really hard in the end to implement in an organization.

Strategists denote organizations fit into categories (A being a innovative quick moving company, B less quick moving all the way to C not moving fast at all and D a company at steady state). Companies in the D category don’t have to innovate as they have a steady state market. Or do they?

As market’s change companies in the C and D range can’t reorg to create new market’s. They have to look towards acquisition in particular finding A companies that aren’t too far away from their culture. You can’t bring an A culture into the center of a D culture company – the clash would destroy one or the other.

The thing that moves towards A companies is innovation. The thing that moves away from D companies is innovation. D companies are at the bottom end of the BCG experience curve. They are looking to build operational efficiencies into the solution to reduce the cost. They are however in a fixed price market that isn’t volatile. Where A companies are in a new, expanding market that has variable pricing and may have operational issues in the startup process.

Where are D companies brittle? Normally the core of their business – so they launch into reorgs attempting to regain their footing. It won’t work because they are doing things the same way in a collapsing market.

Where are A companies brittle? In the reality of operations usually. They build something – it takes off and in the end there was no way to prepare for the reality of the new market. So there are no operational processes or procedures to manage the reality.

B and C companies are also Brittle and interestingly also fall into the reorganization bottleneck as well. B companies normally do well integrating and are successful in their integration efforts. They purchase a number of smaller A companies and take the innovation forward. IBM and Microsoft are good examples of B companies that have developed a strong ability to integrate and stay in the B area. They tend however towards frequent reorgs to address perceived needs that can be market confusing. They also tend towards strategies that are on the innovative end of the market which at times can lead to operations and delivery issues. Frequent reorgs in the B space cause a loss of market momentum.

C companies also have reorg issues but they have moved more towards the D issues (you can’t shed reality to create innovation). Their markets are maturing so they seek operational efficiencies as ways to reduce cost. They have issues integrating A companies (culture shock) and are often the edge of markets that cause brain drains (people leaving because they have an innovative idea that is just outside the company market). So where A and B companies confuse the market with reorgs, C companies lose their ability to react to the innovation component by brain drain and eventually move towards D companies.

So knowing that everyone fails why does everyone reorg? Its easier to reorganization your company than face the bleak reality of a declining market.

more to come…


Scott Andersen

IASA Fellow

Brittle Compute begat a Brittle Corporate Culture…
My Amazon author page!!!!

Expanding on the concept of Brittle Computing, there was a time in the heyday of Brittle Computing when we designed to remove failure from the equation. We built redundant system after redundant system to result in a system that could not fail. We pushed that design not to fail into the business from IT. In fact in many parts of the business its still there.

The entire time I worked on those architectures I thought about Apollo 13. Yes the astronauts made it safely to earth. But that designed to be fail proof system in the end had just enough fail proof to make it home safely. Not complete its mission but to limp home. There Hero’s of Apollo 13 were people willing to take risks and do things differently because it had to be done that way.

At times that has been the reality of Brittle Compute. Limping home alive rather than surviving and growing. You could fairly argue that this in the end gave birth to the concepts of design for failure but in my opinion that would only be half right.

There are operational concepts that Brittle Computing taught us. ITIL, Cobit and other operational frameworks grew out of the B.C. time. You need to regiment processes to make sure there was less chance of failure. The problem with over regimentation in the end is you actually move the Brittle.

A lot of organizations have Brittle people or a Brittle Culture. I’ve quoted from the movie Babe before, in particular the cow character who says “The ways things are, is the way things are.” That is an example of Brittle Computing moved to the people of the organization.

There is a fluidity in the way things are. If you allow the change to flow around you (design for failure) in the end the organization will be stronger. HR professionals present information about good and bad attrition. The reality is we never really know if perceived good attrition was really good. It may have truly been instead an exposure of the Brittle Culture.

The question is once you realize you are living in the Brittle Culture can you get out? No one wants to have to survive what the astronauts survived on Apollo 13. Most of us watched that event (if were were alive) in horror. Every day I rushed home from school to watch CBS news. Hoping Walter Cronkite would tell me it was all over that they were safe. You see a Brittle Culture in the end will break.

The way things are breaks. Murphy said “anything that can go wrong will go wrong.” In fact as we look at and consider Brittle Cultures it is something that has gone wrong. Insulated and unwilling to take risks. There are strategy books galore (Blue Ocean Theory, Good to Great, Why the Mighty Fail) that detail the reality of a Brittle Culture and what happens next. The question is do you see it?


Scott Andersen

IASA Fellow.

Innovation is a moving target…
My Amazon author page!!!!

What’s next. It seems, and I realize I have been beating this drum for awhile so my apologies on that, that innovation has shifted to the right. Perhaps based on recent project’s I’ve seen further right than I had originally thought.

I think it was my 9th or 10th birthday (it could have been before or after) my parents got me a Radio Shack 100 in 1 electronic project kit. I loved that, building and learning about how power flows. I am not the most handy person with tools that require precision. But I understand the what and how of what’s being built. I can design it in Visio, I just can’t freehand what is needed.

Today however you can get a printer that takes that Visio and makes it into a 3d model. Something I can’t do on my own, but I can with a 3d printer.

You can buy a printer that allows you to print circuit boards. From imagination to reality in the time it takes the device to print. What once could take weeks can now be done in hours.

As the cost of “ideas” is decreased how much faster will the rate of innovation become? Oculus Rift – an interesting 3d viewing system was just purchased for 2 billion dollars. I actually almost backed the project (I picked a different project due to the overall shape and design of the actual unit).

There are more than 20 different 3d printer projects. Most of them now include two nozzles to print objects in two colors at once. There are several different circuit printing projects (5 or 6). Build your own robot kits abound on both KS and IG.

My definition of Innovation – solving a problem that exists in a new way is evolving. I would have once said that innovation is as much about the unique way of solving things as it is a new way. But unique has risks, with new prototyping and 3d printing tools the reality of innovation is that you will in the end be able to innovate rapidly to solve interesting and innovative problems. I suspect my definition is going to move to the right as well.

Ah to live in interesting times…


Scott Andersen

IASA Fellow

Oh Brittle how I miss you.. (not)…
My Amazon author page!!!!

The other side of Brittle Computing was the reality of backups. I remember when I first set-up a company mail system where I was working with a couple of other engineers. They built the system and once it was up they didn’t have anyone to pass it to so it ended up on the helpdesk I managed.

Back in the Brittle Days the helpdesk/Ops team ended up with a lot of internal/external customer stuff no one else wanted.

The company I worked for was a technology company so adapting the changes required for the email system to flow within the boundaries of the company was pretty straight forward. Initially it was LAN and dial-up and then LAN/WAN and a couple of dial-ups. MSMail was a 16 bit application and was optimized to run on 386 chips. It had no idea was to do with more than a meg of RAM. It really didn’t care how fast the processor was and in the end its biggest drawback was the reality of hard drives. We didn’t have 15,000 RPM drives back then.

With that solution the critical things for the post office were the externals (that moved mail) and the main post office. As long as those two were functioning everything else would queue. The main post office had three primary roles – center of the spider web for routing mail, in our case eventually the Internet connection and finally as the primary post office for building the main address list. In the MSMail days building the address list had three primary times, T1 where you sent a request out of every post office asking for new accounts, modified accounts and deleted accounts. T2 was where the main post office took all those lists and merged them into a new master list and then finally T3 where the list was set out. Any failure in that 3 step process of course meant you didn’t sync a proper GAL that night.

Over the course of my career I worked with a number of mail systems. I can honestly say the reality of the directory being the point of failure in a mail system is down to nearly never now. Back then the directory was always off just a little. The whole reality of the 3 slots at night to build the new list was risky at times. When you added in other mail systems and attempted to build a master list for all of them it got even trickier.

In essence the reality of Brittle Computing was the willingness to drop troubleshooting and return to a known good backup. Where today I backup my critical files (family pictures, my books etc.) to a cloud back provider I no longer backup my applications. Why? Because I can go to the providers website and download that application and reinstall it easily.

Once, I had a PO drop off my network. It was back in the transition to Novell 4.0 and an entire branch of our company upgraded to Novell 4.0 without bothering to tell anyone. You had to upgrade the servers to include Bindery Emulation if you had 3.1 servers still in the system. The Branch manager called me on day two of the outage and yelled at me for an hour. I kept telling him that if I couldn’t see the drive in MSMail it didn’t matter if Novell had give them the software for free they had to downgrade their server.

You could easily break a Brittle System.


Scott Andersen

IASA Fellow

back to Brittle Computing…
My Amazon author page!!!!

Back to the theory behind Brittle Computing. Brittle is an interesting word in that it implies a loss of tinsel strength. The ability of a solution to withstand pressure in completing a function. Brittle computers were perfectly capability of in fact handling the pressure. We just had to be careful around the edges of the solution.

The compute solutions we have today allow for the design for failure concept. They don’t put you into a reactive support mode when they in the end fail. Well that is the dream, I am not sure every solution does that but we are moving that direction. It comes from the days of brittle compute where we were always worried about what would happen next.

Today we throw out the words COOp, DR and HA frequently. In the days of Brittle Compute we didn’t always have DR and HA capabilities without spending a significant amount of money. That cost factor drove many businesses to rely on solutions that weren’t perfect but hit the good enough mark. Like a rubber-band however, in the end if stretched to far they would eventually snap. Hence the term Brittle Compute, and the reality of the world we used to live in.

There wasn’t a backup. There wasn’t always an Internet we could go to download the latest drivers. There wasn’t a Google where you could in the end download the answer in virtually any IT problem that was plaguing you. You see without a backup system or a duplicate spare or another cloud compute system ready to go you had to fix the problem.

Or wait for the tech.

Downtime was measured in hours not seconds. 4 hours wasn’t unusual. 2-4 outages in a year wasn’t unheard of. It wasn’t 4 9’s we were chasing or even 2 nine’s. Many companies lived with systems that were 98.1 or lower. Human bodies were warmer than the uptime some IT shops could achieve.

OF course from Brittle Computing was born ITIL and Design for Failure. I like to call ITIL the “DBSDTSR” framework. It stands for Don’t be stupid design the system right. It provided a framework to optimize the what we were working on without creating new or additional risks for the organization.


Scott Andersen

IASA Fellow.

Lighting, printers and Monday
My Amazon author page!!!!

Neon accent lighting for the home office is now officially something I looked at. I feel embarrassed to admit that. They say however, the first step to self improvement is admitting you have a problem.

The problem with a home office that you add to a home you’ve purchased is that in the end the lighting is always wrong. For me there are a couple of lights that need to move about a foot one way or the other.

Perfection is not just a state of mind. In the end it becomes an expression of what you want to do.

We are moving to wireless and network printers in the house. I am not sure why in the end we have so many printers. It’s a legacy from the old days. I don’t print anywhere near as often as I used to. Barb still prints frequently using both color and laser printing. I find myself printing less and less often.

Partially because I don’t need to. But I still have a few printers in the house. Mostly the odd format printers (large and photo) that allows you to print various things easily. Its nice to be able to take your phone, plug it in and print a picture you’ve taken. The plotter is from the days when I was a practicing software architect and needed to look at paper versions of networks and larger workflows. I haven’t done a print out on the plotter now since my last big project which ended a year ago.

The thing about printers is, when you need one it helps to have one available. When you don’t need one it kind of sits there taking up space. Catch-22 Mr. Heller, thanks for creating the term. In the end however, you have to have them.


Scott Andersen

IASA Fellow.

Free Cloud Storage–no really its free…
My Amazon author page!!!!

There is an interesting gamble going on right now in the technology/cloud space that I find intriguing. Google slashed the pricing on Google Drive to 1.99 for 100 gigs. In order to get that amount of storage on an HD you would in the end have to pay about 60 dollars. If you ran it for a year or two, you would hit the MTBF and would ultimately need to replace that drive. That means every 2-4 years you buy a new drive. Figure a fair number is 100 bucks.

At the time time for the same space you spend 4 bucks with Google. what a value proposition.

Now, the intriguing part. When you purchase new hardware (a new Chromebook for example) they give you a two year free account with 100 gigs of Google Drive space (or 50 gigs of Dropbox or 50 gigs of Box storage). They are betting that after using the space for two years you aren’t going to want to stop. I am better that pricing will drop through the floor in two years and I will pay a dollars for two more years of Google Drive or OneDrive, or Box, or Ubuntu or Dropbox and so on and so forth.

The freebies model is intriguing but in the end not as risky as you would think. I upgraded to a larger Amazon drive and was willing to continue paying for the storage why? Because Amazon preloads the CD’s I buy into my Amazon drive.

The value of CloudHQ has a solution increases in the new world. I still think their subscriptions are far too expensive overall but now, with 200 or more gig of cloud storage you can find a much stronger case for moving the data around.

This leads me to an interesting evaluation going forward. What are the criteria you need to consider in selecting cloud storage going forward.

Vendor name

Ease of Use

Ease of Access

Storage types supported

Overall cost

X cloud storage        
Y Cloud Storage        

Based on this framework we can begin evaluating the drives spaces on what we need overall. For example I buy new music from Amazon and Apple – so ease of use is high for those two. Ease of access is less so for Apple but improving. In building out the matrix you can evaluate the drives for both enterprise and personal use. I am not sure that you would add anything other than overall security for an enterprise solution in this space.

The times they are a’changing…


Scott Andersen

IASA Fellow!

The best way to move forward is to learn from your mistakes.
My Amazon author page!!!!

Brittle Computing (continued)

When I started out I was a helpdesk professional. There is a certain reality to working on a helpdesk. You learn quickly that there is a human being on the other end of the phone. You also learn quickly what the skills and capabilities of that person are.

I used to have all sorts of support processes memorized. I could without seeing a computer screen walk someone to the configuration tab of a Lotus Notes Connector in Exchange 5.5. I used to (but can’t now too much other stuff in my head) be able to do Windows NT 4 configuration without seeing the screen. That is a skill you perfect on the helpdesk. Seeing what is actually on the other side of the screen.

That said it is also a huge part of what I believe to be Brittle Computing. There is a corollary to Murphy’s Law that says loosely work on something long enough to improve it and it will break. That is the reality of Brittle Computing. You couldn’t make wholesale changes without a path backward.

With Cloud solutions you can easily (and intelligently) remove some of that risk. Virtualization gives us the reality of a snap shot of our VM’s so that we simply roll back easily and quickly. The concept of design for failure is a carry forward from the Brittle Computing days but the meaning has changed a little.

Brittle Design for Failure: Have a backup, have an outage window, start the upgrade. Have warm spare ready. Verify upgrade worked (or failed) if failed go to backup. Bring backup online, and evaluate why the system failed to update properly (root cause analysis).

Modern Design for Failure: Build resilient system that can adapt around failures. If the system fails during upgrade (part b) bring part a online. What’s an outage window?

The best way to move forward is to learn from your mistakes!

That is the motto of a Brittle Computing person.


Scott Andersen

IASA Fellow.