Category Archives: internet

I miss CompUSA

I was reading Brent Leary’s posting Christmas 2011: a Great Example of Smarter Commerce in Action | SmartData Collective tonight and it got me thinking.

I remember going retail shopping with the only motivation being to find something cool to own and bring home.

I don’t do that anymore.

I suspect there are a lot of reasons: I’m older and more careful with the family coffers.  I’ve got all the stuff I need.  But honestly, I think it all boils down to this:

Retail outlets, because of their reduced SKUs (stock keeping units), have less interesting stuff for me to buy. There is less variety in the stores.

And I think this is a vicious cycle that is eating retail.  Retailers, with the high cost of distributed inventory, reduce the number of SKUs to only those items with the most mass-appeal. People go to the stores, can’t find what they are looking for, go home, get on the web and buy it.

And web retailers are getting much better at fulfillment:

  • shipping is increasingly free: 93% for this past Christmas vs. 85% for the season before (let’s be real: it is built into the price) (see USA Today article quoted by Brent)
  • shipping is increasingly fast

Let’s dwell on the last bullet.  My younger son had a small fender-bender with one of the family vehicles requiring a turn signal lens to be replaced.  On Friday, I placed an order with, with normal, non-expedited shipping.  I received the lens on Monday.

I miss CompUSA (the retailer not the current web store).  Toward the end of CompUSA’s retail existence it was awful, but there was a point that they had a wide variety of SKUs and I could always walk out with something I needed but didn’t go in to buy.

Microcenter, the closest of which is over two hours away in Boston, is still like the old CompUSA somewhat. That being said, the last time I was in one, I was disappointed that they, too, had cut back on the SKUs.  I haven’t been to a Fry’s for a while, but suspect it is the same story there, too.

For a reality check, have a look at this article in the NY Times about the problem of malls.

Short of Star Trek transporters or replicators, I’m not sure what the answer is to change this transformation.  One thing that I wonder is whether there is a point when the current just-in-time mass-appeal SKU/local inventory system breaks because of the cost of fuel for 18 wheelers.  One could imagine a point where it is actually makes sense to spend less money on fuel by having very full trucks of goods go to retail stores and create larger local inventories that have to be replenished less often.

I wonder, also, about the fabric of society that shops local retail less.  Is that a good thing?

If you’ve got some opinions on this, I’d love to hear from you by way of comment.

Leave a comment

Posted by on February 8, 2012 in business model, internet, retail


IT, users, tools, and machine shops

I was reading the article “Bring your own apps: The new consumer threat to the CIO” on TechRepublic this morning.  I had to chuckle.

The lead sentence,”The CIO’s control over workplace IT is gradually slipping away as today’s digitally-savvy workforce have decided they want to call the shots when it comes to the technology they use at work,” could have come out of a Computerworld from 1983.  The IBM PC had been released in 1981 and in January of 1983 Lotus 1-2-3 allowed users to take IT into their hands.

Backing up, historically speaking, we find:

“There is no reason anyone would want a computer in their home.” — Ken Olson, president, chairman and founder of Digital Equipment Corp., 1977, and

“I think there is a world market for maybe five computers.” — Thomas Watson, chairman of IBM, 1943.

(both courtesy “Bad Predictions”)

What does this “history repeating” moment tell us?

It has always been about users, taking what’s available to get what they need, when they need it, the way they need it. Technology is not an end onto itself.  Technology is a tool to get something done.

And, ultimately, isn’t how we make, use, and improve our tools fundamental to our humanity?

It has been said that IT departments should be tool-chests for users.  I’d argue that they need to be more like a  machine shop, allowing users to craft their own tools.  IT departments need to furnish the nuts and bolts of their enterprises and the tools to use those parts to fabricate whole new tools.

I think we need to stop clinging to old models. We talk about mashups with its throwback to the term “lash-up” which Merriam-Webster defines “as any improvised arrangement for temporary use.” To use this term actually belittles the activity: it implies that mashups are put together until IT comes up with a more permanent solution.  Actually, mashups are a tool-defining activity onto themselves: users getting what they need, when they need it, the way they need it. Let’s give credit where credit is due: the user!


Tags: , , , , ,

Google can’t afford to sell people’s personal data

You can’t believe the number of people that have asked me about what I think about Google’s new privacy policy.  I’m sure this might not attract a lot of fans, but here’s my take:

  • Plenty of other media sites already integrate tracking of people’s experiences across their offerings on the web.  Google goes out of its way to actual say they are going to do this… and they get criticized.
  • I know of plenty of commercial and educational sites that use Google Apps as the foundation for their business.  I personally know a well regarded security expert at one of the world’s largest medical supply companies that conducted a security and privacy review of Google Apps.  The fact this company now uses Google Apps is a testimony in itself.
  • If a company shares personal data inside the company (say, for example, Verizon sharing people’s wireless and FIOS usage inside of itself) is it really a breach of privacy?  When Verizon does this, they call it bundling and its customers save money.  When Google does this, it is called evil.

I hear now that members of Congress want to call Google before it again because of this issue. I’m not sure this is the best use of their time with lots of other important issues before it… and I’ll leave it at that.

More than anything, I think this fear is about the fact the Google is very big.  I can’t remember which of the talkshows I recently heard this on, but the observation that was made was that right now, Americans fear anything that is big: big government or big business. We fear that which we do not understand.

My fourth and most important bullet is this:

  • Google has reiterated (though some seem not to hear it) that it still isn’t going to sell  personal data.  And do you know why they won’t sell personal data? Because if they were caught selling personal data, it would destroy their entire business model. Not selling personal data is good business for Google. Let’s be real clear about this: if Google were selling people’s personal data, no one would use Google.  If no one used Google, there would be no way it could sell advertising.  Google is making plenty of money without selling people’s personal data.  The bottom line is simple: Google can’t afford to sell people’s personal data.

Tags: , , ,

When “new stuff” is decades old

Of course, this is just going to be the rantings of an old fart… but I guess that’s part of the reason to have a blog, right?  To rant?

Geek alert… this is going to be a bit technical.

The other day we were discussing the need to move an internal website from a rough and tumble, less-than-actively-supported environment to something more mainstream (as in, supported by IT).  I reflected that if the prototype environment was contained to single disk partition, it would be relatively easy to make a virtual image of it and move it to a supported VMware server and run it intact. (Translation: if the website was on a single disk drive, we could copy that disk drive to a file, which could be used by a piece of software to simulate a real computer).

The response I got was to the effect “yeah, that would work and wow do you know your stuff!”

I had to chuckle.  Virtualization, the process of having a real computer simulate one or more computers, has been around since 1967 when IBM created an operating system called CP-40 (and later, CP-67).  I first encountered virtualization more than a decade later when in 1979 I was a system programmer working at Hewlett Packard in their mainframe data center. I was responsible for installing IBM’s then current virtualizing operating system, VM (for virtual machine) on HP’s multi-million dollar Amdahl mainframe.  At that time HP only had one mainframe (hey, they cost a lot of money, even for HP) and if a systems programmer wanted to try out a change to the operating system, you’d need to come in on the weekend for the few hours that the data center wasn’t running. By installing VM, we could run two simulations of our physical mainframe.  One would run our production operating system, and allow business to carry on as usual.  The other we could use to test new versions of the operating system.  No more weekend testing!

Later, as I started to develop PC software, I watched how Intel added capabilities of their microprocessor chip.  By the time Intel announced the 80386 version of their microprocessor chip in 1985, they had added everything needed so that it could simulate multiple computers using virtualization. It wasn’t until 1998 that VMware was formed and created the first software to virtualize the PC.

Being a virtualization affectionado I’ve been experimenting and using VMware’s software since 2001. This included (and still includes) running their Mac OS X specific version, Fusion, on my Mac Book Air.  For those that are PC-only literate, Fusion allows me to run Apple’s Mac OS X operating system AND SIMULTANEOUSLY run Windows 7 on my Mac Book Air.  I can readily switch between the two environments, including cutting and pasting (a version of which had been present in IBM’s VM mainframe operating system in the early 1980’s).

VMware is not the only company that provides virtualizing software.  There are even open source versions.

So, referring back to the comment that set this off… yeah, I know this stuff.  Been there, done that… and even in more than one environment.

But I think the bigger picture is this… in any maturing industry, great ideas are going to be reused.  I’ve watched from a point where computers were so expensive that everyone had to share to a point where computers are so cheap everyone has one (or more!). I’ve watched the data move from that centralized model, where all the data is in one place, to a decentralized model (first with distributed minicomputers and later PCs on local area networks) and now back again… to the “cloud.”

Likewise, social media has its roots in technology that was created in the earliest days of computing.  In the earliest days there was commercial systems like Compuserve, but even before Compuserve there were systems such as Douglas Engelbert’s NLS demonstrated in 1968. Oh, by the way, Engelbert demonstrated the mouse at the same time as part of the user interface to NLS.

Business models need to take this phenomena into account.  Not only do you need to be aware of the side attacks that Clayton Christensen talks about in “The Innovator’s Dilemma,” you need to see where yesterday’s solutions might once again solve the problem’s your product is solving today.

You’ve heard it before: history repeats itself.  Unfortunately, I think some technology companies think they are immune from that law. They are not. And there is nothing more embarrassing than to lose to the past.

Oh, and if you are looking for an industry to model where the past is new again, look west to Hollywood and east to Broadway.  They’ve realized this and profited from it for years.


Tags: , , , , ,

OpenDNS & DNSCrypt – Cool new security for the web

I’ve got to hand it to the folks at OpenDNS… they are really, really, smart cookies.

I think Clayton Christensen would be proud. They took a pretty mundane job that needs to be done, reliable domain name services (DNS), and have created the number one DNS service on the internet.  They are so big, it is difficult to see a competitor trying to come after them.  In fact, one has to wonder why Google didn’t think of this… and why they aren’t competing.

DNS is the internet service your computer uses to find servers on the web.  When you type in “” into your browser, the network software in your computer sends it to a DNS server which returns a numeric IP address.  Think of DNS like a phone book: you search for a name and end up with a phone number.

All internet service providers (ISPs) provide DNS services.  Unfortunately, while DNS is very important to the proper operation of their network, they really don’t spend a lot of time optimizing it.  What does that mean to you?  Slow lookups… which means it takes longer to get to the website you want to view.

OpenDNS solved that by creating a really, really, REALLY fast and reliable DNS.  It is likely that if you change your computer to use OpenDNS, you will have a faster browsing experience.

Best of all, OpenDNS is free.

But the folks at OpenDNS didn’t stop there.  They added things parental controls and usage monitor. Doing it in a DNS server is the perfect place to do it, instead of each computer, because it is centralized in “the cloud.”

What is OpenDNS’s economic incentive to do this?  Just like Google sells information about the click throughs that occur when people search on Google, OpenDNS sells information about who is looking up which websites when.

Additionally, OpenDNS sells premium DNS services to large organizations like BP and Eastern Mountain Sports.

Now OpenDNS has created DNSCrypt.  What’s that and how does it make them money?

Well, when your computer talks to a regular DNS, it does so in a very trusting manner.  DNS was designed for an internet where the network was completely trusted: when a computer wishes to be directed to a computer at a specific numeric IP address, it was assumed the network would reliably and truthfully directed you to that computer.

Unfortunately, when your computer is in the wild, such as on a WiFi hotspot, your computer may not be on a trustworthy network.  The network can actually intercept requests going to a DNS server and provide nefarious responses.

For example, let’s say your computer is asking to go to and you are using your computer on a bad guy’s WiFi hotspot (or a good hotspot that has been hacked).  The evil network can imitate the DNS your computer was trying to use and direct it to a bad guy’s website that looks like your bank’s website.  Then all the bad guy’s website needs to do is simulate the login page of your bank.  You’ll enter your userid and password, they’ll save it, and then pass you to the bank’s website.  It might appear as if you’ve typed the wrong password, so once you are at the legitimate bank’s website, you’ll be prompted again, and this time login into your bank’s website. You won’t even know your userid and password have been compromised.

(You might ask about the lock icon in the status line of your browser and why that can’t be trusted.  Most people don’t even pay attention to the lock icon anymore, and that can be hacked, too, for lots of other reasons we won’t go into here.)

The bottom line is the bad guys can trick your computer into tricking you to give up your userid and password. This is called man-in-the-middle attack.

What DNSCrypt does is this: it causes your computer to only use a DNS server if that server can speak a secret language.  In more technical terms, it encrypts all the requests flowing from your computer to a DNS server.  Since this is encrypted, the bad guys can’t get in the middle and pretend to be a DNS.

What is the business opportunity, then for OpenDNS by providing DNSCrypt?  By providing DNSCrypt, more people will use OpenDNS, resulting in more websites being looked up by OpenDNS, and resulting in more information for OpenDNS to sell.  In fact, one could see a day virtually everyone who has a mobile device uses DNSCrypt and by extension OpenDNS.

Honestly, you got to love these folks… this is brilliant way to make money, much in the same brilliant way that Google makes their money.  Given what OpenDNS has accomplished already, I see them being right up there with the significant web businesses in the future. Further, I’m convinced they’ve got a bunch of other cool money making ideas up their sleeves.

OpenDNS-a company to watch.

%d bloggers like this: