Wi-Fi has finally come to Logan Airport in Boston. Massport thinks they can do a better job delivering wi-fi service than T-Mobile, Verizon, or AT&T.; I tried the service yesterday while in the United Red Carpet club. I won’t buy it again. They are blocking certain network protocols. My VPN client is not functional with their service. With no VPN access I cannot do work. Massport has lost this business traveler and heavy wi-fi user.

posted: 2:38:24 PM      comments ()    

I’m in ITIL training for the next couple of days in Chicago. The majority is the first day is spent doing an airport operation simulation. It has been very interesting to watch the interaction between the teams. We are actually competing with another team for revenue as a result of flight arrivals and departures. One interesting tidbit coming out of the presentation between competition rounds this morning is that 60% to 90% of all helpdesk calls are a result of changes made by IT.

posted: 2:31:12 PM      comments ()    

A great article on blogging over at CCN featuring Joi Ito.

posted: 12:35:55 PM      comments ()    

Whatever happened to this, that, and the other thing? The closest thing to Open Source in storage is Samba, which is purely bridging software. Combining physical storage and the necessary software to manage a large array is, by its very nature, a proprietary act. The truth is that there is no such thing at Open Source when it comes to storage. Could this change? Only if someone compiled a do it yourself storage array blueprint which, in the end, would end up being proprietary. A storage array is very much like a car. Can you build your own? Sure. Do the majority of the people who own cars build their own? No. Why? Cost, complexity, and skills.

posted: 1:47:45 PM      comments ()    

Rangachari Anand dropped by and left an interesting comment about vendor adoption of Open Source. This deserves its own graphic. Vendors adopt Open Source for different reasons, but commonly as a springboard for getting a product to market. Once in the marketplace, the vendor continues to innovate and deliver customer driven features which take them away from their Open Source origins. Very few vendors feed their features back into the Open Source community from whence they came, which is a shame. Technology is about customer driven innovation. This is one reason why companies can innovate much faster, efficiently, and in a focused manner than communities.

posted: 9:29:49 AM      comments ()    

A very interesting post over at EMERGIC.org on storage costs. Rajesh says we are at about $1 per GB of raw, bare metal storage. The cost of a gigabyte in the EMC Symmetrix DMX800 in the data center I manager is about $4 per month, or $48 annually. To be fair, my $4 per month includes the frame, software to manage the storage, and maintenance in addition to the raw disk. However, I just received a quote to add more disks to an EMC frame. If I were to buy just the raw disk from EMC it would be at a cost of $19 per GB. Of course the type of drive that costs $1/GB is very different (ATA or SATA) from the drive that costs $19/GB (SCSI). However, I find it hard to believe that it costs nineteen times more to produce and sell a SCSI drive than an ATA or SATA drive. There are some interesting economics going on here.

posted: 7:22:49 PM      comments ()    

A part of the conversation I have been having with my colleague included the development of a graphic below, which was originally written on a napkin. The point of this graphic is that there are two market segments that will readily adopt Open Source software for their projects and contribute back to the Open Source community. Those two segments represent the two ends of a bell curve. At the low end of the bell curve are independent or small businesses that cannot afford enterprise solutions due to cost. At the high end of the bell curve are organizations that find enterprise solutions too restrictive and have significant funding available to produce highly customized solutions on their own. The middle of the bell curve represents the enterprise market. The reason the enterprise market is reluctant to adopt Open Source software and solutions are that these solutions require expertise to maintain that they are not willing to keep on staff, namely more people with specialized skills. The majority of the enterprise market does not see technology as being a core competency.

(Click on image to see a larger version.)

This diagram will not change until the "trusted technology advisors" (a term which I despise) can put Open Source in business terms, namely operational cost savings. What do you think?

posted: 6:15:38 PM      comments ()    

Why would you open source something in the face of huge profits? I have been having a conversation with a colleague of mine on this very topic for a couple of weeks now. My answer to him is that you would only open source a software package in two instances. One, when there is no money to be made but many people would find value in the software as a utility, such as MRTG and even PERL. Two, when open sourcing the software serves as a catalyst for increased revenue and profits, such as Mac OS X and UserLand's Frontier kernel. In case two the potential revenue and profits must be exponentially greater than retaining the intellectual capital contained in the source code. Unfortunately, the jury is still out on case two as hard numbers have not come in to validate this model. What do you think?

posted: 5:23:40 PM      comments ()    

My brother works for LumenIQ, which produces security and forensics imaging analysis applications. LumenIQ will be featured on a new Discovery Channel series entitled "Proof Positive - The Evidence." The series premieres on June 5, 2004, at 9pm EDT. Check your cable listings for local times.

posted: 5:56:06 PM      comments ()    

A good thing is happening with my RSS subscriptions. More individuals are appearing in my subscription list than feeds from news organizations and vendors and system alerts. This is quite a change from a couple of months ago when my RSS subscriptions were dominated by draw, boring news and system alerts.

posted: 2:52:37 PM      comments ()    

I haven't posted this week because I haven’t' had time. My mother was out visiting from Seattle the first part of the week. I was playing catch up from taking Monday and Tuesday off until today. I had to take a day trip into Boston to collect stuff from an office my company is closing and spend time tracking down the facilities guy in the office I'll be sitting in starting in a few weeks. I was also dealing with some political nastiness which totally demotivated me for a couple of days. I'll start posting again this weekend. I do have some very cool stuff to write up.

posted: 11:05:44 PM      comments ()    

Many lessons can be learned from the Petabox project over at the Internet Archive project. An increasing need for mobile, large scale, inexpensive storage facilities exists. Expanding on the need for these storage facilities is also the need for mobile, large scale, inexpensive computing, power, and operations facilities. Facilities like this exist for the military, oil, and mining industries, but are extremely expensive. The Petabox has a TCO over three years of approximately $2.1MM. This is amazing. A convoy of six vehicles – storage, computing, operations, two power trailers, and a fuel tanker -- could operate a huge mobile data collection facility. Each container could be outfitted with independent communications equipment. External and protected patch panels would interconnect all of the containers in addition to power connections. These are the building blocks for large mobile computing facilities. With ever increasing concerns of terrorism, these types of facilities will be employed by industry as a risk avoidance and mitigation strategy in urban areas.

posted: 1:26:50 PM      comments ()