2011-11-09

Careful Wording

I was listening to one of my favorite podcasts, A Way With Words, this morning when I was reminded of an incident from long ago, where careful word selection almost got me in trouble.

I was an IT director at the time, and my company was in the middle of a break-up, a split, and I found myself needing to write a report detailing the plan for addressing the IT infrastructure during and after this split. Some economies of scale were going to be lost and some resources, once shared, would have to be duplicated to affect perfect autonomy.

The two roughly equal sized business units of the company in question were so different that one could hardly imagine them being run as a single entity in the first place. It was not that their industries weren't aligned or their client bases weren't coincidental; they were. In fact, on paper it was a perfect fit in terms of vertical integration.  The difference between the two business units was the vast deltas in areas of management philosophy, staff personalities, and both short and long-term goals.  The ownership structure itself even perpetuated this difference by having two companies held within a third holding company.

The IT report I was writing was simple enough.  These types of analyses are fairly black and white if you treat them fairly and objectively while giving a little forethought to each entities future.  It also doesn't hurt to have an already established reputation for fairness and objectivity with the ownership groups.  I also tried to keep a balance sheet going as I wrote the report to both keep the expenditures for each entity roughly equal as well as ensuring that each had an appropriate mix of new and legacy systems.  I didn't want either company to be shiny and new, leaving the other to deal with older slower legacy systems to deal with.

The problem arose when I chose the title for the report, "Information Systems Dichotomization Plan".

To my mind, dichotomization was a better word than bifurcation, because dichotomization infers that the split is along a naturally occurring boundary or logical topology feature.  Bifurcation has no such distinction, simply referring to splitting a thing into two separate units.

Apparently it struck a a nerve, because almost immediately after publishing my report to the leaders of both units, one of the principals came straight to my office asking me why I used the term dichotomize instead of split or bifurcation, the terms that they had been using internally for their discussions.  Apparently, to him, dichotomize had a third connotation which inferred that the the two things being split should never have been conjoined in the first place.  Once he said this aloud, I realized that, entirely subconsciously,  I too had a vague sense of that inference when I selected the word.

He then asked me if I thought that this split was the "right thing" for the company.  Unfortunately, I had to concede that while such a split was, on paper, a mistake, that due to the realities of the personalities involved  that I, some time back, had resigned myself to this course of action's inevitability if not its necessity.

The situation ended very well, but I'll never forget getting "called out" for that one word in a twenty page report again.

2011-09-22

Shovel Ready Jobs

OK, I was watching the Republican debate tonight and I've got to weigh in on this.

For the purpose of providing both disclosure and qualifications; I've been working within the Engineering / Construction industry for 22 years, and I am lucky enough to still remain employed by a small engineering firm.

In the last few years, I've seen Engineering firms collapse, reduce their staff by up to 90%, and get acquired by global-sized firms looking for local talent.  All the small boutique firms are gone.

What we have now, here in Florida at least, are a few mega-firms that have  gobbled up most of the mid-sized firms that were left, and about million (or so it seems) one-person firms that are comprised of former mid-to-large firm staffers that have been downsized nowhere to go except to start a one person firm in their garage.  These are engineers, CAD drafters, surveyors, landscape architects, and surveyors, your average middle-class professionals and many are friends and former colleagues.

Now when the few RFPs or RFQs come out, instead of 10 firms submitting qualification packages, 3 getting shortlisted for formal presentations (which cost a lot of money and manpower to put together) and 1 getting selected, we get 100 submitting, 10 getting shortlisted (all that much more money wasted on the presentations) and 1 selected.  This is not a sustainable model.

Shovel Ready?  Really?  If by "shovel ready", you mean that all the design and engineering work is already done and the project is ready for construction, then you're talking about an awful lot of hungry people cut out of the loop.

Shovel Ready?  Where are these stockpiles of pre-designed, engineered, and permitted plan sets supposed to be? Sitting on some shelf somewhere in the basement of the Mayor's house?  What happens when they run out of these magical plans?  There will be no backlog of work for the shovels if the pencil-ready jobs are not commissioned too.

How about job-and-value-creation-ready projects?

Look, a typical infrastructure project first employs private sector engineers, CAD drafters, surveyors, landscape architects, and surveyors as well as all their support staff.  Secondly it provides purpose and extracts value from public sector engineers and planners reviewing and approving projects and construction plans.  Third, it provides jobs for the contractors, carpenters, plumbers, electricians, etc. that put hand to earth.  Fifth, pubic infrastructure projects provide value to the public in terms of their improvement in our quality of life whether it be from improved stormwater drainage, decreased traffic congestion, cleaner water, cheaper energy, or even a rejuvenation of a blighted area. Lastly, these projects provide secondary economic value in the form of opportunities created through the higher property values or higher use potentials for properties in the proximity of these improvement.

Now I know I'm sounding a little Keynesian here, but that's a whole other story.  I don't think that it's either a wholesale Keynesian or Friedman/monetarist world we live in.  What I'm just tired of hearing is this misnomer "Shovel Ready".

2011-08-03

FCC Report on Consumer Broadband Performance is Out!

The FCC Report on the performance of consumer wireline broadband entitled Measuring Broadband America is out. No big suprises, but a few small insights for some customers. The link to the report itself is at the end.

Here's my executive summary:

  • Most ISPs delivered actual download speeds within 20% of advertised speeds, with modest performance declines during peak periods.
  • Upload performance is much less affected than download performance during peak periods.
  • Only Charter, Comcast, Cox, and Verizon (FiOS, not DSL) met or exceeded their advertised sustained download speeds using a 24-hour average.
  • Only Comcast and Verizon (FiOS) met or exceeded their advertised sustained download speeds during peak times (7pm - 11pm Mon-Fri)
  • Advertised upload speeds fared better, with only AT&T, Insight, Mediacom, Qwest, TimeWarner, and Windstream failing to meet their advertised speeds using a 24-hour average.
  • The highest sustained download speed as a percentage of advertised speed went to Verizon (FiOS) by a clear margin of approximately 15%. The sustained upload speed winner was Cox.
  • The latency tests showed tha fiber had the lowest latency, with cable coming in second, and DSL dead last.
  • The Web-Loading times show that, in my opinion, Comcast is doing some good work in the area of traffic shaping or caching with a statistically significant faster page-load time especially in their lower speed tier offerings.
  • Performance variation shows that Verizon (FiOS) keeps a nearly perfect straight line performance no matter the time of day, while most others experience at least a 10%-20% dip during peak hours. Cablevision clearly tanks under peak times loads.

The full FCC Report.

Hope this helps!

2011-07-14

Listen Up Android Team

OK, Google Android team, listen and listen good.  I love my Droid-X, really, I do, but I've just gotta get this off my chest.  I just lost about the thousandth opportunity to take a spontaneous picture on my phone, all because of you.  The camera circuitry is fine, the 1GHz CPU is adequate; it's not Motorola's fault, it's yours.

The pictures are lost because it takes so very long to stop everything and pull up the Camera App.  When I press and hold down that camera button, I want a hardware interrupt to surge through the kernel and have a kernel module (not user-space) to kill/suspend my current process and renice every other process on that phone to 20, and pump up the Camera App process priority to -20. See, because when a photo opportunity presents itself, I don't care if somebody just tweeted the cutest picture of their cat, I just want to take the picture before the opportunity is over.  And it's always over by the time the camera is ready on an Android phone.

For you non-Linux heads, in UNIX/Linux, priorities range from 20 to -20, where 20 is the lowest priority and -20 is the highest.

I've heard it from numerous users both personally (our company uses Android phones) and on the 'Net.  Even Microsoft has poked fun at this problem in their Windows Phone 7 commercial; you know, the one where the guys are sky-diving.

Related to this is the fact that we, the users, need control of the daemons/services running in the back ground.  No, not just the baby game of letting me kill services, we need to be able to control what services are allowed to start-up.  Otherwise, they just re-spawn all over again, eating our resources and causing latencies.  And if the Carriers like Verizon, AT&T, or Sprint try to stop it, call them out in public by name and let the consumers make their voices heard.

There.  It's off my chest.  I feel a little better.

2011-04-13

A Bad Bill in Florida's Senate - SB 0276

While in general I do not typically weigh-in in-depth on political issues, I feel that an issue has arisen that is both one in which I have more than sufficient experience and is an important enough issue with disconcerting far-reaching ramifications that I am compelled to speak.

In the interests of full disclosure and qualification, while I am not a Professional Engineer, Surveyor, or Architect I am currently employed by an AEC firm and have worked in IT within the construction industry for over 20-years.

Let me be perfectly clear here.  The opinions expressed here are strictly my own and may not reflect those of my employer.  I don't know, as they weren't consulted on the matter.

The Florida Senate has a bill on its Committee on Education Pre-K - 12  agenda , SB 0276 -Relating to Procurement of Professional Services , which is a patently bad bill. The bill is sponsored by Senator Bennett, and is  designed to change the standard by which professional consultants including Professional Engineers, Surveyors, and Architects are selected for government work, ostensibly to reduce government costs.  While I applaud the end-goal, I believe that the bill is quite short sighted.  Especially when one considers that the design fees are typically less than 20% of construction costs.

Since 1972, Qualifications Based Selection (QBS) is the law in 47 of 50 states and in Florida our Consultants’ Competitive Negotiations Act (CCNA – FS287.055) mirrors the federal rules in The Brooks Act (Public Law 92-582, 40 U.S.C. 1101 et. seq. and Federal Acquisition Regulations, subpart 36.6).  This standard ensures that professionals, whose decisions directly impact the public good and safety such as Engineers, Surveyors, and Architects are selected for projects through a public selection process based on their qualifications, experience, and measurable past performance.  Once selected, a fee proposal and negotiation is undertaken to ensure that the state is paying only reasonable and customary fees for the work performed.  If a fair contract cannot be reached with the selected firm, the next-highest ranking firm may then be selected.

This CCNA principle, that government agencies select a professional services consulting firm for specific projects based on qualifications rather than on a “lowest bid” basis, is designed to protect the health and safety of the public and ensures the procurement process for these specific professional services are based on qualifications, competence, and track record.

CCNA is PROVEN: Public agencies that use Qualifications-Based Selection (QBS) to procure professional A/E services are better able to control construction costs and achieve a consistently high degree of project satisfaction than those using other procurement methods, according to a two-year study led by Paul S. Chinowsky, PhD of the University of Colorado and Gordon A. Kingsley, PhD of Georgia Tech.  There are no competing or conflicting studies!

CCNA and QBS:
  • Lowers Risk
  • Encourages Innovation
  • Ensures cost-effectiveness and competitiveness
  • Results in better projects by avoiding scoping errors and bid tricks, and meeting the needs of multiple stake-holders.
  • Keeps jobs in Florida.  If cost is the sole driver, then all work functions will be shipped overseas without direct Florida licensed professional supervision.


Again, this is a BAD BILL at its core and no amendments will fix it.

In these difficult times, all government agencies are facing a staggering infrastructure deficit.  A system that is driven by the cheapest solution will produce lower quality results.

Moving to a “lowest-bid” for design work will lead to poorer solutions by less qualified “professionals” which will be far more expensive to construct, will increase the public’s risk, both personal and financial, and will indirectly incentivize the outsourcing of this work to “cheaper” countries in order cut-costs at all cost.

As a 3rd generation Floridian (raising a fourth generation), a tax-payer, and one who is very familiar with the good, the bad, and the ugly of the construction industry, I ask that you contact your legislator to oppose this penny-wise pound-foolish bill, SB 0276.

The Committee:

Name
District   
Phone
Steven R. Wise (Chair)
(R)
5
(850) 487-5027
Larcenia J. Bullard (Vice Chair)
(D)
39
(850) 487-5127
JD Alexander
(R)
17
(850) 487-5044
Lizbeth Benacquisto
(R)
27
(850) 487-5356
Bill Montford
(D)
6

2011-02-15

Mediafly: Such Promise, Such Pain

Since the invention of the podcast, there has been a need for a cross-platform application to manage, download, and play podcast subscriptions. Enter Mediafly, a program that seems to be a mixture of genius and monkey crap-fight.

Having used Mediafly for a long time now, I often picture a genius of a software analyst drawing a 50,000-Ft. view flowchart showing what a top notch podcast management platform application should do. He then hands this system diagram to a group of "too cool for school" web designers to design their aesthetically pleasing but sometime convoluted web interface. Web design in hand, a group of unsupervised code-monkeys with diplomas from Fred's School of App Programming and Taxidermy is then tasked with writing a daunting application which manages multiple concurrent downloads and session state across multiple platforms with bandwidth throttling and multimedia playback.

What could possibly go wrong?

You guessed it. This is a program which seems to do everything you would want it to reasonably do, but runs into serious QA/QC problems when it comes to basic application-domain things like restarting downloads, allowing multiple subscriptions to the same podcast, and video decoding errors.

But when it DOES work, good God, it's a thing of beauty. Having your audio and video podcasts pre-downloaded via WiFi or 3G without having to "sync" to a PC when you head home from work or vice-versa is a wonderful benefit to those with hour-long commutes like mine.

This is an App that I love to hate. It shows such promise that every time I get frustrated with it and start to think about uninstalling it, I keep going back to the Android App Store to check to see if a magical update will be there waiting to take away the pain and fulfill that promise.

2011-01-20

Passing the Buck!

Now, I've been doing IT a while, but take a look at the sorry email alert I just received from our ERP vendor about their upcoming new version due out next month. The names have been changed to protect the guilty.

Attention: System Administrators

In the process of certifying LazyCorp ERP v8.5 with the latest Windows Server 2008 operating system, LazyCorp was notified by Microsoft that a known issue existed with the MDAC components on a Windows XP workstations accessing a Windows Server 2008 database. In the event a XP workstation requested a query (ie. Comprehensive Project Inquiry or viewing of an Attachment) from a database located on a Windows 2008 Server which returned data greater in size than 4MBs, the following NuMega error would be returned.

Microsoft has confirmed the defect and advised LazyCorp that mainstream support for Windows XP expired 4/2009 and no defect fixes or enhancement will be released. Only security related updates will be distributed.

Given that Microsoft no longer supports the XP operating system, LazyCorp will not make any programmatic changes in an attempt to alleviate the issue and recommends using one of the following workarounds to successfully retrieve the +4MB data:

  • Disable HTTP 1.1 in Internet Explorer 7 or 8 on Windows XP. Microsoft recommends this change only be used temporarily, and be set back to the default once the data is retrieved.
  • Upgrade workstation trying to execute the query to Windows 7 or Vista (which a patch has already been released to correct the issue).
My favorite line is:

"Given that Microsoft no longer supports the XP operating system, LazyCorp will not make any programmatic changes in an attempt to alleviate the issue . . ."

Seriously? This Let them eat cake attitude has got to stop! It would be relatively simple and cheap for this vendor, who charges 25% annual maintenance with a 10% annual increase, to fix this or simply put up a notice that the query result would be too big, but instead, they pass the buck and tell their clients to convert all their desktops to Windows 7! Hubris!

I think I'll spend my money on upgrading to Windows 7, and tell them to go to Microsoft to look for the money for their maintenance fee next year. I think we're gonna have to switch vendors if this is going to be their attitude in a recession!