Many years ago, I got burned by Amazon when ordering an eBook.  Their downloading system failed to download the book I ordered and their help desk was absolutely no help.  I lost about $8 to their technical incompetence, and they lost my business for over a year.

Fast forward to today.  I recently killed my Zune subscription account because of its sucky selection.  Much of the music I wanted to subscribe to, Microsoft wanted me to pay for instead.  Thus the subscription service became valueless and overpriced for me.  I'm back to buying CDs which are, by and large, shipped from Amazon.  No, I still don't order their eBooks.

Today I tried another Amazon downloading service, which provides MP3s.  I was going to order Kraftwerk's Autobahn album on CD, until I saw that Amazon had priced it about $27.  This is far above other albums and definitely too pricey for some tracks of catchy and cute atmospheric music.  Amazon also offered the entire album as an MP3 download for $7.83.  It seemed like a good time to try their MP3 download.

I'm glad to say that the download process went smoothly, despite having to install a special Amazon downloader program.  The album has no hidden DRM.  However, it did download directly to my laptop's default folder for Microsoft Media Player music.  Trouble is, I take my music on the road in the Zune, and all the computers in our house share the music files on our Windows Home Server.  Media Player doesn't use the default folder any more.  The Zune software uses the shared folders on Windows Home Server too.

So the Amazon downloader program doesn't understand where the Media Player actually has its library.  It is no big deal for me to copy the files where I want them, I do think Amazon's 90% solution could disappoint a lot of other people.  They will not like the Amazon Help Service.

Our setup for music, while encouraged by Microsoft itself, has been a bit of a pain.  Media Player can autodetect file changes on the local computer, but it is ignorant of folders on file servers.  As we also have a Media Center computer, I get to do regular rebuilds of the music catalogs on Media Center, our laptops, and my workstation.  This is an operation that most consumers won't be able to do.  Microsoft should fix this situation ASAP.

One last thing.  Audiophiles will rail against the poor quality of MP3s and I totally agree.  The album I downloaded is OK, but the fidelity is typical for MP3.  I may try the Zune Marketplace for my next album (in WMA format), if the price is right and there is no DRM.

-- Walter Lounsbery, 8-31-2008

Two weeks ago, I attended the CodeStock conference in Knoxville.  I've just returned from the DevLink conference in Murfreesboro, Tennessee.  August has been the month of software developer conferences.  While the speakers (some working very hard at both events) are worn out, I've had a great time.  I didn't meet anybody at either event that didn't have a good, productive time.

Both conferences have a lot in common.  Both are run by volunteers and well supported by various sponsors.  Speakers may be sponsored, but are not paid from the event proceeds.  Both had excellent content and enthusiastic people.  There was a nominal cost of admission ($25 for the one day CodeStock event, $50 for two days of DevLink).  And of course both events were in Tennessee.

I think it is worthwhile to contrast these sort of events against other conferences.  They are basically modeled after traditional software development conferences (like VSLive!, held by publication companies) or more professional or academic conferences (like SigGraph or ACM conferences).  These conferences generally have high admission costs, are held in expensive venues, and are conducted on massive scales that make you feel like cattle in the herd.  DevLink, in comparison, was limited to an attendance of about 400 due to limits at its venue this weekend.  The DevLink organizers are actively seeking a larger venue next year.

Another difference is a willingness and ability to try different things.  Both CodeStock and DevLink tried the Open Space method for self-organized sessions (conducted by Alan Stevens at both events).  I hope both conferences keep trying new things.

In all cases, I think, the presentations and the speakers make or break a traditional conference.  So at their core, the regional conferences will always have the potential to deliver enthusiastic speakers.  I've been to other types of conferences where the speakers are less enthusiastic, more formulaic, and very guarded because they risk giving away too much of their consulting secrets even as they get to promote their knowledge at large conferences (call them professional speakers).  As long as regional conference organizers recognize and preserve their base of enthusiasm, they will thrive.

My straight appraisal of speakers at DevLink this year: you were all great!  I think last year's DevLink got some world-class speakers, but there were too many pros.  This year was much better and a consistently high quality.  I'll put it this way: I saw a lot of the speakers attending and driving the Open Space sessions, but I wonder how many would have taken the time to do that last year?

I also think CodeStock was great, but I can't get critical about this first Knoxville software development conference.  Maybe next year.  Well, I can criticize the attendees that drove off and missed a great band performance that evening (by Hanover Fist).  The thirty or so folks that came out for the band had a good time.

Although both events were well attended, it would be nice to see more even more participation.  That goes for the user groups that spawn these sorts of events, too.  More software professionals need to think outside the nine-to-five, work up some enthusiasm for their work, and support these kinds of activities.

-- Walter Lounsbery, 8-24-2008

I have enjoyed the progress and popularity of Web journals or blogs over the last few years.  This technology and form of communication just exploded onto the scene.  The technology seemed to lead adoption, there are still facets to blogs that very few blog authors have even heard of.  It is indisputable that blogs, along with Wiki publishing and social network publishing, are well established and common in the Web now.  I've enjoyed reading and writing blogs.  It's great that so many people have published so much information.

The prolific invention of new kinds of Web publishing, which continues unabated, contributes to the vibrant, nonstop nature of the Web.  You would think that in a mature technology like the Web, everything has been invented and reinvented a million times over.  But sometimes when you look around, you don't see the sort of things you expect.  In my case, I'm looking for some simple solutions for simple needs.  I created the SimpleBooks system for Web publishing for a very simple and direct need.

I'd just like to publish Web pages, mainly HTML with styling, on my Websites.  I'd like to write the material in HTML files using an HTML editor (and there are lots of those).  I'd like to start up an FTP utility and just copy the material to my Website, which should pick up the new material and publish it without coding, reprogramming, or having to deal with a database.  My first effort at this sort of publishing, the SimplestArticles system, did much of the job with only 14 lines of code.  But it published from only one directory, did not supply title text for the articles, and could not handle much customization or organization of the material.

So the next step was to create SimpleBooks.  If you want the details, click through to the writeup.  That writeup is actually hosted using the SimpleBooks system, and it has a link to a sample Visual Studio 2008 project for SimpleBooks.  SimpleBooks is a proof-of-concept (OK, it's a demo for the ideas), so don't expect fantastic coding or visual styling.

 

-- Walter Lounsbery, 8-17-2008

I'm trying to log on to the Microsoft Developer Network subscriptions Website this afternoon, and it appears to be down or inaccessible.  My login is rejected.

Could this be due to the release of Visual Studio Service Pack 1 today?  The effect of over 10 million developers trying to get the latest bits at the same time?  I guess we've found the limits of MSDN.

-- Walter Lounsbery, 8-11-2008

My last two posts have taught me some things about layout.  First, you can be surprised by your tools (i.e. Windows Live Writer).  Second, you can be surprised by your publishing application.  In my case, Windows Live Spaces and BlogEngine.NET.  My home page on Windows Live Spaces provides a rather narrow column for my blog posts, which cut off graphics and tables in my initial post about my aircraft design.  Since I get to pick the layout, maybe I should have appreciated what I was doing and remembered to fit the layout better.  My BlogEngine.NET blog does much better and allows you to see all of the tables and graphics.  I still need to respect its limits, but it is more flexible.

I had anticipated these problems some time ago.  There are situations where you the author want as much of the Web page as you can get.  Anything expressed in graphics, tables, video, or interactive content can drive that need.  I applaud the bloggers that suffer publication in a narrow little window on the page (like I do with Spaces).  One alternative is to give the content what it needs when it needs it.  This is possible by publishing an excerpt or a teaser that links to Web pages devoted to the post.

The questions is, where to put the articles so they can be accessed?  My initial solution was extremely simple: provide a Web page that scans all the plain HTML Web pages in a directory, provides an index and a viewing window for those pages, which are basically posts.  The solution precludes hosting the full content on a blog hosting service, you need your own Website and the ability to use your own software.  Here is an example from the Lounsbery Website, which includes an article on the software to publish simple HTML pages (I call it SimplestArticles).  This solution is a bit too elementary for handling lots of content, so I'm going to finish out my project that addresses scaling and other details.  Then I can publish my aircraft design articles or blog posts much better.

-- Walter Lounsbery, 8-8-2008

As I write this, I'm painfully aware how developers (and I are one) can overthink the "user interface".  When that happens, perhaps it should be called the "hardly usable interface" or HUI for short.  In many cases we can even code things to think for the user, or keep that user from doing something "dangerous."  That is absolutely true for a few "features" of this wonderful tool I am using right now, Windows Live Writer.

In my last blog post, I tried to include some highly relevant and important graphics.  I didn't realize that WLW thoughtfully resizes your graphics according to mysterious settings hidden from operators astute enough to check the menu Tools/Options selection.  It is too bad that the clock was running late so that the crappy graphics produced by this wonderful bit of overthinking actually got published.

So here I make an attempt to publish those graphics in the size and legibility they deserve:

 

-- Walter Lounsbery, 8-7-2008

About 15 years ago, I started my first homebuilt aircraft.  I ordered the plans for the KR-2S, anxious to start construction on a homebuilt aircraft and just a little upset that I could not find a home for my Piper Cherokee.  I had moved from Fort Worth, Texas to Savannah, Georgia, and there was no space to park my plane anywhere near Savannah (at a price I could afford).  I had to leave the Cherokee in Texas and sell it long-distance.  It seemed to me that, by the time the KR-2S was ready for an airport, that detail could be taken care of.

Sad to say, my new job in Savannah had lots of overtime and travel.  Other events kept me from starting that construction project.  Here I am ready to start building something much later than I planned.  In the meantime, I also find myself flying in the relatively new Light Sport Aircraft/Sport Pilot regulatory arena.  Regulations that exclude aircraft such as the KR-2S.  Well, I always wanted to design my own plane.  Given how long it took to get started, I'd better get busy on this.

Designing an airplane can be like herding cats.  You may know what you'd like, but there are all kinds of factors between a rough sketch of a sexy-looking airplane and the decisions behind a materials list, drafted design drawings, and that first flight.  With United States experimental aircraft (amateur built) there are few limits.  You can design helicopters, jet packs, aircraft with props or jet engines or turboprops that carry one person or a dozen.  An airplane can take off of water or a runway.  There are a world of choices.  If you want to build a landplane Light Sport Aircraft, the boundaries are a bit more restrictive:

 

• Maximum gross weight of 1320 pounds (600 kilograms)
• Single reciprocating engine (no turbocharging or boost)
• Maximum of two people onboard
• Unpressurized cabin
• Fixed landing gear
• Maximum altitude 10,000 feet above sea level
• Fixed or ground-adjustable propeller
• Maximum speed of 120 knots calibrated airspeed (kcas)
• Stall speed of 45 kcas or lower, no flaps

The last two items are very significant.  In fact, the top speed governs how much engine is legal and practical.  The stall speed governs how much wing is needed.  Unless the wing design is unusual (compared to most light aircraft) the maximum weight and the stall speed should dictate just about the same wing area for all LSA.  This is a reasonable conclusion that actually doesn't work out in reality.  First, let's go through the arithmetic and select a wing area for the LiSA 2 (named after my wife, Lisa).

Lift of a wing depends on only five factors: wing area, wing airfoil, dynamic pressure, angle of attack, and wing geometry.  Dynamic pressure (q) is directly related to the calibrated airspeed, V_e:

                          

In this equation, the velocity is in knots and the dynamic pressure is in pounds per square foot.  If the velocity is doubled, dynamic pressure increases by a factor of four. 

Of the five factors, lift is linear in wing area or angle of attack.  If we use fairly conventional rectangular or tapered wings, the geometry variations don't change lift with angle of attack much and shouldn't have much impact on maximum lift.  So dynamic pressure is a big deal since it varies as the square of velocity.  In fact, due to the LSA stall speed and maximum gross weight requirement, it is the most significant factor in sizing the wing area for the aircraft.

Stall of a wing occurs at its maximum lift coefficient.  While that number can vary dramatically depending on a variety of high lift treatments, when we work with plain wings of fairly conventional aspect ratio and geometry, without flaps, then the maximum lift coefficient is fairly constrained.  Here are examples of production aircraft, based on published data:

 

Certified Aircraft	Max. Gross Weight, lb.	Wing Area, ft2	Stall Speed, kcas	q, lb/ft2	Stall Lift Coefficient
Beech Musketeer	2400	146	63	13.45	1.22
Piper Cherokee	2400	160	60	12	1.25
Cessna 150	1600	160	47	7.49	1.33
Cessna 172	2400	174	51	8.82	1.56

The first two aircraft have rectangular wings, which are less efficient at maximum lift.  The Cessna 150 has a tapered wing and airfoil that is better at producing maximum lift.  The Cessna 172 is the oddball of the group, since the wing design is very similar to the Cessna 150.  My opinion is that the stall speeds I was able to find (and many sources didn't agree) are wrong.  From an aerodynamic perspective, the two-dimensional maximum lift of the Cessna wing airfoil is only about 1.5, which must be greater than the lift of the actual wing due to tip losses and other three-dimensional effects.

Now, if you believe anything you read (just think, you are reading this too!) the next table is going to really confuse you.  How can the following 10 LSA aircraft be so different in stall performance?  The Challenger is a bit of a ringer, since it stalls at much less than 45 kcas.

 

Light Sport Aircraft	Gross Weight	Wing Area	Lift Coef. at 45 kcas
Challenger II	950	173	0.80
Jabiru J170	1320	103	1.87
Opus Super 2	1168	92.5	1.84
Remos GX	1320	118	1.63
Sport Cruiser	1320	141.6	1.36
Dynamic	1212	111	1.59
Mermaid	1430	134.5	1.55
Paradise P-1	1320	136	1.41
RANS S-19	1320	126.9	1.52
Virus 912	1235	118.4	1.52

Confused or not, clearly the wing area has to be big enough to lift the aircraft at whatever maximum lift coefficient you think the wing will be good for.  In other words, it's got to lift 1320 pounds at 45 kcas, pick a maximum lift coefficient and you've got your wing area.  That's what gives the following chart:

                 

I'll design the LiSA 2 for a wing stall lift coefficient of 1.45, giving a wing area of 133 square feet.  In my next post I'll try to cover the influence of span loading on induced drag, or why I picked a 33 foot wing span.

If you dabble in aerodynamics, you've realized that I've just sized the wing without even picking an airfoil.  The airfoil choice really comes from a number of other more important factors that I will cover later.  For now, I'll include a spoiler for my initial airfoil choice, as shown in the Airfoil Optimizer program by DaVinci Technologies:

               

-- Walter Lounsbery, 8-4-2008