19 January 2011

Astronomical Sketching

A poll has been running on the Knightware website for several months that poses this question: Do you sketch observations?

The response looks like this:
Never: 42%
Seldom: 23%
Sometimes: 28%
Frequently: 4%
While the low number for frequent sketchers isn't surprising, the large number that sketch at all (55%) is.

Sketching seems like a dying art. Before the advent of photography in the 1830s, sketching was the best way to convey one's impression from the eyepiece to other people. Today's digital imaging may have displaced more sketching than any other factor.

Be that as it may, sketching is still an important skill for the astronomical observer for these reasons:

1. Sketching forces the observer to look for more detail so that it can be included in the sketch. With limited experience, this is my most compelling reason to pick up the pencil from time to time.

2. Sketches endure. Printed photographs fade over time, though archival storage helps. Black and white prints are less susceptible to fading than color prints or slides. Digital photographs may have storage problems too, but there isn't enough historical evidence to know for sure. Historians are concerned about the durability of digital photos stored on magnetic media because they definitely have a shelf life. CD-R storage is possibly the least durable medium, and many of us use them. Another concern is that digital images require software to be decoded and displayed. We all know that software and file formats change over time.

If you'd like to know more about astronomical sketches, why not check out Carol Lakomiak's monthly column in Sky At Night magazine? Each month she chooses an object and describes the materials and techniques required to produce a nice sketch.

Addendum - I received this note from Carol Lakomiak following the original post ...
  1. You don't need to be an artist to make eyepiece sketches, just learn and practice a few simple techniques. I'm self-taught and have never had any art lessons. My techniques were learned through trial-and-error... if i can do this, anyone can.
  2. Eyepiece sketches are your observations, not works of art. There's no need to reproduce them in a photo program to make them pretty.. that's what we call 'astro art'. Just concentrate on recording your observations as accurately as possible, and your skill will increase with time.
Thanks Carol, and good luck to all with sketching!

09 January 2011

Deep-Sky Planner 3: The Move to 32-bit Windows

In the mid to late-90s, a lot of 16-bit applications were available to run on the new 32-bit OSes, Windows 95 and NT 4, but not so many 32-bit ones. Deep-Sky Planner 3 moved to 32 bits, taking advantage of new facilities in Windows and making full use of the wider address space. Moving to 32-bit OS was not very difficult, but following the new Windows Logo requirements proved to be a lot of work. The Logo requirements provided the formula for developing a compliant application, thereby assuring users of a consistent 'user experience'. What that really meant was that a user should know what to expect in terms of an application's user interface and its interaction with Windows. To an application developer, that was supposed to mean fewer support questions and happier customers. This tenet still guides user interface and system design in Deep-Sky Planner today.

With a larger address space available the database expanded to include variable stars, double stars and quasars. The number of objects in the database grew to 5 times its previous number to 155,582 objects. Several report sorting and filtering options were added to accommodate the new stellar catalogs, and a 'best time to observe' sort option was added that put objects in the best observing order based on one's time and place. This became a nice feature for planning a Messier Marathon or a short observing session for any particular night. Font and color selections in the application made reports customizable in appearance while retaining their memory-efficient nature. The context sensitive help system moved to WinHelp 4 - a significant improvement that required a lot of work.

This release was developed primarily with Borland C++ 5.0 and the Object Windows Library (OWL). A few utilities were developed in the new C++ Builder 1.0. This brief experience with C++ Builder demonstrated a great increase in productivity for this developer. It was obvious that the way forward for Deep-Sky Planner would include a move to C++ Builder.

The production cycle for Deep-Sky Planner 3 was really quick. The contract was signed in June 1998 and it was released on CD-ROM the following month. Advertisements began in Sky & Telescope in October 1998. Despite a large investment in development tools, the price held steady at $49.95.

01 January 2011

Deep-Sky Planner 2: The Move to Windows

Fortunately Deep-Sky Planner 1 was successful enough to warrant a version 2. In the mid-1990s, DOS was losing market share rapidly to Windows 3.1 and Deep-Sky Planner needed to make the move. I stayed with Borland tools for the move, adopting Borland C++ 4 and the Object Windows Library (OWL) for the user interface. OWL was well designed and was similar to Microsoft's MFC in its role for developers. Other new technologies adopted at the time included WinHelp for the online help system and InstallShield for the setup program. (I have long held the belief that installation programs are almost as difficult to develop and test as many applications. Installshield helped, but not much!)

Learning the Windows 16-bit API was the major learning experience for this release. I was not an early adopter of Windows - in fact, I did most of the development for version 2 on OS/2 running Windows 3.1 in a virtual DOS machine. It seemed more stable to me than developing under Windows 3.1 but it really taxed my PS/2 model 70 computer at the time. By the end of the development cycle, I moved to a Pentium machine running Windows.

Memory and disk space were still critical design constraints for 16-bit Windows. Version 2 required an 80286 CPU or above, 8 MB of memory and a whopping 3.3 MB of hard disk space. My, times have changed!

Version 2 gained planet ephemeris and events calculations that users had requested. It contained the same 31,418 deep-sky objects as version 1 and nearly the same sorting and filtering options, just presented in Windows. The program took on the usual Windows behaviors: context sensitive help via F1, the familiar main application menu structure (File, Options, Window, Help), multiple report window management (cascade, tile, etc), and customizable toolbars. By making a search query and the resulting report take on the role of a 'document', it was reasonable to make functions like File|Save and File|Print do what a user might expect with a report. This interface design took some time to evolve. It occurred to me while taking my daughter for a morning walk in the stroller - she's in college now.

Version 2 went much more smoothly on the production side. I began a long working relationship with Rick Fienberg at Sky Publishing in 1995, and he was a true pleasure to work with. The first evaluation diskettes went to Sky in July of 1995. The contract was signed in November 1996 and the copyrights registered in December. The first advertisement appeared in the January 1997 issue of Sky & Telescope. Other venerable software titles appeared in ads of that issue, including TheSky v4, Redshift v2, Voyager v2 and Guide v5. Deep-Sky Planner 2.0 shipped on 2 diskettes and went for $49.95.

By March of 1997, I created web content for the product. I wanted to be able to distribute product updates through the web, and Sky accommodated that by providing space on the skypub.com site - another first for the author of a product published by Sky.

There were eventually 2 'bug fix' updates to version 2, so the web page idea worked out well. Version 1 never needed any bug fixes, but Windows was a different beast.

Next time I'll move on to version 3 and the promise of 32-bit Windows, more catalog data and more new features.