
--------------------------------------------------------------------------------
All     2006    News for NTIA/ITS HF Propagation Models (WINDOWS version)
--------------------------------------------------------------------------------

09 Feb  2006   1. All input programs have been changed to set the default
                  FPROB for VOACAP to 1,1,1,0. This makes the default for
                  the sporadic E (Es) to be zero (ignored).
                  That was always VOA's recommendation. This spells it
                  out more clearly.

               2. It is also recommended to use Method 30 for ICEPAC and VOACAP
                  instead of Method 20. Method 30 does a smoothing algorithm
                  in the 7000-10000km range and also selects the highest
                  signal strength beyond 10000km between the short and long
                  path calculation methods.

04 Mar  2006   1. BATCH process for VOACAP and REC533 Point-to-Point models
                  was not working.  This has probably been a problem since
                  the 21 July 2005 version.
                  This has been fixed.

18 Apr  2006   1. A change was made at a request from a user.
                  He was interested in having the calculations done using
                  an absorption calculation that used to be in IONCAP.
                  It was determined in 1993 that this method produced a value
                  that exceeded the lower limit of an absorption curve.
                  Thus, it was modified to not exceed that value.
                  This limitation was incorporated into VOACAP in June 1999,
                  but I forgot to document it in these news files.
                  A sizeable difference can occur a low frequencies (<3 MHz)
                  at night. Very little data was ever collected at these
                  frequencies when the IONCAP model was developed.

                  It is highly recommended that most users use the "Normal"
                  default absorption model. It will produce a result that
                  is more conserative.

                  For those who are "adventuresome" this option is accessed
                  through the GUI interface by modifying the "SYSTEM"
                  parameters. Once changed, it remains changed for all
                  calculations until changed again.

                  This is only accessible using VOACAP, VOAAREA, ICEPAC, ICEAREA.
                  But will also affect the INVERSE and S/I models.
                  It does not affect any REC533 calculations.

                  This is (internally) accomplished by modifying the VERSION file.
                  Normally the software version reads like:
                       Version 06.0418W
                  If you change the absorption to the old IONCAP model,it becomes:
                       Version 06.0418I

                  This information is output on every type of calculation.
                  Thus, you will know what absorption model was used.

                  If you change the version file by yourself, make sure you
                  know what you are doing.

                  Whenever you download a new version, the "Normal"
                  absorption model will again be implemented.

                  More information about this may come at a later time.
                  Be aware that not everyone involved in the software was
                  in agreement with the change.

20 July 2006   1. This release is the first version using the Salford F95 compiler.
                  Previous versions were with the Salford F77 compiler.
                  If you have problems, please report them to me at:
                     gregory.hand@gmail.com

               2. Fixed a software error in the Area coverage "Overlay Coverages"
                  option. There may not have been a problem with the F77 compiler,
                  but the source code error caused problems with the F95 compiler.

               3. A new "Combine" option "JOINT REL" allows combining
                  the reliability of coverages. Make sure your coverages are
                  "independent" as using this method requires that. Thus,
                  do not combine coverages of different hours or months.
                  This was an option requested and funded by a user.

30 July 2006   1. A user running the software on a Linux system with Wine
                  reported a problem in VOACAP area coverage calculations.
                  This was traced to floating point underflows that Windows
                  systems let pass. But Wine caused execution failure.
                  I was able to track the problem to the evaluation of the
                  geographic variation of the long term prediction maps in
                  the subroutine VERSY. This was eliminated with some well
                  placed double precision calculations. Similarly, REC533
                  had problems in subroutine EVLGOG.  I have been unable to 
                  solve the problem in ICEPAC. Thus, ICEPAC may not run under
                  Linux with Wine.

               2. Jari requested an option in the Point-to-Point interface
                  to allow the Tx and Rx coordinates to be swapped.  Thus,
                  you will notice a new button to perform this task.
                  If you have a directional antenna, it will NOT swap the
                  mainbeam azimuth.

18 Nov  2006   1. A minor change was made to plotting Distance and Time
                  options in the Point-to-Point models. Previously, if you
                  plotted a variable that did not change, (e.g TANGLE for
                  and isotrope antenna), a division by zero could have occurred.
                  This has been fixed.

               2. It seems that sometime after the 05.0119W version I made a
                  change that lessened the loss when the frequency was well
                  above the MUF (and MUFday < .0001). This would sometimes
                  give rise to a reasonable value of REL, when no communication
                  would be expected. This was especially true for HAMs that
                  tend to use very low required SNR values. I corrected this
                  and added more loss when MUFday < .0001. This will only
                  affect circuits that are well above the MUF. MUFday is the fraction
                  of the days of the month that the operating frequency will be 
                  below the mode MUF.
                  This change was performed in both VOACAP and ICEPAC.

               3. Look for some changes in REC533 in the future that have been
                  requested by the ITU meetings held in Geneva in October.

