[aprssig] HAAT calculators

Sean Jewett sean at rimboy.com
Tue Dec 14 11:10:07 EST 2004

On Tue, 14 Dec 2004 stanzepa at earthlink.net wrote:

> >From the July 2, 2004 installment of Surfin':
> "Way back in the March 3, 2001, this column described the HAAT (height
> above average terrain) on-line calculator that was available at the FCC
> Web site. Shortly, after that column appeared, this valuable tool was
> taken off-line for maintenance and three years later, it is still
> off-line. I get an e-mail approximately once a month asking about that
> column. So, let me bring you up to date on the HAAT calculator story.

Too bad it was too early to say it was taken down for DHS reasons.  

Linux users can use Splat! to calculate HAAT and make some crazy colored 
LOS maps:


Note that there is a bug somewhere in the program that appears to manifest
itself under RH 7.x (noted below)  It does run fine under several varients
of YellowDog Linux (Linux for PPC/Macintosh) and may run fine on other
varients of x86 hardware.  This program is very resource intensive.  I did
discover that my 200 MHz 604e based Mac performed only 4x slower than my 
2GHz P4.  

Here is one of the maps generated, it's for a friend's repeater:


Here are the wall clock times of Splat! on linux (all Intel systems are 
RH based, PPC is YDL based):

Intel P4 2 GHz, 512 megs of ram

real    2m11.442s
user    2m5.190s
sys     0m0.650s


PPC 604e 200MHz, 32 megs of ram

real    8m47.735s
user    8m33.700s
sys     0m2.200s


PPC 603ev 180MHz, 80 megs of ram

real    12m33.658s
user    12m26.900s
sys     0m3.460s


Intel Dual Pentium 3 733MHz, 512 megs of ram

real    2m17.968s
user    2m17.750s
sys     0m0.180s


Intel Dual Pentium 3 866MHz, 512 megs of ram

real    1m59.045s
user    1m58.160s
sys     0m0.570s


PPC 603ev 240MHz, 144meg of ram

real    13m58.482s
user    10m15.290s
sys     1m43.100s


Overall I'm pretty pleased with the results.  I expected the Dual P3 to
perform good but not better than the P4.  That said, given that it's a
dual processor I'm not surprised that it did.  I'm not sure why the 240MHz
PPC performed worse than the 180, I think I may have a reason but I'll
need to reboot the 240 system to see if my thoughts are right.  I was
surprised that the 200MHz 604 performed only 4x worse than the P4
(realtime wise).  That said, I'm going to get YDL running on a linux
system (most likely one of the iMacs in my office) and see how the G3
performs.  Given that the 604 predates the P4 by what, 5+ years it's no

Of course, that means I need to give it a try on my dual proc PPC system.

Unfortunately the alpha did not want to play.  It gave some sort of
floating point exception the minute it started trying to crank thru the
analysis.  I'm not sure if it's because it's a software issues (ie, in
SPLAT!), a library issue, or some other issue.  Again, I may give it a
kick in the head to see what happens.

I think I have discovered a bug whenever SPLAT! is compiled on RH7.3
systems.  It will not plot county and cities onto one map... it's either
one or the other.  The PPC systems plotted both without issue.  Not sure
where the deviation is coming from.

Note, this is totally unscientific.  Just a rough look at how well or
poorly this software performs on various hardware.


The punk rock will get you if the government don't get you first.
	--Old 97's
KG4NRC  http://www.rimboy.com  Your source for the crap you know you need.

More information about the aprssig mailing list