I'm rather ambivalent about dishpointer. It's certainly a really neat
idea, but:
1) Their elevation calculation for the Moteck SG2100 seems to be
based on the wrong crank angle, see the link in my original post, p3,
"The reading on the Dish scale should be: 40° - DECLINATION ANGLE"
So my Calculator for my home location (using a 'forward' tilt of 0.66,
so this amount of disagreement with a site using latitude as the basis
for tilt is to be expected) for this rotor gives a dish elevation of
33.23, whereas dishpointer gives 23.
Similarly (using the same forward tilt) PanSat PM900S dishpointer 39,
correct is 23.23
2) More generally their calculations show some strange rounding
inconsistencies, for example one 45 crank rotor rounded the elevation
to 38, another to 39. I suspect they're not looking after enough
decimal places or something. I store and manipulate values to full
javascript accuracy, and only round to display (in 100ths, probably a
bit overkill for most setting mechanisms, but I'd rather give slightly
too much accuracy than too little).
3) When, for five points widely scattered over the UK mainland, I
compared the angles of their marker lines with markers drawn on UK
Ordnance Survey maps such as this demo one I produced (although on
this particular one you can't enter the actual position, just drag the
map and the marker around) ...
http://www.macfh.co.uk/Test/UKOrdnance_Survey_with_OpenLayers.html
... I found that although dp would always calculate the azimuth to
within 0.1 degree, there seemed to be a systematic error whereby the
azimuth actually drawn was too great by around 0.5 degree. Not a
problem when using fixed dish, maybe not even a problem for a rotor,
especially when viewed as an initial pointer just to get a signal, but
it's still rather curious that it's there at all, and doesn't lend
confidence.
4) Then there's the business of perspective distortion in satellite
maps. Over the same five points as above, when I compared photos
instead of the map with the OS version, that introduced about another
0.5 degree random error.
5) When I experimented with their widget in a web-page, as often as
not it failed to load and left the map located in New York
6) If you look at the print preview, the absolute killer is the
marker line doesn't print, making the tool much less useful - as you
can't look a computer monitor while perched on top of a ladder angling
the dish, you need a good printout. Also, on the actual site, there
is so much advertisement clutter that essential information is less
obvious in the print preview, and the marker has a block background. I
know how to fix the marker and line printing problem in principal, but
I don't have the control over the widget to allow me to do it.
7) The maps take ages to load. Given that I've already got a local
OS hardcopy, it's actually quicker to use that!
I haven't entirely given up on this, but besides the dishpointer
problems, the state of play is:
1) Already just about there loading UK Ordnance Survey into Open
Layers, that would be UK only though.
2) Google Maps and all derivatives, such as dishpointer, have the
markers don't print problem.
3) Trying to combine the best of both worlds by loading Google into
Open Layers, the marker isn't stable, if you move the map East West,
it's fine, if you move the map North South or zoom, it shies away from
the edges of the viewport like it's frightened of something.
Post by Char JacksonCould you plug in information from one of several excellent online
sites such as http://www.dishpointer.com ?