[Difx-users] [Non-DoD Source] Re: generate fixed JPLEPH calc9 file for calcif2? {External} {External}

[Jan Wagner]

Hi David,

interesting to hear about the internals, esp. that the record lengths of
DE403 and DE421 are identical. Apparently, data structures are, too. I did
not expect DE403/DE421 files to be interchangeable in terms of structure.

"Can you try the calc9 server as is using DE421?" --> Indeed, that fixed
it! Wow.

Now VLBI track m25020 of 20th January 2025 works under calcif2. As does a
copy of m25020 shifted to year 2045.

The steps were:

cd $DIFXROOT/share/CalcServer/
cp $DIFXROOT/share/difxcalc/DE421_little_Endian .
mv JPLEPH DE403_little_Endian
ln -s DE421_little_Endian JPLEPH
startCalcServer

It's a bit ridiculous that it was actually that simple! :P

Many thanks again!!

Kind regards,
Jan

On Thu, Jan 23, 2025 at 9:33 PM Gordon, David CIV USN NAVOBSY DC (USA) <
david.gordon126.civ at us.navy.mil> wrote:

> Jan,
>
> The record lengths (8144 bytes) appear to be the same for DE403 and DE421.
> I put together a short program using the relevant subroutines from the
> calc9
> version of cpepu.f and ran it with the DE421 ephemeris. It ran with no
> problems.
> DE421 seems to be good up to Jan. 2, 2050. Can you try the calc9 server as
> is
> using DE421?
>
> If that doesn't work I think I can make a new DE403 file with data from
> 1900
> to 2100. I think JPL may have made some changes in DE403 since I originally
> made a binary version around 26+ years ago. I recall ascii files being in
> chunks
> of 25 years. Now they are 100 years. And the record length may have been
> 7144 bytes. John Benson may have made the file you have sometime later
> with the larger record lengths. Or whatever. Let me know how things go.
>
> David
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listmgr.nrao.edu/pipermail/difx-users/attachments/20250128/6a6b9245/attachment.html>