[vsnet-alert 3986] MuniDOS error exac description

[Rudolf Novak <rudolfn@physics.muni.cz>]

Dear colleagues,

according to discussion about problems with MuniDOS. As reported by Lukas Kral and Dieter Husar, there were some problems with MuniDOS output file, especially in the first column containing a Julian Date.

Where was the problem? Unix/Linux users should be calm - the fault was in DOS version, exactly in file MUNIAL.EXE which makes selection of variable and comparisons and read all files to make final light curve.
I attach part of o Turbo Pascal code where as you'll clearly see was very embarrassing  error:

...
se:=se/60+expo/2;mi:=(mi+se)/60;ho:=(ho+mi)/24;day:=day+ho;
...
where 
            se ...second
        expo... exposure time
            mi...minute
	ho...hour
           day...day

       

as you may see "se" is divided by 60 to be converted to minutes and the problem is that EXPO is NOT DIVIDED by 60 and 2 but only 2. So the result was shifted by factor strongly dependant on exposure time. 

in the version available right now on 
	http://ian.cz/munipack
is the bug fixed and also Y2K problem solution so I strongly recomend you to dowlnoad. 

AS IN THE BUG VERSION - AUTORS DON'T GUARANTEE FOR ANY DATA LOSS AND PROBLEMS.

Personaly I'd like to excuse to all who had a problem. It is very pity and depressing to me that so stupid bug was in the package but I hope that you'll understound and won't beat me :)

I'd like to ask all of you to check this. Time is the only think which could bring some problems. Photometry core was written very carefully and is almost identical with Daophot II. For more information and news go to a page please.

With best regards

Rudolf Novak
author of a bug and DOS version.

P.S. You can also contact directly me via email. (codel@ian.cz)

VSNET Home Page

Return to Daisaku Nogami

---

vsnet-adm@kusastro.kyoto-u.ac.jp