NavList:

I have now finished the comparison of my 1970 volume 3 with the online volume 3, using the same methodology described before, to determine if there were any pages duplicated in this volume and am happy to report that I found no duplicated pages. Nor did I find any other errors. I found 10 cases of the one minute difference discrepancy that I described before. I checked each of the discrepancies I found in both volumes 2 and 3 (there were 5 in volume 2)  and found the same round off error as before. The actual Hc was never more than 3 seconds away from the 30 second dividing line for rounding off. For example, on page 134  LAT 22, DEC 7 contrary name, LHA 69 the Hc in the 1970 edition is 16º 31' while in the online edition it is 16º 30'. The computed Hc is 16º 30' 27.5" just 2.5 seconds below the 30' round off dividing line and is correctly rounded down to 16º 30' in the online edition. For some reason the computer algorithm used in calculating the Hcs in the 1970 edition incorrectly rounded up to 16º 31'. This was the same pattern with all 15 cases that I checked, the online edition had the correct values and the 1970 edition differed by one minute due to this systemic rounding off error.

I am attaching pages 133, 139 152, and 158 from the online edition, two copies of each, one copy downloaded in November 2007 and the second copy downloaded today. By comparing them you will see that 133 is a duplicate of 139 and 158 is a duplicate of 152 and this error has not yet been corrected. Also by comparing both copies of page 133 you will see an example of the missing minus sign type error e.g. LAT 22, DEC 0, LHA 91 in the 2007 copy that has been eliminated in the 2008 copy.

I have also scanned and attached pages 133 and 158 from my 1970 edition that you can print and place in your copy of volume 2 to use instead of the ones you have that are duplicates of 139 and 152.

Gary LaPook


Gary LaPook wrote:

It would appear that editions before 1999 are O.K. but those after that date are suspect as is the online version. A quick check to make on your edition is to compare pages 152 and 158 and also 133 and 139, if these are duplicates you have the bad editions. Also look at page 148, LAT 24, DEC 27 same name, LHA 256 for an example of the lack of the negative sign type of error.
 I have compared my 1970 edition volume 2 with the online version and the only pages that I have found that are incorrect are 133 and 158 and I will check volume 3 tonight. My methodology was to compare the Hc given on the first line of each page for DEC either 7 or 22, as appropriate, in the 1970 edition with the online version which should find any pages that are duplicates or containing many errors. Using the method I found several pages where there were the one minute types of discrepancies that I described before that I do not think shows an actual error in the online (newer editions) tables. These were on pages 47, 77, 134,179, and 235.

gl





Ken Gebhart wrote:

My sincere thanks to all who have responded to my call for help, and especially to Gary La Pook.  My apologies also for not having been attentive to past posts concerning the same subject, nor to having searched the archives first.  My company (Celestaire) will have to re-print this book in about 3 months, and will not print a known erroneous edition.  I will let you know how we have solved this problem as time goes on.

Ken

On Nov 13, 2008, at 2:06 PM, Gary LaPook wrote:

To keep this in perspective, a 2.1 second error is only 213 feet!

glapook@PACBELL.NET wrote:

 I found some additional discrepancies between my 1970 edition of H.O
249 and the online version. For example, on page 151 of volume 2, the
first entry (LAT 25, DEC 0, LHA 70) shows 18º 04' in the 170 edition
and 18º 03'  in the online version, a difference of one minute of arc.
There are additional one minute discrepancies on this page and on
other pages. I was going to point out this additional discrepancy to
the government office but decided to check these computations myself.
Using the standard sin cosine formula I computed the Hc for this
example and calculated an Hc of 18º 03' 27.9" which should be rounded
to 18º 03' showing that the incorrect value is in the 1970 edition. I
checked some of the other discrepancies and got the same result, the
online version has the correct values and the 1970 edition contains
the errors. The "errors" are extremely small, my example results from
a 2.1" error in the 1970 computation resulting in rounding off in the
wrong direction to the wrong whole minute which would result in a half
mile error in the LOP. But, if  rounded in the other direction you
would still have a half mile error in the LOP, only in the opposite
direction so this is nothing to get real upset about since you are
expecting only a one NM accuracy in LOPs when using H.O. 249. Since
the math is straight forward I wonder why the original computations
were off by even this small amount.

gl

On Nov 13, 5:14 am, glap...@PACBELL.NET wrote:
  

I sent this email today to the government office that computed these
tables:

Dear Mr. Doherty,

Last year I called your attention to some errors I had found in the on-
line version of  H.O.  249 for Hc's near zero and I checked recently
and these appear now to have been corrected. However, when I was
reviewing the updated version I found some additional errors.
In volume 2 page 158 is completely wrong and appears to be a duplicate
of page 152 and page 133 is also completely wrong and appears to be a
duplicate of page 139! I suspect that there are many other similar
errors hiding in this volume that my cursory review didn't reveal.

Gary LaPook

On Nov 13, 3:34 am, glap...@PACBELL.NET wrote:

    

Further comparison shows that page 133 is a duplicate of page 139!
      

gl
      

On Nov 13, 1:56 am, glap..@PACBELL.NET wrote:
      

Ken,
        

My copy, 1970 edition, doesn't have the error you pointed out. I then
looked at all pages from 122 through 161 looking at the first entry on
each of these pages and found additional differences between my print
edition and the online version. Page 155 has a one minute different Hc
as  does page 151 for the first entry but I didn't check other values.
page 133 is completely different, all tabulations are different ! I
computed several of the entries on page 133 and found that the online
version has the erroneous values while my 1970 edition has the correct
values.
        

gl
        

On Nov 12, 4:08 pm, Ken Gebhart <GEBH...@CELESTAIRE.COM> wrote:
        

To all List Members,
          

Need help to track down errors!
          

I have been made aware of  many errors on page 158 of Pub 249 Vol 2
Apparently all of the Hc values on the page are wrong!  For example,
for dec 0, LHA 0, Lat 26 (page 158), Hc is shown as 65 deg. When it
should be 64 deg by simple inspection.  All the other numbers are
also off by varying amounts as can be found by computation.  There
are some other errors too, of a different kind on other pages.  For
example, page 160.  Look at the column for dec 15 deg, and go down to
LHA 98.  By simple inspection you can see that there should be a
minus sign before Hc 00-25, but it is not there.  I believe this is
typical of many Hcs  near zero.
          

These errors are present in the US Government Edition, privately
published editions, and the PDF on the NGA website.  I have spoken to
the UK Almanac Office, and they report that the errors do not exist
in their equivalent to 249 V2, or their Rapid Sight Reduction Tables,
and can find no trace or knowledge of them.
          

I am asking anyone who has an older copy of this volume, to check it
for these errors and report the publication date.  Maybe we can track
down when they first appeared.  By the way, a friend did phone the
NGA to report this, and was told that “they would look into it”.
          

Thanks, Ken Gebhart
          

--~--~---------~--~----~------------~-------~--~----~
Navigation List archive: www.fer3.com/arc
To post, email NavList@fer3.com
To , email NavList-@fer3.com
-~----------~----~----~----~------~----~------~--~---