[Babase] PsionLoad Important Problem

Karl O. Pinc kop at meme.com
Sun Jun 17 20:13:55 EDT 2007


On 06/17/2007 01:40:47 PM, Lacey Maryott wrote:
> Hey Karl,
>   Leah and I are working on uploading Psion data to babase.  We  
> encountered an error that reads /Unknown ADL type (field 4): O/:  
> Line: 227.
> 
> The sample it is referring to is
> <HDR>,0609121203LOM,PTSAMPLR_JUL03,SETUP_JUL03,4,12 Sep  
> 2006,WEA,JKW,LOM,FEM
> <PNT>,0609121203LOM,12:05:07,W1NN,XXX,XXX,XXX,
> <PNT>,0609121203LOM,12:06:10,W1NN,LUD,XXX,XXX,
> <PNT>,0609121203LOM,12:06:59,OS,,,,
> <PNT>,0609121203LOM,12:07:59,W1NN,XXX,XXX,XXX,
> <PNT>,0609121203LOM,12:08:57,W1NN,XXX,XXX,XXX,
> <PNT>,0609121203LOM,12:10:02,R1NN,XXX,XXX,XXX,
> <ADL>,0609121203LOM,12:10:18,A,EDU,OS,LOM
> *<ADL>,0609121203LOM,12:10:51,O,LOM PRESENT EDU*
> <PNT>,0609121203LOM,12:11:04,W1NN,EDU,EDU,XXX,
> <PNT>,0609121203LOM,12:11:59,W1NN,XXX,XXX,XXX,
> <PNT>,0609121203LOM,12:13:00,W1NN,KUT,XXX,XXX,
> <PNT>,0609121203LOM,12:13:57,W1NN,XXX,XXX,XXX,
> 
> Adlib type O should be a valid code, and there are entries in babase,  
> in the allmiscs table, with this O code, which is used for things  
> like embrace, mount etc.  Therefore, we aren't sure why it is not  
> clear what the real error might be and we were wondering if it might  
> be a coding issue?

You are on the right track.  There's no provision for it in the
program.  I don't believe I found any mention of 'O' in
the psion documentation I reviewed to produce:

https://papio.biology.duke.edu/babasewiki/PsionFormat

The foxpro uppoints program just checked to see if it
was a grooming and if not then threw the data
into adlibs (which is now ALLMISCS.)  The foxpro
program did not validate any of the Psion data
and we got a lot of conversion errors because
of it.  Babase 2.0 does not validate ALLMISCS.Txt,
so when we ran the conversion we got no conversion
errors.

The question then is, what does 'O' mean and how is it
different from 'U'?  Note that the best documentation
I know of is the docs on the Pison format on the
wiki (above).  Should 'O' be an error?  I'd guess not
and have _vague_ recollection of something like: O is the
same as U except that O is free-form and U is structured
-- but nothing checks the structure of U so it may as
well not be.  But I've also vague recollections of
there not being an 'O' any longer.  Maybe they should
all be 'U'?

So, we don't have the documentation and need to know more
about what the data should be.  This is a job for...
(pregnant pause)
Scientists!

Meanwhile back at the ranch, we should check to see
if there's any other strange codes hiding out on
the ALLMISCS table.

When we find the answer I should probably do
more validation of ALLMISCS.Txt.  Sound right?


Karl <kop at meme.com>
Free Software:  "You don't pay back, you pay forward."
                  -- Robert A. Heinlein



More information about the Babase mailing list