GLIMS Meeting Notes [2/20/02]

[Bruce Raup]

Vince,

Thanks for doing the minutes.

One request:  Can we use names, rather than initials, to identify people?
Initials are not unique within the data center (RB could be Roger Barry or
Rob Bauer, for example), and typos completely obscure the intended
meaning.  For example,

> Nancy A. imported the shape files into ArcView.
> We learned that the shape files contain
> .DBF files. These are DBASE files. DBASE was/is
> an RDBMS that operated in DOS. The table attributes
> in the DBF files have a name length limit of 10
> characters. BM said the file names generated when
> attempting to import into the GIS utility contained
> ~ (maybe an ISO standard constraint; 8.3 is 10 characters).

Who is BM?  Is that a typo?  I take it it should be BR (me).

> We discussed issues with consistency with names used to
> describe the attributes in the GLIMS data base and names
> used in shapefiles the .DBF file restricts attribute names
> to 10 characters.
>
> Recommendation: We do not want to change our attribute
> names.
> Is there an alternative utility that does not rely on .DBF
> files nor have the limitation for attribute name length.
> BR will evaluate if an alternative utility is available.

Nancy, who uses state-of-the-art ESRI products, wrote:

Bruce,
Looks like they are now limited to 10 characters... I tried
creating a coverage with names of 16 characters (size limit
for coverages) and it truncated them to 10.
--Nancy

So unless we get away from shapefiles altogether (not a good idea, IMO),
we're stuck with short attribute names within the shapefiles.

I see two possible solutions:  1) list and publish long and short forms of
the attributes, and use the short forms only in the .dbf files; 2) use
only shortened forms for attributes.

I personally prefer option 1.

> L.Copeland and Flagstaff folks are writing c++
> code for for analysis package based on OGC/open
> source guidelines. Using Qt graphics librariy
> (KDE is based on this library).

Minor correction:  should be Copland.  However, the software is being
written only by the Flagstaff folks.

> BR out 2/28 - 3/6 for visit to Alaska and
> participate (as spectator?) in the Ididerod.

Definitely as a spectator!  We did get rides on a dogsled that wasn't in
the Iditarod, though.

Bruce

-- 
Bruce Raup
National Snow and Ice Data Center                     Phone:  303-492-8814
University of Colorado, 449 UCB                       Fax:    303-492-2468
Boulder, CO  80309-0449                                    braup at nsidc.org