census 2000 tiger 108th org2ogr errors

First, I want to thank the ogr2ogr author for writing
a tool to digest the tiger files into postgis tables.

I downloaded all GA counties from census site for
tiger 2000 108th congressional district release
(I believe current official release) and crunched
through most of the ogr2ogr features supported.
The following record-type/files and record #
errors were noted by ogr2ogr:

 RECORD TYPE
     RECORD NUMBER
           COUNTY
 RT2  6194 -021
      2941 -023
      4223 -049
     12540 -139
      1805 -201
      7148 -223
 RT5  1013 -001
 RTS  3509 -001
 RT2    40 -001
 RTI 10106 -001

What should I do to correct these?  I'm assuming
the census tiger data files are dirty, but don't
know that.  If it would be of help I could extract
the relevant records noted above and email them
to a developer.

Assuming the census data files are dirty, should
we create a small set of patch files that can
be run against the directory of *RT? files to clean
them up?  Just a thought.

The features I processed into postgis tables:
        ("CompleteChain", True),
        ("AltName", False),
        ("FeatureIds", False),
        ("ZipCodes", False),
        ("Landmarks", True),
        ("AreaLandmarks", False),
        ("KeyFeatures", False),
        ("Polygon", True),
        ("EntityNames", False),
        ("IDHistory", False),
        ("PolyChainLink", False),
        ("PIP", True)
        ]

featuresNotProcessed = [
        ("ZipPlus4", False)

(note the second parameter controlled whether
a geometry index was created by my crunching script)




Thanks,
Heitzso