<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7651.59">
<TITLE>Ingres driver for gdal</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<BR>
<P><FONT SIZE=2>Hi -<BR>
<BR>
I am another recent addition to the GDAL community. I am currently using the Ingres driver for the latest svn trunk version. I had originally downloaded the driver back in May, and had it working, but the latest version wasn't working for me; specifically ogr2ogr was seg faulting.<BR>
<BR>
After some digging, I found out that the command line syntax had changed between May and now. Previously, this syntax worked:<BR>
<BR>
ogr2ogr -overwrite -append -skipfailures -f ingres INGRES:mydbname road.shp<BR>
<BR>
Now however, the syntax is (as documented on the GDAL website):<BR>
<BR>
ogr2ogr -overwrite -append -skipfailues -f Ingres "@driver=ingres,dbname=mydbname,user=mydbuser,password=mypwd" road.shp<BR>
<BR>
I was wondering if there's a reason to have the syntax so different from some of the other formats, such as Postgres? (with the @ sign at the beginning). Also, out of curiousity, is the new driver parameter used to separate the vector format (specified by "-f Ingres") from the driver used? <BR>
<BR>
Finally, I fixed the seg fault by initializing the data source pointer as such:<BR>
<BR>
from ogr/ogrsf_frmts/ingres/ogringresdriver.cpp<BR>
<BR>
OGRDataSource *OGRIngresDriver::CreateDataSource( const char * pszName,<BR>
char ** /* papszOptions */ )<BR>
<BR>
{<BR>
> OGRIngresDataSource *poDS = NULL;<BR>
<BR>
.....<BR>
<BR>
Thanks,<BR>
Denise</FONT>
</P>
</BODY>
</HTML>