Hi Martin, I am looking at Debian pciutils bug #411314,
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=411314 and it appears that there are a couple problems causing compressed pci.ids files to not work as expected. 1.) update-pciids is not saving the files with the right extension. Here is a patch that fixes the naming. --- update-pciids.sh~ 2007-02-23 03:33:32.000000000 -0800 +++ update-pciids.sh 2007-02-23 03:37:21.000000000 -0800 @@ -7,16 +7,21 @@ GREP=grep if [ -n "$PCI_COMPRESSED_IDS" ] ; then + # Compression explicitly turned off DECOMP="cat" - SRC="$SRC.gz" - GREP=zgrep -elif which bzip2 >/dev/null ; then - DECOMP="bzip2 -d" - SRC="$SRC.bz2" +# For future bzip2 support +#elif which bzip2 >/dev/null ; then +# DECOMP="bzip2 -d" +# SRC="$SRC.bz2" +# DEST="$DEST.bz2" +# GREP=bzgrep elif which gzip >/dev/null ; then DECOMP="gzip -d" SRC="$SRC.gz" + DEST="$DEST.gz" + GREP=zgrep else + # Compression on, but we couldn't find it DECOMP="cat" fi 2.) The file at "http://pciids.sourceforge.net/v2.2/pci.ids.gz"; is not gzipped. Likewise "http://pciids.sourceforge.net/v2.2/pci.ids.bz2"; is not bzip2'd (although that's not an issue yet since we don't have bz2 support). Could this be curl/wget or the sf.net web server automatically decompressing? So when update-pciids grabs them they don't arrive on the system in the right format. Due to the naming bug above this wasn't obvious, so when fixing this it will mean existing 2.2.4 update-pciid's will be getting "pci.ids" files that are compressed. This seems to work fine though. 3.) I think the update-pciids target in the top level Makefile should have a dependency on lib/config.mk in order to ensure that PCI_COMPRESSED_IDS is set properly (like if you used 'make ZLIB=no update-pciids' on a clean tree)? Comments? Thanks, -- Matt Taggart [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
