I've identified the problem. It does eventually produce some output, but it takes ages.
bash while read loops really suck when it comes to performance, so when it has to read every line (egrep for '.') in 40K+ line file, it takes far too long. in the long run, i'll be re-writing dlocate in perl. in the short term, i'll try knocking up a quick awk script to replace the current ugly while/read loop. [ a little later ] hmmm. that was easier than i thought. dpkg: # time dpkg -l \* > /dev/null real 0m4.179s user 0m0.847s sys 0m0.140s original dlocate using a bash while/read loop: # time dlocate -l . > /dev/null real 1m11.527s user 0m20.705s sys 0m53.480s fixed dlocate using awk: # time ./dlocate -l . > /dev/null real 0m1.342s user 0m1.210s sys 0m0.217s problem solved. and dlocate is faster than dpkg again. will upload updated dlocate package soon. craig -- craig sanders <[EMAIL PROTECTED]> BOFH excuse #290: The CPU has shifted, and become decentralized. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
