You beat me to it. Thanks :-) /J
2008/1/13, Petter Reinholdtsen <[EMAIL PROTECTED]>:
> [Johan Walles]
> > If you define at least one way of including a file by mistake I'll try
> > to come up with something.
>
> I had a closer look at the results, and found what I thought ws false
> positives. I
[Johan Walles]
> If you define at least one way of including a file by mistake I'll try
> to come up with something.
I had a closer look at the results, and found what I thought ws false
positives. I used this approach:
% cut -d" " -f3- result-nolib.txt > nolib
% cut -d" " -f3- result-withli
[Johan Walles]
> If you define at least one way of including a file by mistake I'll
> try to come up with something.
I can't of the top of my head. The only cron jobs I know that mess
with atime is the manual page indexer and the shared library updater,
and your patch seem to avoid both those pit
If you define at least one way of including a file by mistake I'll try
to come up with something.
Last time I attempted this I verified that all newly searched
directories had files in them that were accessed long enough ago that
they obviously hadn't been updated by any automated scheme. Would
t
[Johan Walles]
> This patch gives me a 13% improvement (166/1232 packages) in the
> number of discovered packages.
>
> Here are the results from using the above patch on my system with 1232
> packages installed:
>
> newpackages: 166 new non-NOFILES packages
I had a look at the file, and it look
This patch gives me a 13% improvement (166/1232 packages) in the
number of discovered packages.
Here are the results from using the above patch on my system with 1232
packages installed:
newpackages: 166 new non-NOFILES packages
results-nolib.txt, packages-nolib.txt: 531 non-NOFILES packages
res
6 matches
Mail list logo
