inotify-aware tail replacement

Martin Ebourne lists at ebourne.me.uk
Thu Dec 6 22:02:43 UTC 2007


On Thu, 06 Dec 2007 13:41:24 -0800, Jesse Barnes wrote:
> On Thursday, December 06, 2007 1:38 pm Bill Nottingham wrote:
>> Is there any reason the user should *care* about the implementation of
>> tail -f?
> 
> Well, the manpage says -f is affected by the sleep argument, so it's
> conceivable that there are scripts depending on that behavior...

So an inotify patch could be automatically disabled if -s is given. In 
that case use the old behaviour.

> -w (for watch) may be a good name for a new flag.

I don't think there's much point adding it as a new flag since it 
wouldn't get used, thus nullifying the benefit.

I agree with the others, much better to integrate it into coreutils tail 
than provide another version.

Cheers,

Martin,




More information about the fedora-devel-list mailing list