Re: [gentoo-laptop] laptop-mode-tools and updatedb (or cron jobs in general)

[Devon Miller]

That should have read:
if /usr/bin/on_ac_power ; then
    ...
fi

dcm

On 9/27/06, Devon Miller <devon.c.miller@×××××.com> wrote:
> Thanks for the tip on laptop-mode-tools, I hadn't see that.
>
> You may also want to check out sys-power/powermgmt-base. It adds /usr/bin/on_ac_power which simply exits with 0 if laptop's on AC and with 1 if it's on battery.
>
> This lets you make easy changes to the scripts in /etc/cron.* such as wrapping the body of /etc/cron.daily/slocate in
>
> if /usr/bin/on_c_power ; then
>     ...
> fi
>
> dcm
>
>
>
>  On 9/27/06, Fred Labrosse <ffl@×××××××.uk> wrote:
> >  All,
> >
> > I recently discovered laptop-mode-tools and it does replace advantageously the
> > scripts I had written to respond to acpi events.  And it does much more!
> >
> > One thing it does not do (nor were my scripts) is turn off things like
> > updatedb when running on battery.  I looked into that and thought about the
> > various options and here's the result:
> >
> > - One could just move the slocate script from and in /etc/cron.daily to
> > control it running or not.  The problem with that approach is that the script
> > needs to be saved somewhere and then moved (or linked) when it is needed.
> > Moreover, if people want to run it weekly instead of daily, then the moving
> > needs to be changed.
> >
> > - One could remove the slocate script from /etc/cron.daily (or wherever it is)
> > and have a line in the crontab file.  This has very similar limitations to
> > the method above.
> >
> > - One could have laptop-mode-tools write something like "updatedb=[off|on]" in
> > some state file (let's say in /var/state) and have the slocate script  source
> > that state file and run or not updatedb based on the value of the defined
> > variable.  This would not have the above limitations and could be applied to
> > all the cron jobs.  However, this would need modifying system scripts,
> > therefore introducing more work at every update of these.
> >
> > I do think that the final method is the way to go to solve that problem and
> > would be interested in developing/experimenting that.
> >
> > Has anybody done anything like that?  Is that something that could be
> > integrated in gentoo at some stage?  Do people have
> > suggestions/comments/criticisms, etc.?
> >
> > Cheers,
> >
> > Fred
> > --
> > gentoo-laptop@g.o mailing list
> >
> >
>
>
-- 
gentoo-laptop@g.o mailing list