Automation update to search.html (probably boring, committed).

[Hans-Peter Nilsson]

On Mon, 9 Jul 2001, Joseph S. Myers wrote:
> On Sun, 8 Jul 2001, Hans-Peter Nilsson wrote:
>
> > If that's seen as necessary, then that'll be a separate change.
> > I've already added the scripts in question to the "global"
> > (non-gcc-specific) script repository, and the update script is
> > planned to run as the same user (same crontab) that updates the
> > ht://Dig database.
> >
> > My thinking was that this is more site-like infrastructure than
> > gcc-specific infrastructure; you don't see the scripts doing the
> > (other) newmonthsgohere updates for the mailing list indexes in
> > maintainer-scripts.
>
> As long as this is all (including the crontab) in a publicly readable CVS
> repository - and I think the details of such global repositories that
> affect GCC should be documented somewhere in the GCC web pages.

Perhaps.  Right now it isn't (at least the crontab isn't), just
as for other similar tasks running on the system.  I suggest we
let overseers@gcc.gnu.org chime in (CC:ed) to get the bigger
issue out rather than focusing on just this one.  I don't know
about moving all gcc-related cron things to gccadmin, but
perhaps this one can be moved some time.  (BTW, the
maintainer-script repo doesn't have the magic commit-script
things in place to copy the files to the "live" place on
committing, right?  There's a README that says so.)

For thing event in isolation: I don't have authority to write
stuff for gccadmin, but this was a search-related thing and
within things allowed for me.  I wasn't going for doing the
biggest related change I could think of all at once, and am
still not.  Perhaps later.

brgds, H-P