On Thu, 23 Jul 1998, Carol Geary wrote:

> Richard Hartman wrote:
> > 
> > >   To save
> > >disk space we wrap the file #s (at around 50).

> No, you can't control the users cache. If you are trying to save the client's
Actually, this isn't entirely correct.
One way to set up a "non-cacheable" image is to output the image from a
CGI program with a header included of "Cache-Control: no-cache".
This will cause the image to not be cached by the browser (HTTP/1.1
specification.. if you want compatibility with browsers only
understanding HTTP/1.0, use the "Pragma: no-cache" header.. or both :)

Be warned, however.  CGI programming is not for casual programmers.  There
are so many ways to open up your system to attack through CGI that I would
suggest finding a real professional to set it up for you.  Just for
funzies, tho, here is a VERY skeletal program that will do what you need.
I know this doesn't belong here, but it was posted here, so I will
respond.. this time.
----------------------------------------------------------------------------
#!/usr/local/bin/perl
$basepath = "/www/htdocs/newlong/img/";
$file = $ENV{"QUERY_STRING"};
$file =~ s/[^\w.-_]//g; # strip out all non-filename characters
if (open(IMAGE,"< $basepath/$file")) {
   print "Caching-Control: no-cache\n";
   print "Pragma: no-cache\n";
   if ($file=~/.gif$/i) {
      print "Content-type: image/gif\n\n";
   } else {
      # tag any file not ending in ".gif" as a JPEG
      print "Content-type: image/jpeg\n\n";
   }
   while (<IMAGE>) {
     print;
   }
} else {
   print "Content-type: text/plain\n\nError opening $basepath/$file\n";
}
exit(0);
----------------------------------------------------------------------------

To use that program, just call it as:
<IMG SRC="http://www.blah.blah/whatever.cgi?image.gif">
Where image.gif is replaced with the name of the image you are wanting to
display.

If you don't want to worry about CGI, you might try doing a similar thing
in the HTML.. this may or may not do what you want, and probably won't
work in all browsers, but give this a try in the HEAD portion of your
page:

<META HTTP-EQUIV="Cache-Control" CONTENT="no-cache">
<META HTTP-EQUIV="Pragma" CONTENT="no-cache">

This should, in theory, cause the page to be reloaded every time the
browser goes to it (using a link, not the back/forward navigation)
It may or may not cascade to the objects (images) on the page as well..
you will just have to try and see :)

> itself. If the client has set the cache to a large capacity, why should you
> worry about it?
I believe they are changing images.. IE autoupdate or something.. that's
what the post sounded like, anyway.

> Are you trying to save the server's disk space? How about a different
> approach? After you generate 50 files, delete them on the server, and keep
> going with the numbers to generate new files for the client. 
Personally, I am still at a loss as to how the client knows which of the
50 images is the current one.. but it's not my project, so I don't really
care :)

Pat

HWG: hwg-graphics mailing list archives, maintained by Webmasters @ IWA