>From the fingers of Marc Slemko flowed the following:
>
>Why was this added? Someone said something about their client reloading
>the page each time they scrolled(!?!?)
No, each time he hit "back" or resized the window.
>Using Navigator3, it was fine for me before and when I wanted an update I
>just reloaded it now. Now, when I have the index up, it keeps reloading
>it because it expires. eg. if I load the index, look at a PR, go away for
>30 minutes, go back to the index, it reloads it. I find that quite
>annoying.
Dirk's browser was reloading it any time he changed the geometry of
his screen or went from a PR display "back" to the index. Since
it's a CGI, it's almost certainly being marked as "no-cache", so this
behaviour on the part of his browser would be marginally defensible.
I doubt that it's as annoying to you as having to reload the entire
bloody thing over in Italy under slightly more frequent circumstances.
Whatever. I haven't found any way to determine a last-modified date
for the whole GNATS database. If there is one, it might make things
a wee bit better to use a Last-Modified header field instead. On
the other hand, maybe not - because then the pages you were looking
at might be marked modified even if it was an unrelated piece of the
db that got modified.
I'm open to suggestions, but I strongly suspect that your pain now
is considerably less than Dirk's was before the change. I also
think it's appropriate for people with high bandwidth (e.g., in the
U.S) to make some minor sacrifices for those who aren't as close to
the source.
#ken P-)}
|