Re: Expires, Last-Modified, Pragma: no-cache etc.

> On Tue, 15 Aug 1995, Lou Montulli wrote:
> 
> > I recently changed netscape to interpret "Pragma: no-cache" and not
> > cache the object.  This is slightly different than a "Expires" header
> > because the object will not even be cached for history navigation.
> 
Dave Morris writes:
> I believe this is the wrong design.  The user *MUST* be able to always
> flip the page backward and see the same material just viewed.  THis
> is a critical user interface usability issue. Within the same session/
> instance of UAgent execution. I believe it would be/is wrong to
> silently refresh a document during history navigation when the history
> cache has overflowed. I (and many users I know) expect the history
> to be a record of what I've seen. I would have not objection to a 
> browser which checked currency and via a non-modal message advised that
> the history copy wasn't current ("Current copy not current, RELOAD for
> the latest copy" for example).
I agree, this is the right approach. According to this, I change my statement
on 'Pragma: no-local-cache' from doubtful to useless.
I has experience with client cache only using netscape 1.1N.
Netscape has 3 options on contolling local cache freshness:
a) re-check on every access.
b) re-check once per execution.
c) never.
I preferred the b) variant normally, and c) every time when I worked off-line
and never choosed a).
When I follow a link in a not fully readed document, I want to go back into
the same document, preferably into same position where I left - when I'm done
with the followed link. Now I'm reading web documents this way, and netscape
supperts me in doing this.
I'll be worried, if I can't do the same in the future.
Or somebody in the WG prefers to see a changed document when returns back from
a link to the original, not yet completely read document? I will be suprised if
somebody says yes!

Andrew. (Endre Balint Nagy) <bne@bne.ind.eunet.hu>

Received on Wednesday, 16 August 1995 13:20:10 UTC