Re: Form reuse doesn't update content [PATCH]

[Boruch Baum <boruch_baum@xxxxxxx>]

On 2018-04-05 09:19, Katsumi Yamaoka wrote:
> Thanks.  I realized how the problem is serious, and your fix
> (304 check) seems a good approach.   I've committed your patch
> excluding your (message "Boruch:...") probes and:

Wait a minute. The patch that I sent along with the (304 check) approach
wasn't the working patch; it was the FIRST patch that was the successful
patch.

The (304 check) was an idea that I haven't been able to successfully
test because I couldn't find a server that actually sent a 304 response.
Also, because I haven't actually been able to receive a 304 response, I
don't know whether emacs-w3m will require more code to actually respond
correctly to it.

The patch that actually works is the first patch I sent; that's why I
divided the thread. The reason the second patch commented out the lines
from the first patch was because I wanted to be sure that what was
happening was the result of dealing with the etag header instead of just
denying cache-use when a textarea tag exists. That's also the reason why
the second patch included all those test messages; it was just a testing
patch. the email that included that patch said I was asking for help in
getting it to work.

It may be that the second patch doesn't *hurt*, but it might just be
part of the solution and I don't think it actually does anything yet.
What I mean is that if you want to include the both patches (except for
the parts in the second one that sent debug messages and commented out
prior code) then the addition to the FIXME comment should say that the 304
check isn't completely coded and may not actually work.

Also, the comment about the cdr of the cons cell being an ETAG value
isn't exact: it's a header line to be added to the HTTP GET request that
asks the server not to send the url page if the ETAG value is current.

>
> In [emacs-w3m:12947]
> On Tue, 03 Apr 2018 17:00:46 -0400, Boruch Baum wrote:
> [...]
> > --- w3m.el	27 Feb 2018 06:23:35 -0000	1.1706
> > +++ w3m.el	3 Apr 2018 20:26:28 -0000
> [...]
> > -	      (when (eq (car w3m-current-forms) t)
> > -		(setq w3m-current-forms (cdr w3m-current-forms)))
> > +;	      (when (eq (car w3m-current-forms) t)
> > +;		(setq w3m-current-forms (cdr w3m-current-forms)))
>
> This is still necessary even if the way is not so smart.  See
> the function `w3m-fontify-forms' (w3m-form.el) and how the
> function `w3m-goto-url' (w3m.el) pushes the `t' flag.

If that snippet was from the second patch, the only reason those lines
were commented out was to be sure that the 304 check could be tested.


-- 
hkp://keys.gnupg.net
CA45 09B5 5351 7C11 A9D1  7286 0036 9E45 1595 8BC0