url-insert-file-contents

url-insert-file-contents

[Lars Magne Ingebrigtsen]

It seems that some people are experiencing problems with nnweb in
relation to `url-insert-file-contents' in some versions of url.  nnweb
assumes that it returns nil when the file is retrieved successfully
and non-nil when not.  Did this change in recent versions of url?

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@ifi.uio.no * Lars Magne Ingebrigtsen

Re: url-insert-file-contents

[William M. Perry]

Lars Magne Ingebrigtsen <larsi@ifi.uio.no> writes:

> It seems that some people are experiencing problems with nnweb in
> relation to `url-insert-file-contents' in some versions of url.  nnweb
> assumes that it returns nil when the file is retrieved successfully
> and non-nil when not.  Did this change in recent versions of url?

  url-insert-file-contents has not changed in ages.  It has always
returned:

	(kill-buffer url-working-buffer)

  docs for kill-buffer say: 
Value is t if the buffer is actually killed, nil if user says no.

-Bill P.

Re: url-insert-file-contents

[Jan Vroonhof]

wmperry@aventail.com (William M. Perry) writes:

>   url-insert-file-contents has not changed in ages.  It has always
> returned:
> 
> 	(kill-buffer url-working-buffer)
> 
>   docs for kill-buffer say: 
> Value is t if the buffer is actually killed, nil if user says no.

This is probably related to the gnuserv.el problem that aliasses
kill-buffer to server-kill-buffer which does not have the above
behaviour. It would not be that difficult to fix, but it probably
needs a rethink anyway.  My be all the functionality should be in
kill-buffer-hook.

Jan

Re: url-insert-file-contents

[Tony Bennett]

William>   url-insert-file-contents has not changed in ages.  It has always
William> returned:

William> 	(kill-buffer url-working-buffer)

William>   docs for kill-buffer say: 
William> Value is t if the buffer is actually killed, nil if user says no.

My XEmacs 19.15 kill-buffer returns nil on success.  But I just noticed
it has following note in docstring, so gnuserv is overriding at least
some its behavior:

    NOTE: This function has been enhanced to allow for remote editing
    in the following way:

            ...

--tony

Re: url-insert-file-contents

[William M. Perry]

Tony Bennett <tbennett@chapelhill.hp.com> writes:

> William>   url-insert-file-contents has not changed in ages.  It has always
> William> returned:
> 
> William> 	(kill-buffer url-working-buffer)
> 
> William>   docs for kill-buffer say: 
> William> Value is t if the buffer is actually killed, nil if user says no.
> 
> My XEmacs 19.15 kill-buffer returns nil on success.  But I just noticed
> it has following note in docstring, so gnuserv is overriding at least
> some its behavior:
> 
>     NOTE: This function has been enhanced to allow for remote editing
>     in the following way:
> 
>             ...

  The correc thing to do is not rely on the return value of
url-insert-file-contents.  I did not intend for there to be any semantics
for its return value at all.

-Bill P.

Re: url-insert-file-contents

[Lars Magne Ingebrigtsen]

wmperry@aventail.com (William M. Perry) writes:

>   url-insert-file-contents has not changed in ages.  It has always
> returned:
> 
> 	(kill-buffer url-working-buffer)

Ok; I've now removed all dependencies of the return value from this
function from nnweb.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen