emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Emacs-diffs] /srv/bzr/emacs/trunk r112045: * doc-view.el Fix bug#13


From: Tassilo Horn
Subject: Re: [Emacs-diffs] /srv/bzr/emacs/trunk r112045: * doc-view.el Fix bug#13887.
Date: Mon, 18 Mar 2013 20:50:34 +0100
User-agent: Gnus/5.130006 (Ma Gnus v0.6) Emacs/24.3.50 (gnu/linux)

Stefan Monnier <address@hidden> writes:

>>>> +            ;; We're not already displaying an image, so this is the
>>>> +            ;; initial window showing the document.
>>>> +            (run-with-timer nil nil
>>>> +                            (lambda ()
>>>> +                      ;; In case a conversion is running, the
>>>> +                      ;; refresh will happen as defined by
>>>> +                      ;; `doc-view-conversion-refresh-interval'.
>>>> +                              (unless doc-view-current-converter-processes
>>>> +                        (with-selected-window (car winprops)
>>>> +                          (doc-view-goto-page 1)))))
>>> I don't understand the scenario in which this doc-view-goto-page would
>>> be needed.
>> This is the case where the document is initially shown in a window.
>
> I must admit I do not know what you mean precisely with the above
> sentence.

I mean, the code above runs only when the doc-view buffer is shown for
the very first time in some window.

>> My change to `doc-view-insert-image' prevents insertion of image data
>> if the window isn't shown,
>
> Do you mean "if the window is not *live*" or "if the *buffer* isn't
> shown"?

If the window associated with the current doc-view overlay isn't live.
See line 1273.

>> and here the `doc-view-goto-page' call will do just that.
>
> But between the moment you checked window-live-p and the moment you run
> the above timer, if the buffer ends up displayed, there must have been
> an operation that ran window-configuration-change-hook.
>
> So IIUC this doc-view-goto-page shouldn't be required because the
> window-configuration-change-hook should have caused an equivalent
> doc-view-goto-page already.

window-configuration-change-hook runs image-mode-reapply-winprops, but
at that point in time, there's no winprops entry for that window.
However, image-mode-reapply-winprops calls image-mode-winprops, and that
runs image-mode-new-window-functions because of the missing winprops
entry, and now we are in the situation I handle with the code in
question.

Bye,
Tassilo



reply via email to

[Prev in Thread] Current Thread [Next in Thread]