Re: Bug references in newsticker's docstrings

[Chong Yidong]

Since no one responded, I went ahead and made the suggested changes.

> net/newsticker.el contains some bits from its pre-CVS past that should
> be updated:
>
>  (defun newsticker--insert-image (img string)
>   "Insert IMG with STRING at point.
> This is a work-around for a strange behavior of Emacs versions before
> 21.3.50.  Images inserted with `insert-image' vanished after doing
> `fill-region'."
>   ;; This should work:
>   ;;(insert-image img string)
>   ;; but it does not. Therefore we do this, which should be equivalent!
>   (let ((pos (point)))
>     ;;(insert string)
>     (insert ":-)")
>     (add-text-properties pos (point) (list 'display img))))
>
>  (defun newsticker--buffer-redraw ()
>   "Sometimes (CVS) Emacs forgets to update the window..."
>   (if (fboundp 'force-window-update)
>       (force-window-update (current-buffer))
>     (redraw-frame (selected-frame)))
>   (run-hooks 'newsticker-buffer-change-hook)
>   (sit-for 0))
>
> At the very least, the docstrings should be changed IMO. But the
> question is: are they still relevant, i.e., do they refer to real
> bugs, or it is now cargo cult and should be fixed?
>
> AFAICS, the comment in `newsticker--insert-image' is not true anymore
> (at least, no image has vanished in my tests).
>
> `newsticker--buffer-redraw' is used mostly after calling
> `add-to-visibility-spec' and `remove-from-visibility-spec', and in two
> cases after adding or removing text properties. Should
> `force-window-update' (or even `redraw-frame') be necessary after
> that? (If so, `force-window-update' can be directly called, as it is
> always going to be defined anyway.)
>
>                    /L/e/k/t/u