[emacs-wiki-discuss] Re: planner/muse update speed

[Joseph Kiniry]

Hi Jim,

On 11 Jan, 2006, at 10:58, Jim Ottaway wrote:

> > > > > > Joseph Kiniry <address@hidden> writes:
>
> > Hi Jim,
> > On 10 Jan, 2006, at 20:42, Jim Ottaway wrote:
>
> > > > > > > > Joseph Kiniry <address@hidden> writes:
> > > > Hi Jim,
> > > > I see no change in behavior with this small patch.
> > >
> > > > E.g., if I go to a given day my moving to a calendar day (I link
> > > > calendar to planner) I still see "Highlighting buffer...." messages
> > > > for every muse buffer mentioned in the day page regardless of  
> > > > whether
> > > > or not that page is actually loaded into emacs.   On my current day
> > > > page this takes nearly 30 seconds.  It seems the pages are loaded,
> > > > highlighted, then saved, even though nothing was touched.
> > >
> > > > Joe
> > >
> > > Hmm. That patch does seem to stop highlighting when tasks are  
> > > updated
> > > or changed.
> > >
> > > Is your problem exclusively with calendar movement?
>
> > No.  It also happens anytime plan is called, anytime I move to a day
> > (e.g., with planner-goto-yesterday, etc.), anytime I move a task
> > (e.g., with planner-copy-or-move-task), etc.
>
> > > It seems to me that there is no check for pre-existing day page
> > > buffers in the page-saving mechanism between planner-calendar- 
> > > show and
> > > planner-kill-calendar-files.
>
> > My muse-config.el is available via
> >    http://secure.ucd.ie/~kiniry/tmp/muse-config.el
> > if that'll help.  I have tried turning on an off various settings to
> > no avail thus far.
>
> > Joe
>
> I have tried another small change that might work.  This time, I have
> disabled the call to muse-colors-region in planner-mode; I don't know
> why that was there, though, so some other highlighting issue may
> appear.

That definitely speeds things up.  Updates now are nearly  
instantaneous and the time to refresh corresponds to the length/ 
complexity of the page.
> I removed the previous change, to ensure that this is the real source
> of the problem.

It is.
> Does this work for you?  If not, I am more or less stumped.

It does.  Thanks.
> On a separate matter: Looking at the calendar insinuation behaviour,
> it looks as though all pages visited by moving through the calendar
> get removed when the calendar is exited, including any that were there
> before M-x calendar was invoked.  Is this reasonable behaviour?  I
> don't use this feature regularly, but I would have thought that any
> pre-existing page buffers should be retained.

I have never noticed this behavior as planner/muse will load a page  
anyway when I access it.  I guess the proper emacs way would be to  
leave them in memory rather than removing them though.  I am  
sometimes surprised by how few planner buffers are loaded after  
several days of us, and I guess this is due to this calendar  
insinuation behavior.

Best,
Joe

> Regards,
>
> --
> Jim Ottaway
>
> --- orig/planner.el
> +++ mod/planner.el
> @@ -821,8 +821,7 @@
>  (defmacro with-planner-update-setup (&rest body)
>    "Execute BODY then save buffers according to `planner-tasks-file- 
> behavior'.
>  Also sets some variables to modify font-lock behaviour while  
> updating."
> -  (let ((muse-mode-highlight-p nil)
> -        (live-buffers (make-symbol "live-buffers")))
> +  (let ((live-buffers (make-symbol "live-buffers")))
>      `(save-window-excursion
>         (save-excursion
>           (save-restriction
> @@ -977,9 +976,9 @@
>      (planner-setup-highlighting)
>      (when (fboundp 'easy-menu-add)
>        (easy-menu-add planner-menu planner-mode-map))
> -    (planner-prepare-file)
> -    (when (and font-lock-mode muse-mode-highlight-p)
> -      (muse-colors-buffer))))
> +    (planner-prepare-file)))
> +;;     (when (and font-lock-mode muse-mode-highlight-p)
> +;;       (muse-colors-buffer))))
>
>  (defvar planner-date-regexp
>    "\\<\\([1-9][0-9][0-9][0-9]\\)\\.\\([0-9]+\\)\\.\\([0-9]+\\)\\>")
>
>
>
> _______________________________________________
> emacs-wiki-discuss mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/emacs-wiki-discuss

---
Joseph Kiniry
School of Computer Science and Informatics
UCD Dublin
http://secure.ucd.ie/
http://srg.cs.ucd.ie/