On Mar 15, 2009, at 5:26 PM, Bastien wrote:

>>> property's value, have a way to use a calc or a tbl formula?
>>> Additionnally, have a _DEFAULT suffix to define defaults.
>>
>> There was a patch recently on the mailing list, subject "elisp
>> formulas....."
>
> Seen it, will test it closely.

Please do, I am interested in this patch, maybe you see ways
to expand the possibilities.

>
>>> - FR: A function to display the link the cursor is on (the same way
>>> the
>>> mouse displays a small popup.)  This function would particularily be
>>> useful in the agenda view.
>>
>> C-c C-l  LOOK  RET RET   :-)
>
> :)
>
>> Hard to find new key bindings for these.
>
> But C-c C-l is not accessible from the agenda, right?

No, it is not, and I'd rather not make it.  There could be several links
in a headline, and the agenda is really meant to have commands that
do not depend on where in the line the cursor is.

I guess you could add to post-command-hook in the agenda something
like this, to use the echo area if it is empty:

(add-hook 'org-agenda-mode-hook
	  (lambda ()
	    (add-hook 'post-command-hook
		      (lambda ()
			(unless (current-message)
			  (if (looking-at ".*?\\[\\[\\(.*?\\)\\]")
			      (message "Link to: %s"
				       (org-match-string-no-properties 1)))))
		      t t)))

>
>>> - FR: When selecting an attachment, if there is already a directory
>>> defined for this the current entry, then use this directory as the
>>> default directory for browsing and select the attachment.
>>
>> You mean if there is an attachment directory defined?
>
> Yes.
>
>> What do you mean by "when selecting and attachment"?
>
> C-c C-a should be aware of any previously defined directory, and take
> this directory as the default directory when browsing for other files
> to attach.  Am I more clear?

No, not clear to me.

Do you mean that the attachment directory should become the default?
No, maybe you mean that if I have already attached a/b/c/file,
the next time you want to attach a file from the same directory, a/b/c?



>
>>> - FR: Allow to insert "+1d"-type strings when picking up a date with
>>> `org-time-stamp'.
>>
>> This works for me, what am I missing here?
>
> Weird.  C-c C-s + 1 d RET does only insert today's date here, no "+1d"
> string.
>
>>> - FR: Symmetric to individual lead time, I wish I can have  
>>> individual
>>> "remaining" time - for example, the number of days during which a
>>> deadline is still visible in the agenda view.  (That would happily
>>> steal 90% of what org-expiry.el is trying to achieve.)
>>
>> I think the purpose of deadlines is for them sting you until you do
>> something about them?  For me it seems that this would be counter
>> productive.  Why don't you tick off these items ok archive/kill  
>> them????
>
> Well, let's say I plan to do something next week but it's not high
> priority.  I want it to pop up in my agenda view next week, and stay  
> for
> three days, because passed this delay, I know I won't have any  
> interest
> in doing it anymore.  Having something like +3d looks consistent to me
> -- but I agree I can find workarounds.  Or just fix my habits :)
>
>>> - FR: Following a sh: or an elisp: link prompts the user for
>>> confirmation.  Can we have a variable that let the user to globally
>>> turn this off?
>>
>> Yes, there is such a variable: org-confirm-elisp-link-function
>> and org-confirm-shell-link-function.
>
> Great, thanks!
>
>>> - FR: Visually mark (with #) entries that are marked in the agenda
>>> view?
>>
>> I am not sure what you are talking about, what marking???
>
> Sorry I meant when marking an entry with `k m' in agenda view.
>
>>> - FR: When setting a mark in the agenda view, go to the next line?
>>> (Like in buffer-menu and other menu mode.)
>>
>> I am not sure what you are talking about, what marking???
>
> Same here.  `k m' would jump to the next unmarked entry.
>
>>> - FR: Maybe `?' could provide a little help in agenda mode.
>>
>> Hmmm, yes, why not.  What should it say?
>
> Will think of it.
>
>> Because, the keys are available in the menu.
>
> Yeah, but maybe some people don't use menus (I don't).

I use them to look up keys, nothing else :-)

>
>>> - FR: Grouping several :clock: lines when clocking periods are very
>>> close to each other (with a user-defined threshold).  For example,
>>> clocking 5 times 10 minutes within the same hour would be rounded up
>>> to one hour.
>>
>> With things nicely tugged way into drawers, I see no reasons to do  
>> this.
>
> The main reason was to cheat and try pretend you've been working 1  
> hour
> when you really worked 50 minutes :)

Yes, I can see that, but anything automatic would probably create  
crap.  Editing time stamps is easy enough though....
>
> Great!
>
>>> - Q/FR: How to go at the end of a field in a table?  Maybe C-a and  
>>> C-e
>>> could be contextually redefined to do this, since a table is such a
>>> specific editing context.
>>
>> Word-motion commands?  I am not comfortable with overloading C-a  
>> and C-
>> e.
>
> forward-paragraph and backward-paragraph would make sense, no?  For  
> now
> M-} jumps out of the table.
>
>>> `C-c C-d SPC', `C-c C-s SPC' could respectively delete deadlines and
>>> scheduled properties.  Or maybe C-c C-u could remove any timestamp
>>> (this keystroke is free in org-agenda-mode since outline-mode is not
>>> loaded.)
>>
>> I believe C-u C-c C-s and C-u C-c C-d do this for Deadlines.  You men
>> the time stamp w/o any keyword????
>
> The problem is that you don't see whether an entry is scheduled or  
> has a
> deadline in agenda view.  How do you decide which one to delete?

Well, the prefix in the agenda does tell you if it is scheduled or a  
deadline!

- Carsten