Re: Re: Clock report (R from the agenda)

[Erik Iverson <eriki@ccbr.umn.edu>]

I don't know what this thread is about, but it seems related.
Aapologies if I'm
hijacking it, but as of my latest pull just this morning,
my agenda has no clocktable in it, where as it previously
did.  I have not changed anything as far as I know.

Thanks!

--Erik

Sébastien Vauban wrote:
> Hi Carsten,
> 
> Carsten Dominik wrote:
>> On Nov 4, 2010, at 11:39 AM, Sébastien Vauban wrote:
>>> Carsten Dominik wrote:
>>>> you should now be able to use `C-u R' to achieve this.
>>> OK.
>>>
>>>
>>>> The lighter in the mode line will then switch from "Clock" to "Clock{}",
>>> A detail: I'd eventually would have written "Clock/" to remind the "/" used
>>> for applying the filters.
>> I did use {} because the current filter is actually listed in the mode line,
>> surrounded by {}.
>>
>>> Another: could we append, in the modeline, the tags (or their abbrev, such
>>> as "w" for "work") used in the filtered view?
>> The filter *is* shown in the mode line.  Just not as part of the Clock
>> lighter.
> 
> I wanted to (re-)test these, but now, when doing R, nothing is appended into
> my buffer with logged times. Did I do something wrong?
> 
> Same with C-u R.
> 
> 
>>>> and the current tags filtering should apply to the clock table in the
>>>> agenda.
>>>>
>>>> Please test this and report back.
>>> Only minor thing: while the logged lines are correctly shown or made
>>> invisible in the grid time, you need to refresh the table with "g" for it
>>> to display the correct values.
>>>
>>> Until that, what's above is not in sync' with what's in the table. Isn't
>>> there a way to make this refresh happen automatically?
>> That would be possible. However, the whole idea of filtering is to be *very*
>> fast, it works by hiding lines that are already in the buffer. Doing a
>> refresh for each change in filter would be time consuming. So I'd say having
>> to refresh by hand if the clock is showing filtered stuff is the smaller
>> evil.
> 
> I would privilege coherency of sums above small delay in table appearance. You
> know, when we look at tables for chasing time, we really need trustable
> figures.
> 
> In fact, I don't really understand your argument: if I want quick reports, I
> would just choose for the unfiltered view. If I need detailed sums of clocked
> times, I would go for the "filtrable" view (by C-u R) and would accept a
> little delay.
> 
> If you really don't share this vision, could you at least make this
> customizable?  TIA.
> 
> 
>>>>> : | File             | L | Headline             | Time    |      |
>>>>> : |------------------+---+----------------------+---------+------|
>>>>> : |                  |   | *Total time*         | *10:15* |      |
>>>>> : |------------------+---+----------------------+---------+------|
>>>>> : | Clock-Report.org |   | *File time*          | *10:15* |      |
>>>>> : | Clock-Report.org | 1 | Work                 | 8:09    |      |
>>>>> : | Clock-Report.org | 2 | Client A             |         | 3:23 |
>>>>> : | Clock-Report.org | 2 | Client B             |         | 4:46 |
>>>>> : | Clock-Report.org | 1 | Personal             | 2:06    |      |
>>>>> : | Clock-Report.org | 2 | DONE Lunch with Mary |         | 2:06 |
>>> Nice new layout. Much clearer for the levels...
>> Yes, I think so to.  Try :compact t, that is also nice, I think.
> 
> I'll have a look -- when reports will come back to life.
> 
> 
>>> | File             | Headline                 | Time    |      |
>>> |------------------+--------------------------+---------+------|
>>> |                  | ALL *Total time*         | *10:15* |      |
>>> |------------------+--------------------------+---------+------|
>>> | Clock-Report.org | *File time*              | *10:15* |      |
>>> |                  | Work                     | 8:09    |      |
>>> |                  | \__ Client A             |         | 3:23 |
>>> |                  | \__ Client B             |         | 4:46 |
>>> |                  | Personal                 | 2:06    |      |
>>> |                  | \__ DONE Lunch with Mary |         | 2:06 |
> 
> A couple of days ago, when it still worked, I've just seen something really
> painful: all the Org buffers were referenced in that table, even those which
> should not participate -- because I did not clock any time in them (for that
> or those days).
> 
> Plus, having many Org files, in fact, I did not see anymore the lines with the
> real time, as there were many lines with 0:00 time before them...
> 
> Best regards,
>   Seb
>