Clocking in the current task should clock it out first

Clocking in the current task should clock it out first

[Daniel Clemente]

Hi,
  in recent org-modes a new behaviour was added: when doing C-c C-x
C-i on the current task, it isn't clocked out first. It shows the
message „Clock continues in "[task]"“ and adds a new line for the
clock in.
  This creates a clock section like:

#+BEGIN_EXAMPLE
*** after pressing many successive C-c C-x C-i …
	:CLOCK:
	CLOCK: [2010-03-17 dc 10:25]--[2010-03-17 dc 10:30] =>  0:05
	CLOCK: [2010-03-17 dc 10:20]
	CLOCK: [2010-03-17 dc 10:20]
	CLOCK: [2010-03-17 dc 10:20]
	CLOCK: [2010-03-17 dc 10:20]
	CLOCK: [2010-03-17 dc 10:20]
	CLOCK: [2010-03-17 dc 10:20]
	CLOCK: [2010-03-12 dv 16:38]--[2010-03-12 dv 16:39] =>  0:01
	:END:
#+END_EXAMPLE


  They are later correctly found to be dangling clocks.
  I presume this is a bug?

Daniel

Re: Clocking in the current task should clock it out first

[Bernt Hansen]

Daniel Clemente <n142857@gmail.com> writes:

> Hi,
>   in recent org-modes a new behaviour was added: when doing C-c C-x
> C-i on the current task, it isn't clocked out first. It shows the
> message „Clock continues in "[task]"“ and adds a new line for the
> clock in.
>   This creates a clock section like:
>
> #+BEGIN_EXAMPLE
> *** after pressing many successive C-c C-x C-i …
> 	:CLOCK:
> 	CLOCK: [2010-03-17 dc 10:25]--[2010-03-17 dc 10:30] =>  0:05
> 	CLOCK: [2010-03-17 dc 10:20]
> 	CLOCK: [2010-03-17 dc 10:20]
> 	CLOCK: [2010-03-17 dc 10:20]
> 	CLOCK: [2010-03-17 dc 10:20]
> 	CLOCK: [2010-03-17 dc 10:20]
> 	CLOCK: [2010-03-17 dc 10:20]
> 	CLOCK: [2010-03-12 dv 16:38]--[2010-03-12 dv 16:39] =>  0:01
> 	:END:
> #+END_EXAMPLE
>
>
>   They are later correctly found to be dangling clocks.
>   I presume this is a bug?

Hi Daniel,

Yes I believe this is a bug.  I think I've also run into this issue but
I have auto clock resolution disabled so this is not leaving open clocks
in my setup.

(setq org-clock-auto-clock-resolution nil)

This of course is only a temporary work around until a real fix occurs.
I haven't had the time to investigate this yet but it is on my list of
things to look at.

-Bernt

Re: Re: Clocking in the current task should clock it out first

[Carsten Dominik]

Hi,

strangely enough, this does not happen for me.  Maybe you
have some setup for clock resolution that I do not have?

- Carsten

On Mar 17, 2010, at 2:06 PM, Bernt Hansen wrote:

> Daniel Clemente <n142857@gmail.com> writes:
>
>> Hi,
>>  in recent org-modes a new behaviour was added: when doing C-c C-x
>> C-i on the current task, it isn't clocked out first. It shows the
>> message „Clock continues in "[task]"“ and adds a new line for the
>> clock in.
>>  This creates a clock section like:
>>
>> #+BEGIN_EXAMPLE
>> *** after pressing many successive C-c C-x C-i …
>> 	:CLOCK:
>> 	CLOCK: [2010-03-17 dc 10:25]--[2010-03-17 dc 10:30] =>  0:05
>> 	CLOCK: [2010-03-17 dc 10:20]
>> 	CLOCK: [2010-03-17 dc 10:20]
>> 	CLOCK: [2010-03-17 dc 10:20]
>> 	CLOCK: [2010-03-17 dc 10:20]
>> 	CLOCK: [2010-03-17 dc 10:20]
>> 	CLOCK: [2010-03-17 dc 10:20]
>> 	CLOCK: [2010-03-12 dv 16:38]--[2010-03-12 dv 16:39] =>  0:01
>> 	:END:
>> #+END_EXAMPLE
>>
>>
>>  They are later correctly found to be dangling clocks.
>>  I presume this is a bug?
>
> Hi Daniel,
>
> Yes I believe this is a bug.  I think I've also run into this issue  
> but
> I have auto clock resolution disabled so this is not leaving open  
> clocks
> in my setup.
>
> (setq org-clock-auto-clock-resolution nil)
>
> This of course is only a temporary work around until a real fix  
> occurs.
> I haven't had the time to investigate this yet but it is on my list of
> things to look at.
>
> -Bernt
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Please use `Reply All' to send replies to the list.
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode

- Carsten

Re: Re: Clocking in the current task should clock it out first

[Daniel Clemente]

El vie, mar 19 2010 a les 18:36, Carsten Dominik va escriure:
> Hi,
>
> strangely enough, this does not happen for me.  Maybe you
> have some setup for clock resolution that I do not have?
>

No. Just Emacs:
GNU Emacs 24.0.50.1 (i686-pc-linux-gnu, GTK+ Version 2.18.6) of 2010-03-13

I used this .emacs:
(add-to-list 'load-path "/w/org-mode/lisp") (require 'org-install)
(add-to-list 'load-path "/w/org-mode/contrib/lisp")
(require 'org)


Commit 29d945720095a65852f69b7c628c1199eb4961fe (Date:   Fri Nov 27 08:09:10 2009 +0100) was done precisely to leave clocks open; the bad behaviour comes from this change.
Via bisect I found that the previous revision was good and the next revision bad.


-- Daniel

Re: Clocking in the current task should clock it out first

[Bernt Hansen]

Hi,

I tried playing with this again last weekend and I can't get it to break
on my Linux machine with GNU Emacs 22.2.1

I've seen it break (at least once) on windows at work... but I haven't
tried to reproduce that in my spare time.  Somehow I just can't get
myself to use Windows for 'fun' :-P

I tried the auto clock resolution with my normal clocking setup (where
the clock runs all the time, and I explicitly clock in and out
regardless of todo state changes).

I can't get the clock resolution/idle time code to do anything really
useful in my setup.  If I do org-resolve-clocks when my current task is
clocking it asks for how many minutes to keep etc, and then clocks in
from *now* leaving a hole in my clock data which I don't want.

I'm not sure exactly how this stuff is supposed to work - maybe John can
shed some light on this.  I can't find documentation about resolving
clocks in the regular org-mode documentation either.  I remember an
article John posted on the mailing list but I don't think that got into
the official org-mode documentation other than the lisp functions and
docstrings.

I have also notice that clock resolution (in the distant past) would
overlap clock times.  If it finds more than one open clock it can
resolve them so that they overlap with other clock entries and for me
that's _really_ _really_ bad.  I'd rather have it to nothing than create
hard-to-find overlapping clock entries.

For now I'll continue to run with org-clock-auto-clock-resolution set to
nil until I understand how this is supposed to work and make life better
for me :)

I would really like it to be more useful than it is in my setup but I'm
not really missing this functionality alot at the moment and my time to
play with it is extremely limited at present.

Regards,
Bernt


Carsten Dominik <carsten.dominik@gmail.com> writes:

> Hi,
>
> strangely enough, this does not happen for me.  Maybe you
> have some setup for clock resolution that I do not have?
>
> - Carsten
>
> On Mar 17, 2010, at 2:06 PM, Bernt Hansen wrote:
>
>> Daniel Clemente <n142857@gmail.com> writes:
>>
>>> Hi,
>>>  in recent org-modes a new behaviour was added: when doing C-c C-x
>>> C-i on the current task, it isn't clocked out first. It shows the
>>> message „Clock continues in "[task]"“ and adds a new line for the
>>> clock in.
>>>  This creates a clock section like:
>>>
>>> #+BEGIN_EXAMPLE
>>> *** after pressing many successive C-c C-x C-i …
>>> 	:CLOCK:
>>> 	CLOCK: [2010-03-17 dc 10:25]--[2010-03-17 dc 10:30] =>  0:05
>>> 	CLOCK: [2010-03-17 dc 10:20]
>>> 	CLOCK: [2010-03-17 dc 10:20]
>>> 	CLOCK: [2010-03-17 dc 10:20]
>>> 	CLOCK: [2010-03-17 dc 10:20]
>>> 	CLOCK: [2010-03-17 dc 10:20]
>>> 	CLOCK: [2010-03-17 dc 10:20]
>>> 	CLOCK: [2010-03-12 dv 16:38]--[2010-03-12 dv 16:39] =>  0:01
>>> 	:END:
>>> #+END_EXAMPLE
>>>
>>>
>>>  They are later correctly found to be dangling clocks.
>>>  I presume this is a bug?
>>
>> Hi Daniel,
>>
>> Yes I believe this is a bug.  I think I've also run into this issue
>> but
>> I have auto clock resolution disabled so this is not leaving open
>> clocks
>> in my setup.
>>
>> (setq org-clock-auto-clock-resolution nil)
>>
>> This of course is only a temporary work around until a real fix
>> occurs.
>> I haven't had the time to investigate this yet but it is on my list of
>> things to look at.
>>
>> -Bernt
>>
>>
>> _______________________________________________
>> Emacs-orgmode mailing list
>> Please use `Reply All' to send replies to the list.
>> Emacs-orgmode@gnu.org
>> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>
> - Carsten
>
>
>
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Please use `Reply All' to send replies to the list.
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode

Re: Clocking in the current task should clock it out first

[John Wiegley]

On Mar 22, 2010, at 8:25 PM, Bernt Hansen wrote:

> I can't get the clock resolution/idle time code to do anything really
> useful in my setup.  If I do org-resolve-clocks when my current task is
> clocking it asks for how many minutes to keep etc, and then clocks in
> from *now* leaving a hole in my clock data which I don't want.
> 
> I'm not sure exactly how this stuff is supposed to work - maybe John can
> shed some light on this.  I can't find documentation about resolving
> clocks in the regular org-mode documentation either.  I remember an
> article John posted on the mailing list but I don't think that got into
> the official org-mode documentation other than the lisp functions and
> docstrings.

Can you be a bit clearer here, Bernt?  I use the clock resolver constantly and have not been having the kinds of problem you describe.  I'd like to get these hammered out now, though.

> I have also notice that clock resolution (in the distant past) would
> overlap clock times.  If it finds more than one open clock it can
> resolve them so that they overlap with other clock entries and for me
> that's _really_ _really_ bad.  I'd rather have it to nothing than create
> hard-to-find overlapping clock entries.

Yes, this is a possibility. :(  Org-mode doesn't handle multiple open clocks gracefully in other ways too, so I didn't try to hard to avoid situations like these in the resolver.

John

Re: Clocking in the current task should clock it out first

[Bernt Hansen]

John Wiegley <jwiegley@gmail.com> writes:

> On Mar 22, 2010, at 8:25 PM, Bernt Hansen wrote:
>
>> I can't get the clock resolution/idle time code to do anything really
>> useful in my setup.  If I do org-resolve-clocks when my current task is
>> clocking it asks for how many minutes to keep etc, and then clocks in
>> from *now* leaving a hole in my clock data which I don't want.
>> 
>> I'm not sure exactly how this stuff is supposed to work - maybe John can
>> shed some light on this.  I can't find documentation about resolving
>> clocks in the regular org-mode documentation either.  I remember an
>> article John posted on the mailing list but I don't think that got into
>> the official org-mode documentation other than the lisp functions and
>> docstrings.
>
> Can you be a bit clearer here, Bernt?  I use the clock resolver
> constantly and have not been having the kinds of problem you describe.
> I'd like to get these hammered out now, though.

I'll try :)

I just turned it on in my setup by setting the following variables for
this session:

,----
| org-clock-auto-clock-resolution is a variable defined in `org-clock.el'.
| Its value is 
| when-no-clock-is-running
`----

,----
| org-clock-idle-time is a variable defined in `org-clock.el'.
| Its value is 1
`----

I haven't tried the idle time stuff yet.  I have a clocking entry to
test with that looks like this:

,----
| * NEXT Respond to John's email on clocking
|   SCHEDULED: <2010-05-16 Sun>
|   :CLOCK:
|   CLOCK: [2010-05-16 Sun 15:29]
|   :END:
`----

and I've been playing with starting clocks with open clock entries.  I'm
a bit confused about the prompt when I clock in with an open clock.  The
prompt seems to be geared towards the idle time stuff but I currently
don't have any idle time -- I'm just clocking in an open entry.

If I clock out, change the clock entry as above so that it is dangling
and then clock it in I get the following prompt:

"Dangling clock started 36 minutes ago [(kK)p (sS)ub (C)ncl (i)gn]"

As far as I can tell k and K are identical, s, S, and C are identical
and i leaves it untouched.

k and K asks me how many minutes to keep and then clocks in the entry
from now.

s, S, and C removes the open clock and clocks in the entry from now.

i leaves it alone and clocks in from the time of the open entry.

For this case I think it would make sense to simplify the prompt to just
k, s, and i (or whatever makes sense in this scenario).

Setting org-clock-auto-clock-resolution to t doesn't make any sense in
my setup since it asks for every clock-in what I want to do with the
minutes on my currently clocking entry.

I'm about to go light the barbecue for dinner... so I'll see what the
idle time stuff does shortly I hope.

...

Well that didn't seem to work.  It's been 15 minutes and as far as I can
tell the idle time stuff didn't do anything.  I clocked in another task
when the clock was still running and it simply closed the clock of my
open task and started the new clock normally.

Am I missing something in my setup to enable the idle time detection?

Regards,
Bernt