BETA: Project/Folder/Task parsing
pottster
Thanks for all the extra functionality you are adding to the Betas.
Just a couple of observations on the new parsing options for Projects/
Folders/Tasks.
Shouldn't the parsed task be added to the bottom of the list of
siblings as happens currently with a non-parsed task?
It would be useful if, in the RTE Parsing Result dialog, a destination
column could be added to confirm the task was going to land in the
right place.
Regards,
Ken
Andrey Tkachuk (MLO)
1) agree since it would be more consistent. Implemented.
2) It is hard in current architecture. The good news however that I
have just added Undo/Redo for inserting tasks from RTE. Together with
selection in the Outline all just inserted from RTE tasks it is much
more useful I think. So you can type in RTE, Ctrl+Enter to parse and
add to Outline, see where in the outline your tasks are placed, if
they are in wrong position use "Edit->Undo insert from RTE" and you
get your RTE text back and may correct it if needed and try again.
Hope that makes sense.
Andrey.
pottster
Appreciate your efforts Andrey.
On Feb 18, 8:06 am, "Andrey Tkachuk (MLO)"
pottster
Another quick observation.
Parsed tasks appear to be allocating themselves to already completed
tasks if there is a match with the parsed word. Is there any way to
restrict this to insertion under uncompleted tasks only?
Ken
Andrey Tkachuk (MLO)
A.
pottster
Forgive me for making the case a little more...
1) As it stands, from a practical point of view, users would probably
be restricted to parsing for very unusual words otherwise they would
have to remember the text in hundreds (maybe thousands!) of completed
tasks.
2) If you are adding parsed tasks to completed tasks you are
effectively uncompleting the parent. Even if this is what you wanted
to do (which I doubt) the current setup still leaves the parent task
completed.
I realise the requests made of you are growing all the time but
parsing in RTE is becoming a really important part of MLO and ninja
task management ;-)
Ken
On Feb 18, 3:03 pm, "Andrey Tkachuk (MLO)"
Ron Stockfleth
While you have made a good point, restricting parsing to to
uncompleted tasks will not help me all that much.
In April I will become a 5-year user of MLO. I have 12 thousand tasks
in my MLO file (I know, I really do need to purge my data file). I
have 3,500 uncompleted tasks in MLO (a significant reduction from my
total, but still too unwieldy for parsing into Project, folders,
tasks). Fortunately, I have only 150 active tasks right now (thank you
MLO for helping me keep my sanity)!
This added parsing feature is useful only to the extent that it can
assign the RTE task to the appropriate project, folder, task on the
very first attempt. If the user has to undo the assignment, or the
user has to look in his outline for the appropriate project/folder/
task name to assign his RTE task to, then the feature loses its
usefulness.
While on the surface this feature seems to be a good idea, its value
falls off as the size of the MLO data file grows. I have tried this
feature for the past 2 days, and I haven't got a correct assignment on
the first try yet. Limiting the parsing feature to uncompleted tasks
will not help me as I have just too many tasks in my data file.
Fortunately, my old tried-and-true approach still works for me:
1. Capture my thought and enter into RTE, automatically dropping it
into my Inbox (Collection phase)
Here, I simply capture the thought in the task caption. If that
thought is an actual appointment or deadline, I do enter date and
possibly the time.
2. Enter appropriate parameters to the tasks in my Inbox at a
convenient time for me (Process)
3. Drag fleshed out tasks to their appropriate place in my outline,
again, at a convenient time for me (Organize).
These 3 steps occur a number of times naturally throughout the day and
happen quite rapidly.
I use many of MLO's features, but this is one feature that will
definitely not work for me, with no fault to either Andrey nor to
those who suggested the idea.
There is one way that I think that this feature might prove useful to
me and to others that have reasonably large data files. And that is if
it operated similarly to the search function. That is, if for example
I added a parsing parameter such as -toprjMLO, and then MLO gave me a
list of all projects with MLO in the task caption, to which I could
select the appropriate project for my RTE task, then this definitely
would improve the usefulness of this parsing feature. Of course, if it
produced a list of 100 projects, then not so much.
One last comment...What if I can't remember whether the assignment I
wish to have is a project, folder, or simply a plain task? I would
think that you would want this feature to work if you could remember
the task caption but couldn't remember if it a project, folder. or
task.
I do not know how practical my suggestion would be to implement, and I
am not requesting that it be done this way. I am simply saying that
for this feature to be useful to me it would have to work something
like I just described.
Best regards,
Ron
pottster
Thanks for such a considered response. It's always interesting to hear
about other people's work methods and how they are using MLO.
The conclusion I'm rapidly coming to is that nobody works in exactly
the same way as somebody else. This is why MLO is so good because it's
flexible enough to accommodate most approaches. I suspect that this
has been one of Andrey's guiding principles - to allow users to work
the way they want to. We're probably all guilty of making requests
which suit our personal setup but Andrey has to assess whether they
will be of use to more than just one individual. I don't envy him, it
must be difficult.
Regarding your comments about the new parsing options: -
• I agree, it's usefulness is probably in direct proportion to the
size of the Outline. Having said that I don't think it's intended to
be anything more than a shortcut to regularly used places. I don't
think it's a sensible way to attach tasks to any task in your Outline.
• I don't know if you use folders as placeholders but that would
definitely help with restricting the number of parsing options you
have to remember and increase the likelihood of a correct assignment
of the task.
• Trying to think creatively, you could perhaps name your tasks with a
built in tag to allow for more accurate parsing e.g. a task called
"#Wash car" would be found by an RTE entry of "Buy a bucket -to#wash"
even if there were another task with "wash" in it. You would have to
remember the tags though, so again, it might only work for a limited
number.
• I think your suggestion for a search as you type function wouldn't
be implemented through parsing but more likely via the Parent button
in the RTE dialog. I agree this would be useful. Although, having said
that, there is already an incremental search as you type available
from the Parent Task button but I think it has two main drawbacks.
Firstly, it stops when it finds the first match rather than offering a
choice of matches and secondly, it makes the selected location become
the new default.
• Regarding the point about not remembering whether a task is a
project folder or task; I may be wrong but I think tofld and toprj are
just for narrowing the selection and to"any word in the caption" will
still find it even if it's a project or folder. Try it.
Regards,
Ken
PS hope you've got an external backup of that data file!
PPS congrats on your impending 5th anniversary - perhaps Andrey will
send you a card ;-)
Ron Stockfleth
That's what I love about MLO, 12,000 tasks in a 1 MB data file and MLO
is still lightning quick! Yeah, I run Windows Home Server on my home
network with my MLO data file residing on the server with the MLO
folder set to offline folder availability. The result is whether I am
working on my desktop computer or if I grab a netbook computer when I
run out of the house, I always have the latest MLO data file with me.
Sweet!
When Andrey has the iPhone application up-and-running, I will have my
iPhone data with me even if my house burns down! :) I guess that
should have been a :( but us MLO users can be fanatical.
Happy parsing!
Ron
Lisa Stroyan
The good news however that I
have just added Undo/Redo for inserting tasks from RTE. Together with
selection in the Outline all just inserted from RTE tasks it is much
more useful I think. So you can type in RTE, Ctrl+Enter to parse and
add to Outline, see where in the outline your tasks are placed, if
they are in wrong position use "Edit->Undo insert from RTE" and you
get your RTE text back and may correct it if needed and try again.
Lisa
Lisa Stroyan, mailto:lstr...@gmail.com
www.empathic-parenting.com
Lisa Stroyan
Andrey Tkachuk (MLO)
> restrict this to insertion under uncompleted tasks only?
I have implemented it this way:
A new Project for a task on -toXXX command may be completed only if
there are no any there matches for your input.
Here is the full list of priorities when selecting a task from the
whole list matched to the -toXXX command:
1) Task without completed parent has more priority
2) Task with full match of the name has more priority
3) Task which starts from the input string has more priority
4) Task which is closer to the root has more priority
Andrey.
On Feb 18, 3:52 pm, pottster <kenwarren...@googlemail.com> wrote:
> Andrey,
>
> Another quick observation.
>
> Parsed tasks appear to be allocating themselves to already completed
> tasks if there is a match with the parsed word.
>
> Ken
pottster
Look forward to trying it.
On Feb 20, 11:58 am, "Andrey Tkachuk (MLO)"