Hi Jed,
 

> I think that you should allow for the possibility that a note may have 
> more than one question associated with it, and that a question can be 
> associated with multiple notes.


You're right. I am starting to think I was a little too focused on 
mimicking pen and paper.

Having these many-to-many relationships between notes and questions does, 
however, seem to require that all questions and all notes are individual 
tiddlers rather than fields of a note-tiddler holding all notes and 
questions.

Maybe the draft was by far stretching the abilites of fields too much,
in that it may not be very desireable for fields to hold too much content.

As for visualising such a many-to-many relationship,
I wonder what Cornell suggests how to show that on paper, I guess it 
doesn't (?).
So, maybe, mimicking the pen-and-paper approach 1:1 isn't the best idea, 
after all.

Looks like those requirements will need a thorough reqrite with such a 
paradigm shift...
That's ok, it's why I drafted them anyways to get a meaningful design 
before implementing anything.

That could probably be done by giving each question a field that contains a 
> list of all of the rowidex values for the notes associated with it.


Nope, with the current spec that won't be possible,
because in the above draft, a question dooesn't have a field, it is a field.

If I wanted to save a many-to-many mapping based on that paradigm,
I would need some extra field storing these relationships, e.g.:

*cn-mapping: cq1-cn1 cq1-cn2 cq3-cn5*

And then split and parse all these relationships so as to display them, 
perhaps by having dim-styled clones of questions showing up in the 
corresponding RowIndex, either duplicated (one question for many notes) or 
deferred (many questions to one note, thus more than one question listed in 
a row, those that are mapped not being directly editabe).

It may be a good idea to have the ability to add tags or something 
> equivalent to each note to allow searching for notes/questions on specific 
> topics.


That is probably the #1 reason why storing notes or questions as fields is 
a really bad idea.

To me, it now seems to boil down to this:
Questions and notes will (have to be) tiddlers with id-style names. (see 
below)

Since deleting a note breaks indexing I was trying to use in 
> http://inmysocks.tiddlyspot.com/#Note%20Taking%20Test%202 I made an 
> action-increment widget in the MathyThing plugin to increment the id but I 
> haven't gotten around to making a version of the note taking that uses it. 
> The widget doesn't have an option for leading zeros, but I don't think that 
> would be too hard to add. The plugin would benefit from the ability to give 
> a date/time or other format to be incremented so I should probably do that 
> anyway.
>

Please, let me know when you have it working. 
 

> If there is a button to add a new note/question as part of the view 
> templates, and that button was the only thing used to add them, then that 
> button could trigger the incrementing each time it was pressed but some way 
> to detect when the template is used would be better because it would help 
> avoid duplicate ids. 
>

As long as we have one reliable way to increment, I am fine.
If users managed to use the template without that button doing the 
incrementing,
then they're clearly doing something wrong.
 

> As you said, the prefix part can be done using WikiText. 
>
And the prefix would probably be a good way to distinguish between 
> classes/projects/whatever is being tracked.
>

I guess, after all, it would make sense that there is only one index and 
that questions, in fact, increment the very same index as notes, however, 
are somehow distinct, e.g. by having a *Question* tag, whereas notes could 
have a *Note* tag.

For example, a note would thus be stored as...

title: MA101-1
tags: Note

contents of the note with index=1 for the class Math-101

...and a question as...

title: MA101-2
tags: Question

contents of the question with index=2 for the class Math-101
this question is not linked to any note yet!

The relationships between questions and notes could then be entirely 
created using tags, e.g.

title: MA101-3
tags: Question MA101-1

contents of the question with index=3 for the class Math-101
referring to note MA101-1

One could even have a note referencing another note, e.g. a "dependent 
note"...

title: MA101-4
tags: Note MA101-1 MA101-2

contents of the note with index=4 for the class Math-101
referring to note MA101-1 and
referring to question MA101-2

I don't think there would be many new things required to make the backend 
> part. The view editing ability and the drag and drop list sorting would all 
> have to be implemented, but the way you have it laid out the backend looks 
> mostly independent of the display methods.


When you say "backend"... what exactly do you have in mind? Configuration 
such as...

   - the global tag for notes
   - the global tag for questions
   - the note prefix for a given note tiddler, e.g. a "class"

?

Best wishes, Tobias. 

-- 
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/tiddlywiki.
For more options, visit https://groups.google.com/d/optout.

Reply via email to