Horizontal Parson problems

[Rob Beezer]

For Brad - working on the details for horizontal Parsons problems. Perhaps you
can refresh your memory on these and we can discuss tomorrow via Zoom? Don't
bother to send back a lengthy email. I'm going to work on static for a bit, and
can then move on to other things that also need attention, so this is not
holding me up.

1. There's a div.hparsons inside div.hparsons_section that is always empty.
I guess it is necessary?

2. textarea/@data-dburl="/_static/test.db" A 100% mystery to me. And perhaps
irrelevant if not on a Runestone server? Problems seem to function for me
locally (well, one SQL statement version) but there is an allusion to feedback
based on execution in the overview samples.

3. Is there support or considerations for "natural" language versions?

4. Randomization. For vertical Parsons, we author the blocks in the correct
order (the answer), and they get randomized on presentation. Horizontal Parsons
has a textarea/@data-randomize attribute (true/false). I'd presume the
default is "true". And then there is textarea/@data-blockanswer which always
has the values "0 1 2 3" or "0 1 2 3 4 5" in the samples. I'm guessing you can
author out-of-order, set randomize to "false", and specify the ordering for the
answer this way? There is an option to "reuse" blocks, which is simple enough
to implement (I have), but perhaps @data-blockanswer is a way to specify
reuse's for the answer? I guess I'd have a PTX "block" point to another one to
indicate reuse for the answer.

Anyway, this feels flipped on its head, authoring vertical v. horizontal. For
horizontal, I also have authors prescribe a fixed "random" order for the
presentation in static forms. Which might be the idea in the
no-randomize/block-answer scenario?

I'm looking for consistency for authors, not ease of coding. Reuse is new, but
I think I can handle that gracefully. And I can probably build any necessary
bijection. And I may be totally confused about the above.


Chat Tuesday? Perhaps at 1 PM Pacific, if there are other things going on
during the abbreviated Drop-In.

[Zihan Wu]

Hi Rob,

I'm Zihan Wu, and I am a third-year PhD student. I implemented the horizontal Parsons problems in another web platform and later migrated it to Runestone, which might have caused the inconsistency for authors. I would be happy to help with editing/fixing the directive, and if you have a meeting set up to discuss some works around it I would love to join to see what I can do!

Thanks,

Zihan

[Rob Beezer]

Dear Zihan,

There will be a developer Drop-In tomorrow from 11-Noon Pacific. It is usually
two hours, but Brad M and I both have another meeting at Noon.

There is another one on Fridays (not this Friday?) Noon-3 Pacific.

These are announced the morning-of on the "pretext-announce" Google Group, which
does not require permission to join. So you can find a Zoom link there.

I'll look forward to chatting with you sometime soon when it fits your schedule.

Thanks,
Rob

> --
> You received this message because you are subscribed to the Google Groups
> "PreTeXt development" group.
> To unsubscribe from this group and stop receiving emails from it, send an email
> to pretext-dev...@googlegroups.com
> <mailto:pretext-dev...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/pretext-dev/25160fc8-25e9-49da-80b6-e7231874fc9dn%40googlegroups.com <https://groups.google.com/d/msgid/pretext-dev/25160fc8-25e9-49da-80b6-e7231874fc9dn%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Rob Beezer]

Mostly for Brad, Zihan, and Barb. A real rough cut, but I think all the
features are supported. Nothing is pushed yet. I know how to do static
versions, but that code is not ready.

Problems 7-11
https://pretextbook.org/beta/2022-11-23-horizontal-parson/interactive-exercises.html

Problem 9 points to a SQL database ("test.db") that I ripped out of the
Runestone Overview. And has unit tests. I can't tell if it is working. I
changed one unit test to be incorrect and got no indication. And nothing in the
console. And tried a leading slash on the path to the DB as well. So maybe
there's some debugging that can be done there. Or maybe I still don't
understand that scenario.

I've seen a few small things. Maybe this is an easy one: the id is at the top
of the blue box, not in the empty parentheses at the bottom.

Rob


On 11/21/22 12:09, Zihan Wu wrote:

[Rob Beezer]

Horizontal Parson problems are now available. But maybe need some more work, so
a soft opening.

Documentation:
Subsection 4.12.4: Horizontal Parsons Problems
https://pretextbook.org/doc/guide/html/topic-interactive-exercises.html#horizontal-parsons-problems

Samples:
Exercises 5.9: Horizontal Parsons Exercises
https://pretextbook.org/examples/sample-book/noparts/html/horizontal-parsons-exercises.html

1. The previously reported problem with IDs in the wrong place was my fault.
Fixed now.

2. These are built on the assumption that the blocks are pieces of computer
code. Both in the Javascript, and therefore in what I've done for PreTeXt. I
think a "natural language" variation would be useful. Rearrange words of a
sentence, or terms of an equation.

3. There is some inconsistency in the features of a vertical problem v. a
horizontal problem, which I think are independent of the layout. Vertical
allows distractors, but I did not implement that for vertical (having seen no
examples). Horizontal allows reuse of blocks, I think that is impossible right
now with vertical, barring new Javascript.

Perhaps some discussion on Tuesday would provide the way forward.

Rob

[Rob Beezer]

And now with horizontal distractors. I believe I've made the necessary
adjustments so that these blocks will be part of the problem, and not part of
the solution. But there are lots of combinations of features and ways to
exercise these exercises. So please test yours out if you are an early adopter.

Exercise 5.: Parsons Problem, Natural Language, with Distractors
https://pretextbook.org/examples/sample-book/noparts/html/horizontal-parsons-exercises.html#horizontal-parson-natural-distractor

A note for Brad and Zihan. It appears the "blockanswer" attribute requires
*exactly* one space as a separator between indices. Not sure if that is
documented somewhere. Didn't hang me up for too long.

Rob