TiddlyWiki in Education

[Steven Schneider]

There seems to be some interest in this group in using TiddlyWiki as a platform for teaching, and for delivering educational resources such as textbooks and presentations. I thought we might use this thread to identify those interested in this kind of work, and see if there is critical mass to spin off a few collaborative projects, or at least to bring together some existing resources. 

So, here goes:

I'm a professor in Information Design at SUNY Polytechnic Institute in NY, USA (hence my user name, stevesuny :). I've used TW in classes for the past 6-7 years, both wikifying texts and teaching students to design and write interactive texts. At the moment, I am interested in:

# Rendering open texts published as ePub (or mobi or wxr or odt: this is one example) in TW so that students can submit assignments as comments on the text. 

# Developing tools that use resources from the Internet Archive as teaching materials.

# Expanding and enhancing DesignWriteStudio as a platform for teaching reading & writing of interactive texts, using TiddlyWiki as both the object of study as well as object of activity.

If there is interest, perhaps we might spend the next few weeks writing brief case studies of what we've done, and what we'd like to do? Or, especially for those in higher education, perhaps there is a conference we could collaboratively prepare a session proposal for? Anyone with particular interests in education or textbooks planning on attending the Euro TW Meet-Up in Oxford next month? 

//steve.

[@TiddlyTweeter]

Ciao Steve

Very interested in this. I have long career as an ethnographer (backround: LSE, MPhil, anthropologist) and archivist (some on-line archives).

I have a large live project right now I am working on with other people. I hope to use TW for it.

It is all about Charles Dickens' Great Expectations and works towards a fully annotated version of the text. The project also examines numerous films that have been made of the book for possible new film versions of it.

I'm interested in TW as a solution because it seems to be able to combine rigor of fidelity to original texts, whilst allowing a lot of customized freedom on annotation, commentary AND new innovation.

I am not a programmer, which is my handicap, but good on specifying need. I can willingly, given a real group doing real things, take time to document in detail the requirements of the project, so long as there is a chance it helps me actually instantiate it in TW.

Best wishes
Josiah

[Lost Admin]

On Monday, May 8, 2017 at 8:00:17 PM UTC-4, Steven Schneider wrote:

...

# Rendering open texts published as ePub (or mobi or wxr or odt: this is one example) in TW so that students can submit assignments as comments on the text. 

...

A couple of times now, I've seen this talk of TW and ePub. I've done a bit of digging into the epub format and I don't understand how you can integrate TW and epub. ePub is a container format for text documents that uses HTML to encode the document but it also places some pretty tight limits on the HTML and per-html-file size. On top of all that, the document needs to fall-back gracefully to the limits imposed by ebook readers with limited RAM. So, how do you propose to merge the two?

I suppose, if you kept the TW file small enough you could wrap it in the structure of an epub and it might render properly in some ebook readers but I don't think you could save changes back into the epub file without a custom epub reader.

[Mark S.]

I think he's saying that he wants to convert existing texts into TW.

I've gone through this exercise with some classical texts (e.g. a couple plays at shakespeare.tiddlyspot.com, bibles). Classical text doesn't have special formatting to worry about. Well structured data can be imported fairly quickly. The questions for the mentioned epubs would be, do you want to bring in all the text formatting, and if so, do you want to do it as TW markup, Markdown markup, or original HTML ?

The broader questions I would have is, how do you want to "chunk" the text into TW, and then how do you want overlay or link commentaries? The smaller the level of chunking, the easier it is to comment on precise bits of text, but the larger the file size and the slower the TW will be.

I can imagine that you could show all the verses in a text and then have a highlighted asterisk after the text. Click on the asterisk and there would appear a note box that a student could record notes about that passage. A report tiddler would aggregate all the comments made. I'm guessing that that would be too inflexible.

Good luck,
Mark

[@TiddlyTweeter]

Ciao Lost Admin

I agree with you that beyond a certain scale native TW may not work for ePubs. In tests I did full scale novels (where the unit was 1 tid = 1 paragraph of novel of over 4,000 paras) became unwieldy. HOWEVER, it may be possible to optimize further. And TW has potential for authoring texts ready for ePub-ing inside a wrapper. I think there is definite potential.

Josiah

[@TiddlyTweeter]

Ciao Mark S.

I agree that "chunking" is the central issue on performance. It is my experience in testing that the more fine grained the chunking the slower the performance. And it gets worse the larger the file.

An interesting issue is, once the text is completed, is whether it would be possible to output the "ePub version" with optimized JavaScript dedicated to specific limited functions & not all of what TW normally does. It might be worth thinking on since ePubs are increasingly ubiquitous.

Josiah

[Steven Schneider]

Hi folks - I made a major faux pas in the group by cc'ing this message to others -- I meant to forward it to them. If you could, please take care not to "reply all" so we don't flood their mailbox. My apologies to all. //steve.

[Mark S.]

I've never seen an epub with javascript ability. I know PDFs can have some ability, but I don't think I've ever seen an example of one. The problem is that javascript could be used in theory to hack a system, so adoption has been slow.

Each tiddler adds 100 to 200 bytes of overhead, depending on additional fields and tags.

That book you mentioned must have been huge. Unless my math is off, or your paragraphs are really small, it could be 10 to 20 megs. Even Moby Dick is only 1.2 megs

Text compresses really well. Thinking out loud, I wonder if there could be a Tiddler Template to decompress text. Then huge books could be packed into a TW. You would view the content through a viewer page that would display the content normally. 

-- Mark

[Steven Schneider]

On Tuesday, May 9, 2017 at 10:02:13 AM UTC-4, Mark S. wrote:

I think he's saying that he wants to convert existing texts into TW.

Yes - that is exactly what I am saying.  Here is an example: http://open.lib.umn.edu/americangovernment/

How would I begin to import this into TW?

//steve.

[@TiddlyTweeter]

Ciao Steve

Its also an issue using Google Groups on web. As a web user of it I'm not really thinking its also an email list. Looking at it I see there is a CC field I should uncheck.

BUT the problem is, you as the author maybe want copies to go to those people, so how would I know when to UNcheck it?

C'est la vie. I will now :-)

Best wishes
Josiah

[@TiddlyTweeter]

Ciao Mark

Mark S. wrote:

... your paragraphs are really small,

Absolutely right. Novels that have lots of dialogue like ...

"Did you eat the sandwich?", she said

"Are you questioning my motives?", he replied

She responded, "No, just whether you liked cheese with pickle."

... etc

 
End up as vast numbers of Tiddlers if para chunked.

Josiah

[Steven Schneider]

OK, then, here is a workflow to go from epub -> html -> tiddlywiki.

(1) I download the epub version of http://open.lib.umn.edu/americangovernment/

(2) Use pandoc to convert to html (http://pandoc.org/)

pandoc -f epub -t html American-Government-and-Politics-in-the-Information-Age-1479837299._3.epub -o Amgov.html

(3) Import Amgov.html to a text-slicer edition of TiddlyWiki  http://tiddlywiki.com/editions/text-slicer/

(4) The results are here:

http://american-government-in-the-information-age.tiddlyspot.com/

Pretty good for 5 minutes of work - not counting the time to download pandoc & teach myself to use it :)

(5) Next steps: work on displaying TW tiddlers in an organized fashion using the list fields. That should be relatively straightforward. I'll report back to this thread soon...

//steve.

[Mark S.]

There are multiple approaches, depending on how you want things chunked. Chunking matters because your students will probably be commenting at whatever "chunk" level you provide. So if you want paragraph-by-paragraph chunking, then you're going to need to chunk things into single-paragraph level tiddler chunks.

The more chunks, the more overhead, which is why it becomes tempting to chunk at the chapter or section level.

One way to proceed would be to install tiddlyclip. Then go through the original text and clip the text you wanted piece by piece. You could do this as plain text or as converted TW markup text. I don't know if tiddlyclip can clip the original HTML or not -- I'd have to study the docs. The problems with this approach is that you would need to tag or add fields to your resulting tiddlers so that you could link a series of paragraphs up together as chapters and sections. So, it's labor intensive.

Another approach is to drag and drop chapters (or selected paragraphs) into your new TiddlyWiki and then import the text. This brings in the original HTML formatting so it looks like the original text but the actual text is harder to edit and it's going to be a little bit bigger (because the text has HTML tags in it). You will want to tweak the HTML to remove the next/back links (either eliminate them or point them at the appropriate tiddler). Of course, you probably won't need to edit the text after this tweaking so HTMN might be all right. You will need to add your own titles to each tiddler chunk you import. You will probably want to either come up with titles that will allow the entire text to be sorted by title and/or add fields and tags that will allow the text to be reassembled as one document.

Another approach would be to find the original HTML source text. I think the epub could be decompiled to do this. Put all the source HTML into one big HTML file and then convert it by a series of replacements into TW markup. Then paste the whole thing at once into a giant tiddler. Jeremy has a slicer tool/plugin that can then break out the text into smaller tiddlers of connected text. This last approach may actually be the easiest in the long run.

As you can see, there's a bunch of things to think about. What size chunking? How much of the original formatting? How to connect the pieces together?

HTH
Mark

[Lost Admin]

On Tuesday, May 9, 2017 at 10:34:18 AM UTC-4, Mark S. wrote:

I've never seen an epub with javascript ability. I know PDFs can have some ability, but I don't think I've ever seen an example of one. The problem is that javascript could be used in theory to hack a system, so adoption has been slow.

As I understand it, EPUB version 3 is supposed to support some sort of interactivity through javascript. However, I don't know how many ebook readers that support epubs actually fully support version 3.  

[Mark S.]

Wow - amazing. While I was typing up some approaches, you already did it!

[@TiddlyTweeter]

Ciao Mark & Steve

I have two comments ...

1 - CHUNKING is also about the semantics of the original text. In novels the working optimal seems to be the paragraph. Though in some cases (Like Finnegan's Wake) it might be the sentence. For poetry its the line. I think its a mistake to allow practical workability to override the meaning units needed. Of course, there are cases where chapter units might work. It very much depends on the nature of the work. And on the purpose of doing it.

2 - HOW to CHUNK is I think a very important and interesting issue. Personally I use a Regular Expression engine that lets me input a vanilla text and directly output fully formed Tiddlers. Of course I have to put hours into getting it to work. But the output requires me to do nothing. Its already fully chunked and ordered with embedded fields. I don't even JSON. I just inject it directly into a TW.

Best wishes
Josiah

Steve wrote:

OK, then, here is a workflow to go from epub -> html -> tiddlywiki.

[Mark S.]

I didn't realize that Jeremy's slicer could handle HTML directly. That's a real time saver.

I see that some formatting is lost. In particular, the <em> tag used in the preamble of chapter 1. This might mean that other formatting (like bold) might also be lost in some places.

It appears to me that you might want to start over with a different starting tiddler name, since "Sliced up Amgov.htm..." appears in every tiddler and in every link. I can imagine how students would respond to that.

I think if it was me I'd want some way to get rid of the pink text.

Have fun!
Mark

On Tuesday, May 9, 2017 at 8:20:43 AM UTC-7, Steven Schneider wrote:

[@TiddlyTweeter]

Jeremy's Slicer is excellent for simple cases. It requires certain criteria to be met to function well.

One thing about PRE-PROCESSORS is how much of the heavy lifting they do or not.

Its all very well to chunk some vast document into Tiddler units. But what also matters is what they are capable of for generating organizational metadata.

I'm personally very interested in how smart pre-processing could be. My aim being to absolutely reduce to as low a level as possible later manual editing.

Josiah

[@TiddlyTweeter]

There is a fairly good overview of Interactivity on EPUB 3 here: http://epubzone.org/news/epub-3-and-interactivity

I think the point is that its not gonna universally embrace freed JS. That does not stop US developing a different approach that approximates them AND adds other functions.

A sort of honed TW-Pub with better performance at scale than TW native. Or a neat way to produce ePubs that then get wrapped conventionally. Or both.

Just thoughts
Josiah

[Steven Schneider]

Got a chance to do a bit more on the epub > html > tiddlywiki workflow.  

See http://american-government-imported-text.tiddlyspot.com/ which imports all tiddlers matching filter from  http://american-government-in-the-information-age.tiddlyspot.com/, which used text-slicer to parse the html of an epub. I then transclude all tiddlers using a series of <$list> commands.  Not bad for a quick import of epub. Work to be done, but proof of concept, at least.

//steve.

[Stephen Wilson]

To bring the thread back a little to where it started, I would be interested in developing tiddlywiki for education.
I would also be interested in looking at ways to export tiddlers easily into a read only format.
And finally, I would definitely be interested in looking at systems to either generate questions/ mark questions and give feedback automatically.

My use case is for engineering at level 3 (A level/ Pre University stuff).  My tiddlywiki heavily nested eg : http://stephenteacher.tiddlyspot.com/#Statics  I would need a simple way to parse that tree, collect the tiddlers and assemble them in order.
On the exercise side I suspect my needs are a little different as ideal I need maths focused solutions.  Essentials would be random question generation eg the question is a+b=? but a and b can be any number from 1 to 10 so each student gets a slightly different question.  Feedback from the student answer would then be good.  The recent if plugin springs to mind as something which could be leveraged to this.    There also needs to be some way for students to own their work.
I have been curiously looking at http://webwork.maa.org/ and http://www.u-psud.fr/fr/universite/organisation-generale/services/direction-de-l-innovation-pedagogique/utiliser-wims.html (Google translate on Chrome to the rescue) for this functionality but have not had the time to install and play with them.  I have also looked at doing similar in tiddlyspot using the matthcell plugin http://stephenteacher.tiddlyspot.com/#Interactive%20Steps%20to%20Solving%20Problems
So that's where I'm at and what I'd like to do educationally.

Stephen Wilson

stephenteacher.tiddlyspot.com/

[Lost Admin]

I'm curious about something... your student base is smart and technically savvy (based on your statement of them being pre-university engineering students), how can you possibly prevent them from gaming the system if you use tiddlywiki? They will have full access to all the code (all the logic is in JavaScript that runs on the browser),

[Stephen Wilson]

That is a very good question.......I guess I hadn't really thought that far ahead :)

At least it would take a little more effort to try and game the system than simply submitting a friends assignment and hoping I don't notice.

They are less tech savvy than you think...

Stephen

[Mark S.]

Owning the work. You could distribute to each student their own password. Before uploading the assignment, they would encrypt using the password you provided. This would be effectively like "signing" the document. The documents could even share the same open file folder safely. Not as sophisticated as GPG, but should work unless students share their passwords or have trouble copying/pasting them.

Mark

[Lost Admin]

On Wednesday, May 10, 2017 at 9:33:57 AM UTC-4, Stephen Wilson wrote:

That is a very good question.......I guess I hadn't really thought that far ahead :)

At least it would take a little more effort to try and game the system than simply submitting a friends assignment and hoping I don't notice.

They are less tech savvy than you think...

Stephen

That is disappointing.

[Steven Schneider]

Stephen, 

First, apologies for cc on initial message.

Second, thanks for the response. I suppose if we get 4-5-6 folks interested in teaching / education / making resources, we can spin off into another group to talk? But for now, stick with this thread....

Responses inline below:

On Wednesday, May 10, 2017 at 5:46:24 AM UTC-4, Stephen Wilson wrote:

To bring the thread back a little to where it started, I would be interested in developing tiddlywiki for education.
I would also be interested in looking at ways to export tiddlers easily into a read only format.

I think this is best done using node.js, the way that tiddlywiki.com static is presented. I'm just on the intro side of learning node.js, but it seems necessary for anything along these lines.

 

And finally, I would definitely be interested in looking at systems to either generate questions/ mark questions and give feedback automatically.

YES!! But feedback/grading gets tricky. It used to be more possible in the tiddlyspace days, but TW Federation is not super easy (again, probably requires node.js). One way I've done this is to use google forms for student submissions. For example, https://sunypoly-com216.updog.co/syllabus.html#Reflections%3A%20Chapter%202 asks students to submit their thoughts in a google form, which I then import (using xlsx import tool; see demo) to create this: https://sunypoly-com216.updog.co/syllabus.html#Chapter%202%20Reflections. Grading could be done in the results spreadsheet...

I've also created a google form in which students "voted" using pre-formed google forms to record their submissions

My use case is for engineering at level 3 (A level/ Pre University stuff).  My tiddlywiki heavily nested eg : http://stephenteacher.tiddlyspot.com/#Statics  I would need a simple way to parse that tree, collect the tiddlers and assemble them in order.

I think that we can do that already using toc filter http://tobibeer.github.io/tw5-plugins/#toc

On the exercise side I suspect my needs are a little different as ideal I need maths focused solutions.  Essentials would be random question generation eg the question is a+b=? but a and b can be any number from 1 to 10 so each student gets a slightly different question.Feedback from the student answer would then be good.  The recent if plugin springs to mind as something which could be leveraged to this.   

I think you are referring to Thomas Elmiger IfAisB work - if so, this might be the approach what we need. 

 

There also needs to be some way for students to own their work.

 

Well, yes, but let's not miss the opportunity to do half of what we need. And, I believe, that as Xememex moves forward, that opportunity will present itself.

So that's where I'm at and what I'd like to do educationally.

Very exciting! Thanks for the response. 

//steve.

 

Stephen Wilson

stephenteacher.tiddlyspot.com/

[PMario]

On Wednesday, May 10, 2017 at 3:33:57 PM UTC+2, Stephen Wilson wrote:

That is a very good question.......I guess I hadn't really thought that far ahead :)

At least it would take a little more effort to try and game the system than simply submitting a friends assignment and hoping I don't notice.

They are less tech savvy than you think...

If they trick the system, they have to dig relatively deep. .. So you can use this "assignment work" as the base for a grade :)) ... once

-m

[Handoko Suwono]

nice discussion here. I am posting to follow the updates. I am not an educator, however been a mentor to google code-in which provides TiddlyWiki some intro tasks for teens pre-university students. And perhaps doing it in some workshops in conferences. So it's good to know more.

handoko -

On Tuesday, May 9, 2017 at 7:00:17 AM UTC+7, Steven Schneider wrote:

There seems to be some interest in this group in using TiddlyWiki as a platform for teaching, and for delivering educational resources such as textbooks and presentations. I thought we might use this thread to identify those interested in this kind of work, and see if there is critical mass to spin off a few collaborative projects, or at least to bring together some existing resources. 

So, here goes:

[@TiddlyTweeter]

The context of teaching & "cheating" is interesting. That TW could be cheated by someone who knew what they were doing is a plus in my view :-). A+ for smarts. Though B- for not playing fair:-)

On a more serious note, how you could solve this without a server solution? There is a point at which TW, as currently implemented, isn't really anywhere an elegant solution.

I am very interested in the potential of Danielo's TW version, NOTESELF, that (1) saves to browser storage, not files; (2) can co-ordinate its running on multiple devices via a Cloudant server. I think its a very interesting approach that might be leveraged towards multi-user approaches and passworded levels of authentication. Danielo created it for himself, not for explicit multi-user situations, but both the Cloudant (server database) and PouchDB (local storage database) have built in potential in the required direction. Its very interesting because its keeps the full local ownership of TW, yet combines it with a remote database. It runs with or without the remote database.

My 2 cents
Josiah

Stephen Wilson wrote...

... it would take a little more effort to try and game the system than simply submitting a friends assignment and hoping I don't notice.

PMario replied ...