Query: What is a fragment?

[TiddlyTweeter]

TW "philosophy" often refers to "fragments". They are the lowest sustainable units out of which "kosher" TW are built. 

But there is an issue. WHAT actually is the size/scope of the resolved/composed input chunks?

This is where it gets interesting. Is that unit a semantic whole? Or merely a (e.g. phrasal) sub-part of it?

"Fragment" looks trivial at first. But as soon as you introduce PURPOSE it becomes less stably minimal.

Anyone has thoughts on this?

Do ask if the query is unclear. I can give examples, though rather hoped you'd "get" the issue without.

TT

[TonyM]

TT is this fragment not the tiddler?

since the tiddler title is a unique key then the tiddler is a record, contains columns / fields. The only difference is the text field especialy can contain a rich functionality and perhaps there we can have fragments of macros and transclusions.

I am not sure where else a fragment could be.

Regards
Tony

[TiddlyTweeter]

Tx TonyM

Right. Tiddlers ARE the fragments.

But the OP is not about that per se.

It is about WHAT the size and scope of a "fragment / Tiddler is".

HOW do you decide that?

And in what ways does it relate to PURPOSE?

TT

[Mat]

My thoughts:

https://groups.google.com/d/msg/tiddlywiki/s9Y_w85282I/ZsIToC88AAAJ

<:-)

[TiddlyTweeter]

Mat wrote:

https://groups.google.com/d/msg/tiddlywiki/s9Y_w85282I/ZsIToC88AAAJ

this ...

(me) HOW small is a good fragment? And how would you know?

(mat) I can answer this... in a more general way but also with an exact answer:
A tiddler, i.e the smallest semantically meaningful bit of information, should be EXACTLY as small as the context demands for it to be meaningful.
There is not reason to split up, say, the huge Encyclopedia Britannica tiddler if one never needs any subpart of it (and if the system can handle such a big tiddler).
And it is pointless to have a tiddler for each ingredient in your pancake recipe if those tiddlers are never used in any other context. 

 Apart from fact you more aware than me I asked a related question afore, very good answer fit to the issue.

What I think interesting. Very muted in previous notes on this. Is the enormous role of ...

1 -- cognition in knowing IN ADVANCE. It may be a mystery left as such. But HOW do you know what is the right approach on chunking?

2 -- I think the main thing you, Mat, flag is CONTEXT. That matches my experience. Yet knowing that context is not trivial to proving.

Is "FRAGMENT" scope ONLY intuitively known, rather than formally definable? If so .... then ...?

Asking.

TT

 

[Mat]

TiddlyTweeter wrote:

Is "FRAGMENT" scope ONLY intuitively known, rather than formally definable? If so .... then ...?

...then the important thing is that we have a system that allows the data to change along with the changes of our feeble minds. To merge or split the fragments as we see fit. IMO TW does this better than any other software I know of BUT there is definite room for experimentation and improvement. For example, it would be cool if we could drag'n drop to merge tiddlers. And if the excision functionality was more accessible.

<:-)

[bimlas]

Similar thread: https://forum.zettelkasten.de/discussion/comment/5842

PS.: The Zettelkasten community is as philosophical as TiddlyWiki, maybe we could learn something from each other.

[TonyM]

TT

How do I decide? For me its based on decades of experience in IT with a conceptual approach to understanding and negotiating my way through complexity.

Although I actively try and find ways to make my implicit knowledge explicit. It is difficult to communicate with people of many different experiences and skills. I have a consulting business called Interpreting IT to back this up.

To be frank one way to decide is to pay for professional help from someone with subject matter expertise. Never the less I continue to give freely and try and answer this very kind of question.

Not withstanding the fact that experience can answer these questions it is quite easy to "rebase", reconfigure or redesign the way data and processing is organised in tiddlywiki using tiddlywiki itself so we are free to design then redesign.

One of the meta concepts I am developing on tiddlywiki is rapid and reiterative design methods and as a result I am considering researching and writing a book with the working title "Occams electric shaver" a computer equivalent of "Occams razor", and this shows we must build then simplify repeatedly, abstracting many times. In some ways this idea shows how the answer to your question is a process rather than a rule.

So fragments will be the size they need to be depending on how mature the design is and in tiddly wiki this keeps returning to the tiddler.

I hope this makes sense.
Regards
Tony

[TiddlyTweeter]

Ciao bimlas, I am aware you think about these kinds of issues too. What is information? How can I divide and organize it?

You may find this post I made on Zettelkasten interesting https://groups.google.com/d/msg/tiddlywiki/pCcWB2zVD40/3YILA5jsBQAJ.

Zettelkasten was originally a MANUAL system. 

I think there is a lot of merit in deriving systems later from previous manual systems. 

Not least because they are cognitively graspable. Luhmann was a genius of manual organisation.

Best wishes

TT

[TiddlyTweeter]

Ciao TonyM ...

How do I decide? For me its based on decades of experience in IT with a conceptual approach to understanding and negotiating my way through complexity.

Although I actively try and find ways to make my implicit knowledge explicit.

Very good note. I picked out two great points to underline ...

 

... meta concepts I am developing on tiddlywiki is rapid and reiterative design methods

Right. Reiteration. Recursive. Self-aware.

My point is this: TW is quite unusual in actually reasonably supporting human cyclic cognitive process.

Of course NO computer yet can work as well on the liminal that humans constantly sense, build, and act from. 

But I'd guess the "fragment/tiddler" is quite close. 

It encourages appropriate reduction to salient partitions.

... the answer to your question is a process rather than a rule.

Right. The heart parts are "fragment" yet "fragment" is not what it might first look like. 

Good fragments are tempered by your mind working in context for purpose.

Thanks!

TT

[TonyM]

TT

Your welcome, I appreciate yout raising conceptual and philosophical issues about tiddlywiki.

Thanks

Tony

[TiddlyTweeter]

TiddlyTweeter wrote:

Is "FRAGMENT" scope ONLY intuitively known, rather than formally definable? If so .... then ...?

 

Mat responded ... 

...then the important thing is that we have a system that allows the data to change along with the changes of our feeble minds. To merge or split the fragments as we see fit. IMO TW does this better than any other software I know of BUT there is definite room for experimentation and improvement. For example, it would be cool if we could drag'n drop to merge tiddlers. And if the excision functionality was more accessible.

That's interesting. "More ways to re-chunk" ?? Sounds right, to make data and concepts of that data more explicitly re-chunkable ?? That loop, to re-do fragment size & scope, I think fits human meaning making process better than idea of "getting it right"--which completely overlooks we are inherently iterative.

FWIW, In my own case, I am very interested in facilitating associative thinking. For example have a wiki of thousands of phrases (sub sentence) that you can randomise and then transclude and save interesting combos of. Repeat at will.

Basically, a William Burroughs "cut-up" machine. That use of random "fragments" supports associative cognition and pattern (new story) recognition in a helpful low overhead way.

Thoughts

TT

[TonyM]

Hay, I am all for chunking, but we must realise moving a chunk is important, rather than copying it, Chunks need to keep their name, if we do make copies, or we can loose data with duplicates.

Another point is combining chunks can remain virtual with chunks remaining themself. We could find a way to do sections like in TWC to access chunks within tiddlers, but this is a slippery slope.

So things can be chunked until they can be chunked no more, then can be collected and ordered to combine them virtually, but avoid physically combining them unless there is an identified reason to de-chunk.

This can be compared to Analysis and Synthesis but in this case synthesis is virtual and flexible. We never loose the gains of analysis.The chucks we find. Careful chunking will always record what we divided, to do the chunking and maintain this information in the system. 

Regards

Tony

[springer]

On fragments/chunks/tiddlers philosophy,

I entirely endorse Tony's point that one has to think dynamically about this. As a project develops, not only one's familiarity with the subject but also the actual work being done may shift scope. 

Let me also just point out, for this discussion, that sometimes the author/builder of a site will be working at a level that is more fine-grained than the reader/consumer of a site.

Still, it's helpful for me to start with a benchmark. Motivated by my work in studying Japanese, I want to suggest that there's something like a "sentence"-level unit below which one can have reference, but not competent assertion. And focusing on smaller bits can lead to bad habits. (My study has settled in with a brilliant zero-translation SRS program focused on whole model sentences, with gradual accretions in the vocabulary covered.)

Switching back to TW, we can ask what are the sentence-like (assertible/endorsable) units for the project at hand, and treat those as natural starting-points. (In practice, it may be more like a paragraph, but a paragraph with a single sentence-like point at issue.) That doesn't mean that there's no point in having tiddlers for, say, a single word (like a noun), but then the function of that tiddler will be something like "This noun's function and connections are like so."

One great use of the dynamic tables plugins is that when I realize that a set of tiddlers has gotten "overgrown" (carrying content beyond the "sentence"-worth that has emergent metadata structure to it), I can more easily set up a field, move content over into the field, and feel confident that the shift has been a clean one. 

Still, when I go to present information to students, I often want to emphasize various compound tiddlers... and though each one of these still has a single "sentence"-like point, the compound tiddlers (usually template and/or TOC-srtuctured, in my case) put the most pedagogically useful supporting bits in view as well.

-Springer

[TonyM]

Folks,

Give splitregexp[\n] try below code fragment on tiddlywiki.com to see how lines can be read as fragments.

<$list filter="HelloThere">
   <$list filter="[all[current]get[text]splitregexp[\n]]" variable=abstract>
       <$link to=<<currentTiddler>> ><$text text=<<abstract>>/></$link><hr>
   </$list>
</$list>

Using the recent xml tools adds even more possibilities, including finding a way to extract html sections that can be address elsewhere.

A more advanced version of the above in my tiddlers contain a comment section 

<!--
Abstract: A abstract about the content of this tiddler
-->

And the filter I use in the above code is

[all[current]get[text]splitregexp[\n]prefix[Abstract:]]

So I am fragmenting content as if I through a grenade into it.

The prefix operator is a friend here; it could tell the difference between, https:// and file:// and anything else that may be present in a line

Regards

Tony

[bimlas]

Unfortunately, a lot of things are still obscure about Zettelkasten, but I’m very interested in the topic because basically it can be applied regardless of device: any note-taking app, wiki, even an issue tracker, and of course it works on paper.

I started working on a TiddlyWiki Zettelkasten edition, but it’s pretty hard because in Tiddly it’s hard to handle tiddlers named with a Unified Identifier (for example, the tag popup list becomes unusable). Is there any example TW from which ideas could be gained?

[TonyM]

Bimlas,

Do you mean when you say "Unified Identifier" are you talking about a unique ID? I found an algorithm and way to implement in tiddlywiki so any tiddler can be referenced by its Unique ID regardless of its title.

Just ask if that is what you are looking for.

Regards

Tony

[TiddlyTweeter]

bimlas wrote:

Unfortunately, a lot of things are still obscure about Zettelkasten, but I’m very interested in the topic because basically it can be applied regardless of device: any note-taking app, wiki, even an issue tracker, and of course it works on paper.

Right. Especially if you can't read German. But some of the "concepts" the Zettelkasten community use in discussion are very, very good. This is likely because of the exceptional ability in German to combine words to easily create exact new concepts. The discussions are able to differentiate many variant ways of linking, for instance, accurately & without ambiguity.

Zettelkasten is, essentially, a "theory" of connection building with a strong commitment to emergent meaning, yet, at the same time a steadfast commitment to each item, on creation, being UID'd forever, yet with no limiting hierarchy.

The several implementations in software do vary. Personally I find understanding Luhmann's original manual method extremely helpful in dealing with the TW relevant issue of "fragment evolution to new wholes". For instance: https://takingnotenow.blogspot.com/2007/12/luhmanns-zettelkasten.html and https://niklas-luhmann-archiv.de/nachlass/zettelkasten (Google translate will give a good enough translation).

Best wishes

TT

[TiddlyTweeter]

bimlas wrote:

I started working on a TiddlyWiki Zettelkasten edition, but it’s pretty hard because in Tiddly it’s hard to handle tiddlers named with a Unified Identifier (for example, the tag popup list becomes unusable). Is there any example TW from which ideas could be gained?

Good point!

TonyM replied about idea of using a subsidiary UID he has worked on. Maybe a useful step.

But think your point in using the TITLE field as a unique identifier is right, as that is what it IS, built in. There are several ways of thinking about it. To give an example of a possible Luhmann style indexing that auto disambiguates for "Insomnia" ...

467-1a INSOMNIA (sleep disorder)

732-2c INSOMNIA , 1976 (film)

1034-1a INSOMNIA, 1982 (film)

The point is this example is the shown "title" would be in the caption field; the real Title is both a UID and human informative.

Now, to your real QUESTION: "the tag popup list becomes unusable"

Just wondering if, in the end, that popup could be modified to display a different field than "title".

The PROBLEM with going with an arbitrary UID (e.g. random nums; current date-time) is it has no semantic content.

Luhmann's approach was to have "titling" that is both a unique identifier AND has some human meaningfulness

Just thoughts

TT

[TiddlyTweeter]

springer

Excellent example of practically what occurs. This kind of material account of actual usage is invaluable imo.

Its just as important as technical understanding of how to do it.

Best wishes

TT

[Saq Imtiaz]

But think your point in using the TITLE field as a unique identifier is right, as that is what it IS, built in.

Using the title also has the advantage of uniqueness being guaranteed and quick lookups (in terms of performance) via filters, which is important when working with larger datasets.

Now, to your real QUESTION: "the tag popup list becomes unusable"

Just wondering if, in the end, that popup could be modified to display a different field than "title".

This has been my approach when using UUIDs for titles, display always uses the caption field if it exists, and otherwise defaults to the title. However, this is difficult to do as a plugin and yet ensure that it affects everywhere the tiddler titles may be displayed in TW.

Cheers,
Saq

[TonyM]

TT,

Tiddlers are already unique, but they need not be unique over time, if renamed. The relink plugin solves this for many but not all use cases.

That's why I believe using a seperate field to store a serialy issued unique number, issued if that field is empty on the first save, is best way to go where relink does not work.

As you point out you may need to "disambiguate" the tiddler title, but this will only occur if there are titles to disambiguate. So if we used INSOMNIA we would be fine until another need for INSOMNIA arises, at which time a new tiddler with a new serial number could exist, and because we do not want to override the existing INSOMNIA the automatic response is to create "INSOMNIA 1", thus it has being disambiguated, since the tiddler is tolerant of renames you can use a better disambiguation process at anytime.

Personally the way I create tiddlers always permits them to be renamed without consequence, it is the tags or fields already catagorising them, so a name change makes no difference. As soon as a disambiguation need arises I tend to build different tiddler "types" eg film/medical. If you do use a value in the title to do this, I recommend placing this in a field of its own (in addition to the updated title) so later you can programmatically interrogate the title, and subtract the disambiguation from it. 

We could even make a disambiguation tool. That store both the field and appends to the title. Adding it to the import process as well.

So relink can do this most of the time, but having serial number is ideal if one needs to associate many tiddlers to one, and a few other cases, we just provide store the serial number of the "parent" as an example. Then whatever the name the tiddler with that serial number will be found. No need for relink.

The only reason I have not published this is I have three solutions, one of which would need a core hackability improvement. To demonstrate the I need to have all four to demo.

Custom Actions on standard buttons

Automatic or selective unique serial numbers

An auto tag generation facility  ie given tiddler name there automatically exists a tagname (if not unused) for that tiddler

action tags to allow actions tiddlers to be reused in multiple cases.

This is my designer challenge to deal with, and attempt to seek assistance/a collaborator in the dev thread unfortunately got no where.

Serial number may also need to be disambiguated by wiki name to support interwiki transfers.

Regards

Tony

On Sunday, 24 May 2020 18:25:15 UTC+10, TiddlyTweeter wrote:

bimlas wrote:

I started working on a TiddlyWiki Zettelkasten edition, but it’s pretty hard because in Tiddly it’s hard to handle tiddlers named with a Unified Identifier (for example, the tag popup list becomes unusable). Is there any example TW from which ideas could be gained?

Good point!

TonyM replied about idea of using a subsidiary UID he has worked on. Maybe a useful step.

But think your point in using the TITLE field as a unique identifier is right, as that is what it IS, built in. There are several ways of thinking about it. To give an example of a possible Luhmann style indexing that auto disambiguates for "Insomnia" ...

467-1a INSOMNIA (sleep disorder)

732-2c IMSOMNIA , 1976 (film)

1034-1a INSOMNIA, 1982 (film)

[TiddlyTweeter]

TonyM wrote:

Tiddlers are already unique, but they need not be unique over time, if renamed.

Re the specific examples I gave are Luhmann naming. 

467-1a INSOMNIA (sleep disorder)
732-2c IMSOMNIA , 1976 (film)
1034-1a INSOMNIA, 1982 (film)

They are set once forever. Never renamed That was the use case. I maybe did not make that clear enough?

That schema presupposes good "feel" for the "field of endeavor". Such that the maker never goes back to alter titles.

The other situations you are exploring are interesting too.

The OP is broadly concerned with "fragment conception & handling". 

Use cases vary. That is central, I think, some cases you need to 

... redo actual content

others

... combine/list differently for changed ends

others, again

... never change.

Maybe the issue I started with should be reframed with END-NEEDS more central.

I was most interested in exploring how meaning making interacts with tech systems in different ways, I guess. 

All three scenarios are viable in TW. All three imply different mental processes, ontology and realization code.

Best wishes

TT

[Birthe C]

I know this is not zettelkasten, but it is collection of knowledge. I remember all the great work done by Alberto Molina. It is hard to believe that many years have passed since then. But he had to start over several times. Tags, type-field, other fields among other things. And how to put titles together. It changed from the beginning to the last incarnation, I believe it was http://magictabs.tiddlyspot.com/ and http://bottomtabs.tiddlyspot.com/ He obviously knew what he wanted and created beautiful tiddlywikies. But still not quite satisfied though.

It seems the titles always end up being the object of discussion.

Planning ahead for a life of study is not easy.

Luhmann found his system and stuck to it.

Really would it be any easier for everyone to take up his system totally and analogue, stick to it through life. I Doubt it.

Birthe

[PMario]

On Saturday, May 23, 2020 at 11:40:45 PM UTC+2, bimlas wrote:

...

I started working on a TiddlyWiki Zettelkasten edition, but it’s pretty hard because in Tiddly it’s hard to handle tiddlers named with a Unified Identifier (for example, the tag popup list becomes unusable).

Hi, .. I think the Zettelkasten system doesn't need or use tags. ... Tagging is a completely different concept, that wasn't used by Luhmann. ...

The second link TT posted has a very interesting picture at Chapter 3.1 "Umfang und Inhalt" (Scope and Content)

The paragraph ends with: "All drawers were without external labeling, which could have given information about the content of the respective drawer." (Google Translate is wrong here. ... excerpt -> drawer ... They are talking about the physical drawer made of wood.)

If a drawer would have a label, it would have been like tagging in our TW sense. .. But Luhmann's adressing system is based on "linking and backlinking" with a twist. ..

That is discussed at 3.4 Numbering Principle and 4. Conclusion

Especially conclusion is very interesting.

-mario

[PMario]

On Sunday, May 24, 2020 at 3:55:22 PM UTC+2, PMario wrote:

The second link TT posted has a very interesting picture at Chapter 3.1 "Umfang und Inhalt" (Scope and Content)

Link added. sry.

[bimlas]

Josiah,

I write these answers without reading the linked articles (I don’t have much time and writing the answers is quite time consuming with Google Translator).

 

467-1a INSOMNIA (sleep disorder)

732-2c INSOMNIA , 1976 (film)

1034-1a INSOMNIA, 1982 (film)

The point is this example is the shown "title" would be in the caption field; the real Title is both a UID and human informative.

While learning about Zettelkasten, I realized that I think it’s best if the notes don’t have a name, or at least treat it separately from the UID, because if I change its content, its name might give incorrect information about it. For example, if I write a “986 Naming Notes in Zettelkasten” note and then realize later that this methodology doesn’t just work for Zettelkasten, I’d like to rename it “986 Naming Notes”. Therefore, I would like to use 986 as the title (which is never changes), and "Naming Notes in Zettelkasten" would be just a caption.

 A short UID is also good because e.g. if I make a graph from my notes, their display will be uniform and not as space consuming as if they had a human informative title.

 

The PROBLEM with going with an arbitrary UID (e.g. random nums; current date-time) is it has no semantic content.

Luhmann's approach was to have "titling" that is both a unique identifier AND has some human meaningfulness

Even if a note has a name, in fact it will never be completely clear because its context determines what effect its content achieves. For example, if a note is titled "String," it says virtually nothing, because it could be related to programming as well as music. The context is mostly given by the text of the link pointing to it. If we just list the note names, it won’t be so clear why they’re included in a given context. Instead, you can use the text of the links to tell you how they relate to that topic. For example, if we only use their titles in an introduction to programming:

* 123 How to write CleanCode?

* 456 What makes the code stink, or what does the bad code smell like?

If we explain the reason for the connection with links:

* Examples of good coding techniques

* Signs that warn of bad code

Also, I don’t think it’s possible to give a title to every note because there’s something so abstract that it can’t be summed up in a few words. For example, give a title to your most adventurous dream. 

[bimlas]

PMario,

 

 Hi, .. I think the Zettelkasten system doesn't need or use tags. ... Tagging is a completely different concept, that wasn't used by Luhmann. ... 

Because of the implementation of Schlagwortregisterzettel (which I previously known as index zettel), I was thinking of using tags. In fact, of course, links can replace tags in this case.

I think the real problem that makes me unable to marry TiddlyWiki and Zettelkasten is that I still don’t quite understand how the latter works. It takes time for me to understand the theoretical knowledge I read about it through practice.

[TonyM]

Bimlas,

Even if a note has a name, in fact it will never be completely clear because its context determines what effect its content achieves. For example, if a note is titled "String," it says virtually nothing, because it could be related to programming as well as music. The context is mostly given by the text of the link pointing to it. If we just list the note names, it won’t be so clear why they’re included in a given context. Instead, you can use the text of the links to tell you how they relate to that topic. For example, if we only use their titles in an introduction to programming

This context sensitivity is why the links in relationships are best independent from just the title, On the second use of "string" you "discover" the need for a logical disambiguation that arose from one of the contexts in which you found a connection with string. At this point you may split and link to a different string "musical string" and rename the other "text string", or you may maintain the abstract "string" with subtiddlers for alternative definitions, this is uncovering or discovering information and relationships in the data set you are collecting. Except perhaps as a definition of the word "string", I may be unlikely to use this as a title without disambiguating from the start. 

For example, If I created a tiddler "string" to provide the definition of the word, I would realize there is now a set of relationships between word use and their definitions. So I would more likely disambiguate by a namespace $:/glossary/string providing a new context that of definition. his is all a matter of a "process of discovery", thus the underlying system must be capable of adaption.

For the system to be remain adaptive we must;

• Avoid Dependencies on compound keys
• Maintain maximum renamability (Relink or Serial Numbers)
• Maximise the readability of titles independently of how we organise tiddlers. 

If we deal with the context correctly we can use Captions to "re-ambiguate", so it appears as "string" but its link or tooltip is to a different "string".

This is an interesting discussion and two particular thoughts have arisen in my mind

• If we can analyse this enough and empower users to handle any relationships that appear, what a great feature it would be.
• Perhaps we could scan a Tiddlywiki and extract all relationships found and formalise these with a new tiddler for each, representing the relationship(s) even identify different contexts, needs for disambiguation and re-ambiguation etc...
• Could we design "ambiguous tiddlers" that internally disambiguate? eg String has the children "musical string" and "text string"
• "musical string" could store the context in which it is valid eg tiddler has music keyword.
• Discovery of ambiguity and abstractions and the reverse are important, let us capture and support them. 
• and many others

Regards

Tony

Regards

Tony

[TiddlyTweeter]

bimlas wrote:

I think the real problem that makes me unable to marry TiddlyWiki and Zettelkasten is that I still don’t quite understand how the latter works. It takes time for me to understand the theoretical knowledge I read about it through practice.

One of the reasons Luhmann can be difficult to grasp--and by extension Zettelkasten--is actually quite a lot that is implicit is going on (meaning a lot was in Luhmann's "head", not explicated). But Luhmann successfully used the boxes of slip notes to facilitate emergent knowing in an unusually efficient way.

Just a comment

TT

[bimlas]

I just realized that the "created" field can be used as a DTID and we can easily create links by a simple macro:

<$link to={{{ [created[20181122141438931]] }}}><$view tiddler={{{ [created[20181122141438931]] }}} field="caption"/></$link>

PS.: Because of the syntax, I simply can't convert it to a macro because I can't use the datetime macro parameter as a variable in the filter.

[bimlas]

Ok, it works, I just forgot to change "caption" to "title":

\define dtid-link(created) <$link to={{{ [created<__created__>] }}}><$view tiddler={{{ [created<__created__>] }}} field="title"/></$link>

This <<dtid-link "20131129090249275">> link is created based on `created` field. Try to rename the target tiddler!

Try it on TW.com!

[TiddlyTweeter]

​Birthe C wrote (layout changed by me):

​... ​all the great work done by Alberto Molina​ ... ​he had to start over several times.

Tags, type-field, other fields among other things.

And how to put titles together.

It changed from the beginning to the last incarnation, I believe it was http://magictabs.tiddlyspot.com/ and http://bottomtabs.tiddlyspot.com/ 

He obviously knew what he wanted and created beautiful tiddlywikies.

​Great example! Highly pertinent.​

​I'd add the finished beautiful, showcase too:​ http://tw5.scholars.tiddlyspot.com/#William%20Shakespeare

But still not quite satisfied though.

It seems the titles always end up being the object of discussio​n.

​I think it's unavoidable because of the centrality of Titling to ordering/access in both TW and scholarship​.

 

Planning ahead for a life of study is not easy. 

Luhmann found his system and stuck to it.

Really would it be any easier for everyone to take up his system totally and analogue, stick to it through life. I Doubt it.

​Right. But nonetheless interesting.

FWIW, when you dealing with resources external to a TW a kind of "stubborn materiality" (actual singular objects) remains important.

For instance cataloging art works​ and discussing them.

Or, at a slightly more abstract level, film appreciation where one may be interested in different "vectors" (title, genre, style etc., etc)

The Luhmann & Molina's strongly commit towards a solution finding their "practical contexts of meaning".

My point in the OP was to try make clear there is no one way. You can use TW like driving a car made of fragments. And also build a new car that makes-up the fragments differently.

In an a way its obvious. But its not maybe as explicit as it could be that we are overwhelmingly "wet-ware" first :-)

Thoughts

TT

[TonyM]

Bimlas,

Nice simple method. I considered this at one point and did not proceeded with it for a few reasons that I recall.

• I did not know how to guarantee uniqueness, especially for imported tiddlers
• Cloning tiddlers can defeat this with duplicates
• I noticed many system tiddlers are "created" field free.

I did actually intervene and on given tiddlers tested the created date was unique, and if not incremented only the milliseconds to make it unique.

• You can test for uniqueness on tiddlers that need this serial number, and I copied it to the TSN tiddler Serial Number field so I knew this test had being done.
• I decided if this were to be a potentially global serial number it would consume less space to start from 1 rather than in your example 20,131,129,090,249,275
• I even considered converting the generated serial number to a base 62 number system - 26 lower case, 26 upper case and 10 numeric to make the serial number even shorter
• Although I maintained next SN in decimal.

Regardless of how one sets the serial number a small set of tools would help, such as editor toolbar to select a link to a tiddler that uses the serial number.

Regards

Tony

[TiddlyTweeter]

Ha! Of course! That is elegant. And an "elusive obvious" found.

Best, TT

[TiddlyTweeter]

Ciao Bimlas

A few comments you do NOT need to reply to. 

I'm mainly commenting on mental processes prior to software tools. I think it's interesting, but not particularly helpful to you, now.

bimlas wrote (slightly edited):

 me ..

467-1a INSOMNIA (sleep disorder)

732-2c INSOMNIA , 1976 (film)

1034-1a INSOMNIA, 1982 (film)

The point is this example is the shown "title" would be in the caption field; the real Title is both a UID and human informative.

 

... I think it’s best if the notes don’t have a name, or at least treat it separately from the UID, because if I change its content, its name might give incorrect information about it.  if I write a “986 Naming Notes in Zettelkasten” note and then realize later that this methodology doesn’t ... work ... I’d like to rename it “986 Naming Notes”.

 

... 986 as the title (... never changes), and "Naming Notes in Zettelkasten" ... a caption.

Right. Makes great sense. Luhmann predated the PC so his approach was towards "don't revise, get it right at start" for both "index" & "title" and they are not clearly differentiated in his manual system. We have PCs so can hold them separately much easier now. Then they were not so separated.

  

 A short UID is also good because e.g. if I make a graph from my notes, their display will be uniform ...

Right. 

 

The PROBLEM with going with an arbitrary UID (e.g. random nums; current date-time) is it has no semantic content.

Luhmann's approach was to have "titling" that is both a unique identifier AND has some human meaningfulness

Even if a note has a name, in fact it will never be completely clear because its context determines what effect its content achieves.

For example, if a note is titled "String," it says virtually nothing, because it could be related to programming as well as music. The context is mostly given by the text of the link pointing to it. If we just list the note names, it won’t be so clear why they’re included in a given context.

That is a VERY interesting point. And strongly right in one way and maybe wrong in another. The point is Luhmann did not have a computer so his mental process was likely different than ours now. His computer was largely in his head. Nonetheless he was able to achieve massive scale integration.

MY point here is about that in "externalising the brain" to a program do we actually lose out? Get conceptually lazy? Neglect our inbuilt ability at pattern recognition?

I'm not expecting you to answer that :-). 

But Luhmann remains a great example of exceptionally astute use of analogue simply structured data for purpose. (AND anticipating computers.)

 

Also, I don’t think it’s possible to give a title to every note because there’s something so abstract that it can’t be summed up in a few words. For example, give a title to your most adventurous dream. 

Right.

Again it devolves to the scholar's context. Pre computer there was the general idea in social science scholarship that you don't write a note unless it has clear content. In other words, pre computer, writing had less "semantic gaps". With the advent of computers abstract "holding spaces" became much easier. At the same time it probably has also, made one aspect of scholarship weaker by over externalising the "brain" too early in a thought process.

Just comments

TT

[TiddlyTweeter]

bimlas wrote:

PMario,

 

 Hi, .. I think the Zettelkasten system doesn't need or use tags. ... Tagging is a completely different concept, that wasn't used by Luhmann. ... 

Because of the implementation of Schlagwortregisterzettel (which I previously known as index zettel), I was thinking of using tags. In fact, of course, links can replace tags in this case.

Couple of footnotes.

1 - It is true, especially for his second, vast, Zettelkasten, Luhmann had a substantial cross-cutting "key index" he added to on slip creation, and sometimes later added to. In many ways similar to modern "tags" (non-hierarchical indices) before they existed. But they were simply access indexes rather than "labels" (I.e. they existed in the hand written "index" but were not written on the actual record slips). Its sort of tagging, but "looking" only from outside the actual records.

 

2 - It is worth noting that some software implementations of Zettlekasten DO add and use tags for exactly the reasons you, bimlas, advance.

Just notes

TT

[PMario]

On Sunday, May 24, 2020 at 11:39:29 PM UTC+2, bimlas wrote:

I think the real problem that makes me unable to marry TiddlyWiki and Zettelkasten is that I still don’t quite understand how the latter works. It takes time for me to understand the theoretical knowledge I read about it through practice.

I think you are right. The system seems it has been developed over a big part of a lifetime of the developer. It was "learning / developing by doing"

The article mentions a card 9/8g: "(Zettel 9/8g): „man liest anders, wenn man auf die Möglichkeit der Verzettelung achtet“, which means "You read (books) differently if you have the "Verzettelung" in mind".

Where "Verzettelung" in this context means "indexing" as used in a database context.

I think, we should try to post a translated version of the article. ... BUT I don't know, if we would have a licensing problem. Will check it.

-m

[PMario]

On Monday, May 25, 2020 at 8:41:14 PM UTC+2, PMario wrote:

..

I think, we should try to post a translated version of the article. ... BUT I don't know, if we would have a licensing problem. Will check it.

OK it seems to be CC BY-NC-SA 4.0. ... So no real problem. I'll have a closer look. in a different thread.

-mario

[bimlas]

PMario,

As far as I can tell from a glance, this translation is similar to it. Or is it not quite the same?

https://sociologica.unibo.it/article/view/8350/8270

[TiddlyTweeter]

Bimlas & PMario

Both articles are very good. They overlap a bit. But they also differ in perspective quite a lot.

IMO they complement each other really well.

IF the German text were available in English along with access to the Sociologica article, between the two it would provide a rounded case study of, I think, of great interest.

TBH I found exploring Luhmann very enlightening.

Best wishes, TT

bimlas wrote:

As far as I can tell from a glance, this translation is similar to it. Or is it not quite the same?

https://sociologica.unibo.it/article/view/8350/8270

PMario wrote (modified by me) ...

... think, we should try to post a translated version of the article. ...

 

https://niklas-luhmann-archiv.de/nachlass/zettelkasten

[TiddlyTweeter]

bimlas wrote:

https://sociologica.unibo.it/article/view/8350/8270

That article is good. Very good for social scientists.

One thing in it that is well explained is how Luhmann, years before other sociologists, abandoned hierarchy based categories in favour of "surprising linked networks" (my term).

Luhmann's feeling for "linking" prefigures what came on the net. It was radical at the time. And, most important, it worked. In some ways maybe still ahead of what is actually happening.

In relation to TW I doubt Luhmann would like TOC's based on the HTML "li" construct, which is the narrowing tree structure he totally avoided :-).

TT

[TiddlyTweeter]

An issue in fragment handing is TERMINOLOGY. When does a part stop being a fragment?

IF I split a fragment its still a fragment. Even though we have two parts, THE SPLIT & THE SPLITEE.

TT

[TiddlyTweeter]

IF I join fragments (either literally, or through transclusion) how do I recognize its a shift towards whole? 

[HansWobbe]

I think the ultimate answer to "When does a part stop being a fragment?" may  be "when the fragment is Indivisible.".

Of course, that depends on the tools at one's disposal.  I recall Physics teachers explaining that Electrons,Neutrons and Protons were called "sub-atomic" part-icles because the prior generation though that the Atom was the smallest, indivisible object.  And as an engineer, I was taught that the difference between analog signals (messy Fourrier wholes)  and  digiat signals was that, in the digital realm, there were only 0 and 1.

Now, with Quantum Mechanics and Quantum Computing, I am only probably sure I can say "I think, therefore I am.".

In the scope and context of TiddlyWiki, I have come to appreciate that a Character is the smallest practical Part (glibly over-looking that there are 4 bits to a Nibble and 2 Nibbles to a Byte and as many as possible 2^31 characters in the UniCode character set). 

For me, that makes a unicode character the smallest possible fragment.  I am also relatively certain that, since a Character is Indivisible, it can only be divided by 1 or itself ... which is really neat since it means Characters are like Prime numbers!

[TonyM]

Hans/TT

I like the way you think. However to me the beauty of the "Fragment" is it is still a fragment all the way up and all the way down. In the past I have stripped a byte into 8 bits and used each bit as a flag. And whilst I do not know how to fragment bits I can emulate it by having each bit represented by a byte. Or in future we may be using Quantum bits.

And there are quarks as fragments of subatomic particles as well being subatomic particles.

When we join two fragments, we end up with one fragment, that is the point. The whole universe may only be a fragment of something else, we may never know.

Initially in this thread I suggested a tiddler was a fragment, it could be by convention, though lets call it only a tiddler, but the concept of fragments being a piece of something else is more conceptual and perhaps more useful. 

I have said before that the beauty of tiddlywiki is the tiddler, a record in database terms, composed of fields, with a unique title is at eye level, and the tools are all focused around tiddlers. Sure you can fragment the fields further, but the unique tiddler is always addressable.

So given quarks as a guide, where sub Atomic is a line above or below an atom, where everything else is,  perhaps we have Tiddler fragments, tiddlers, and meta-tiddlers would be the best way to classify fragments. 

Regards

Tony

[PMario]

On Thursday, June 4, 2020 at 3:18:26 AM UTC+2, TonyM wrote:

When we join two fragments, we end up with one fragment, that is the point. The whole universe may only be a fragment of something else, we may never know.

I do like the idea that 1 + 1 = 1 :) ... It's like a chemical reaction.

@Hans,

You scaled it down to quarks. ... Now let's scale it up to filaments and add some meta data like a tag ;)

-m

[HansWobbe]

Mario:

I have a Deeper understanding of things Quantum.  I see you suggestion of scale-up as a challenge to elevate my understanding of  StringTheory.  The last time I reviewed that, they had simplified it to only 11 dimensions to handle the "universe(s)".  I am optimistic however since I felt comfortable with the first 4 in the 1970s.  Now, in the 2020s, I am comfortable with models for the 5th an 6th.  I haven't done the math to know how many of 7, 8, 9, 10 and 11 that rate of progress let's me achieve. I keep getting distracted by TiddlyWiki's

N-DimensionalArray capabilities.  :-)

[TiddlyTweeter]

PMario & Hans

Mr friend Mr Darcy considers every fragment is part of a whole always. But sometimes he gets worried.

He wrote me "As I look at me left foot I know it is a fragment of me. But I'm concerned it has no idea what it is."

Just a comment from my friend.

TT

[Ste Wilson]

Interestingly if you create a hologram and then break the glass each fragment contains all of the hologram...

[Ste Wilson]

A fragment of a fractal would be an iteration of the whole.

[HansWobbe]

Please pass the KlineBottle. I need a drink!

[TonyM]

Ste,

Too true re fractals, But I think the fraction of a hologram is loosing some information, like its appearance from different direction?

Tony.

[Ste Wilson]

https://science.howstuffworks.com/hologram.htm

[TiddlyTweeter]

Hans, Ha!

A mathematician named Klein

Thought the Möbius band was divine.

     Said he: "If you glue

     The edges of two,

You'll get a weird bottle like mine."

Actually its a very interesting article. 

But how does it relate to the Partible, in the process of fragmenting? :-)

TT

[TiddlyTweeter]

Ciao Ste

 Ste Wilson wrote:

Interestingly if you create a hologram and then break the glass each fragment contains all of the hologram...

Interesting comment.

What I'm taking from this discussion is there is a vast richness in concepts of "fragmentation" ... They differ between fields of endeavor.

In relation to TW I, me myself and I, think we might look at a few concepts outside computer science to see if we can better pin down some notes on praxis.

FWIW, in branches of linguistics, its often observed that the "body" is not partible (linguistically). Sure there is the "naming of parts" but once your foot is cut-off is it still yours? Just an example.

TT

[TiddlyTweeter]

The whole thing with fragments in TW subsists in some kind of implicit understanding in The Maker that co-ordinates behavior.

IMO this video of a DUCK HELL MARCH encapsulates the issue. I really mean that. HOW does the fragment duck know where its going?

TT

[TonyM]

Ste,

Yes, reviewing the link, it all depend on what you call the whole hologram. If it is "all the 3d information" then dividing it is fragmenting, and loosing 3d data. If you mean a 2d image then there can be many found even after fragmenting the hologram. 

This goes well to my earlier points "However to me the beauty of the "Fragment" is it is still a fragment all the way up and all the way down", but what is it a fragment of is a matter of relativity and perspective?

This is I believe enough to turn fragments into a philosophical subject. Part of fractals, chaos and complexity theory amongst others.

Perhaps this is a value of tiddlywiki it has the power to take black and white deterministic software and allow it to grow into more than that, through emergence into relational, perspective driven, subjective objects, there by transcending the deterministic nature of computers. The subjectivity enters because of the observer being present and part of the system, and multiple observers may see different things, getting closer to an art work.

Regards

Tony

On Saturday, June 6, 2020 at 12:02:10 AM UTC+10, Ste Wilson wrote:

https://science.howstuffworks.com/hologram.htm

[TiddlyTweeter]

TonyM wrote:

... a value of tiddlywiki it has the power to take black and white deterministic software and allow it to grow into more than that, through emergence into relational, perspective driven, subjective objects, there by transcending the deterministic nature of computers. 

Broadly agree. But TW isn't transcending computers really. 

Rather, I'd say its "fragment model" is unusual.

Its common that software "unitises" data. 

What is not so common is that users can endlessly, easily, define and redefine what those units are and how they are combined.

And that "fragments" can be "semantic" and "structural", often both at the same time.

IMO, computer science does not theorise this possibility very well (I hasten to add there may be resources that do; I just can't find them).

So, in that sense, I'd say that TW, at it's best, is especially open to natural, intrinsic processes of thought.

TT

[TiddlyTweeter]

Hans

Has anyone made a wiki of single characters? One per Tiddler.

It would be an interesting experiment.

Let's do it.

TT

[HansWobbe]

I have several and would be happy to share.

I do think it's a bit of a bootstrapping exercise since we would have to agree on what a particular character means.  Fortunately, It has been my experience that this type of communications, within a TiddlyWiki wrapper, can scale up very quickly; so you and I really should do some tests.

I'll think about what a good starting point might be and post further before the weekend ends.

Thanks for expressing an interest in this.

Cheers,

Hans

[TiddlyTweeter]

I think its fascinating.

These kinds of experiment, even if they fail, are hugely informative!

And sometimes they are very surprising.

TT

[HansWobbe]

This is the one of the characters that I use most frequently.  I nominate it as a Candidate for the first OneCharacterWiki.  Obviously, I should at least export it as an JSON object, if only to avoid loading the poor thing into a standard  Empty TiddlyWiki file.  After all, it is supposed to be an isolated character and it would likely be totally over-whelmed by being placed in an Empty TW file.

🐟

As an added benefit of isolation, this little fellow will automatically be appropriately "socially distanced", almost like me!

Care to guess the meaning I am trying to impart?

[HansWobbe]

The single character as JSON

[springer]

Chiming in, philosophically, again:

The analogy with physical parts (atoms, quarks, etc.) may be misleading. While objects might in principle be always further divisible, information, in any practical context, is not so. 

Generally, a single character or sound is not an assertable, and building a wiki around them as "the smallest fragments" would be pretty silly. Just like "Among the things I learned by listening to the radio this morning is... [insert a single .2-second-long phoneme-sound recording]" You're not going to extract a bit of information out of those phonemes until you introduce a tiddler that strings them into an order; only THEN have you got an assertable. OR, you're actually putting a microscope on that sound (for info correlating with its resonance or waveform or transmission quality, etc.), and it's simply not a simple anymore. ;)

In other words, a single character can corresponds to an assertable in some circumstances: a typographer may want to track what's involved in rendering Q in Palatino, for example. But then the information at issue is much more fine-grained than what we normally mean by hailing the character Q as a minimal building-block (treating its many typographical shape-structures as artibrarily fungible instances). 

Similarly, a fish icon (or the 2-byte character for fish: 魚) might count as an inscrutable simple in some context, but I will want a whole tiddler for it exactly when getting that icon/character to render, or being able to reproduce its stroke-order, or understanding its compositional and historical relation to other characters, is a domain for more detailed information. 

If there's no assertable, there's nothing worth tiddling with. And as soon as something's assertible, it's not a conceptual simple anymore, because information involves synthesis. So it seems to me. I welcome a counterexample if you can think of one...

(Where this comes from, in my own thinking, emerges out of C S Peirce's logic and semiotics... with Peirce, along with his penpal Lady Welby, playing a significant role in the (pre-)history of computing...)

-Springer

[Eric Shulman]

Would that be meant to represent a https://en.wikipedia.org/wiki/Three-spined_stickleback (aka, Gasterosteus aculeatus), whose common name in Britain is the tiddler.

Tech note: the appearance of this character, https://emojipedia.org/fish/ (code point U+1F41F), varies depending on the platform/environment you are using:

   * "Apple", "Samsung", and "WhatsApp" versions look sad :(

   * "Facebook" and "Twitter" versions look shocked

   * "Messenger" and "Mozilla" versions look confused

   * "Emojidex" version looks bored

   * "JoyPixels version looks happy

-e

[HansWobbe]

@Eric: Well done!  That is indeed what I use this symbol for.  And your observation that it's rendering depends upon the whims of browsers is also correct.  As a result, I only use the symbol personally, as a way of indicating that any tiddler tagged with it, denotes a block of MicroContent that is also processed in one of several other technologies I need to use.  For other Audiences, I use other "symbols/words".  For some Audiences, a version of ...

That, of course, is largely due to my belief that "know your audience" is basically good advice to Authors.

@springer:  Thank you for your comment.  It stimulated a lot of thinking on my part. Perhaps a few more assertions and clarifications and ideally an example or two, will help set some boundaries that may be useful in establishing design considerations for a wiki.  I'll try to condense a few of those thoughts and post a brief summary later today.

[HansWobbe]

This link ( https://groups.google.com/d/msg/tiddlywiki/AZjiguV9DUU/NonZSOLuCAAJ )  is a start at an explanation that  letters are the (alphabetic) symbols used to form words that are a Special index into a Dictionary of meanings.  Generalizing this yields the realization that there are MANY more "nonWord" strings than there are Words (even in a rich language like Chinese).

It's interesting that https://en.wikipedia.org/wiki/Chinese_telegraph_code was effective as a way of using co-ordinates as an index into a dictionary of Words that conveyed meaningful information very efficiently.

Obviously, with TiddlyWiki's UTF-8 support, we can do very much better by building a vastly larger dictionary of accepted meanings.  This might initially just be thought of as yest another "jargon".

Perhaps, upon reflection, the use of English vocabulary is similar.  I have learned that newspapers are written using a very small subset of English words since few folk have mastered a significant subset of the unabridged Oxford dictionary.  Hence they simply consider their omitted words to be "meaningless" in the contexts of their interests.

[TonyM]

Hans

Looking at the telegraph code I recall a discussion on hexadecimal conversion led me to consider translating numbers, specifically decimal serial numbers to one with a large base formed from 0-9 a-z A-Z just to reduce the characters needed. If a number is unique in decimal it will be unique in base 62 only shorter.

Although in the discussion about fragments we are talking about a field or attribute of a tiddler. A serial number adheres to the rule.

Related to the key (the title) the whole key and nothing but the key.

Regards
Tony

[HansWobbe]

Tony:

I also like this method of reducing the number of characters needed, by increasing the base.  I started doing this in the 1970s using APL (A Programing Language) which made it very easy.  Since then, my "alphabet" has grown to its present size of about a thousand characters.

Since you are interested in in unique serial numbers; I generally strive to implement such a thing as a value that consists of combinations and permutations of the letters in my alphabet.  That effectively lets me use a value that is distinctly encoded with up to five of my dimensions at this time.

On Sunday, June 7, 2020 at 6:01:12 PM UTC-4, TonyM wrote:

...

Related to the key (the title) the whole key and nothing but the key. ...

 

[TonyM]

Hans,

When you say 5 dimensions do you mean in the same key?

That is inspirational, but what do you mean?

I feel we should have a way to register a unique id for any tiddlywiki we make, then that can be part of a compound key to get an universal serial wiki/tiddler. Would that be 2 dimensional?

Then we could have a third for our self eg wiki/my "brand"/tiddler?

Regards

Tony

[TiddlyTweeter]

springer wrote:

The analogy with physical parts (atoms, quarks, etc.) may be misleading. While objects might in principle be always further divisible, information, in any practical context, is not so. 

I agree. Human meaning making is context dependent with vast implicit processes occurring. Necessarily so.

I think what is interesting in TW is its "fragment model" is quite unusual. Because it goes far yet remains agnostic on what is scoped.

If there's no assertable, there's nothing worth tiddling with. And as soon as something's assertible, it's not a conceptual simple anymore, because information involves synthesis. So it seems to me. I welcome a counterexample if you can think of one...

Right. ASSERTIVE is an interesting word there that I think captures the sense a person is doing it FOR something that COHERES  for them. 

(Where this comes from, in my own thinking, emerges out of C S Peirce's logic and semiotics... with Peirce, along with his penpal Lady Welby, playing a significant role in the (pre-)history of computing...)

 (Peirce's Pragmatism was foundational in US philosophy for actually getting stuff done.)

TT

[TiddlyTweeter]

HansWobbe wrote:

This link ( https://groups.google.com/d/msg/tiddlywiki/AZjiguV9DUU/NonZSOLuCAAJ )  is a start at an explanation that  letters are the (alphabetic) symbols used to form words that are a Special index into a Dictionary of meanings.  Generalizing this yields the realization that there are MANY more "nonWord" strings than there are Words (even in a rich language like Chinese).

 "Zabbadabbadoo" (check spelling, please), as Fred Flintstone said ... https://www.youtube.com/watch?v=dptMRtOg__Y

NonWords are not them till they are :-)

HOW does that happen?

TT

[TonyM]

TT,

I would think many of lewis carols words in Jabberwocky have become real in the minds of those who can recite it.

https://interestingliterature.com/2016/01/a-short-analysis-of-jabberwocky-by-lewis-carroll/

The point being something "becomes" when it is seen, used, restated, published.

Clearly there are degrees of "becoming".

Regards

Tony