The power of Title. An experiment & a Q?

I did a test via JSON import of 570+ Tiddlers that had only Title fields … example (great for the use case which doesn’t need Text fields) …

[{"title":"with South Park because he hates Trump, and he's trying to expose through humor the hypocrisy and the meanness and the viciousness."},
{"title":"Trump does not find it funny."},
{"title":"Trust me, he hates it."},
{"title":"I'm Joanna Coles."},
{"title":"This is the Daily Beast podcast."},
{"title":"Who better to have on this week than Anthony Scaramucci
... etc

 [edited for readability]

Very interesting were the stats on that import (timestamps off) …

0.03 Mb in 576 tiddler titles

TW in default is Title centric so everything worked without additional fields.

Q: What I’m wondering if whether IN TW you can create new Tiddlers that are ONLY Titles? (i.e no body text, no tags, no timestamps, just a title field)

Just wondering
TT

Of course you can do this… not without considerable tinkering though.

The easiest step is to turn timestamps off, so your newly created tiddlers don’t get creation and modification fields.

You’d have to tweak the core, though, so that the creator and modifier field (tracking username as shown in control panel) would not be set.

If you’re doing this to shave time off of the import process, or for any other efficiency-related motivation, I’m skeptical that it’s worth the tradeoff in terms of real information lost (especially modification timestamp, which helps tremendously with troubleshooting).

On the other hand, you can also use a tool like the Commander plugin to strip these standard fields from a batch of tiddlers. I’ve occasionally done something like that, to streamline some bundle of tiddlers whose origins really don’t need tracking.

3 Likes

I’m going to push back on this just a bit… if you’re willing to use a custom interface rather than the standard tiddler editor, you can do this with a single tiddler and no core tinkering. And I’d hazard that if you’re creating title-only tiddlers (why? I can’t think of any reasons that wouldn’t be better served by a data tiddler, but perhaps you’ve got a plan) you probably don’t need the full editor experience anyway.

In any case, here’s a demo: Title-only tiddler generator.tid (936 Bytes) I forgot to include the how-to (such as it is) in the tiddler itself, but it’s intended to accept one tiddler title per line (no brackets needed).

And the raw code:

\define title-filter() [<currentTiddler>get[temp]splitregexp[\n]]
\procedure create()
<$list filter=<<title-filter>>>
<$action-createtiddler $basetitle={{!!title}}>
<$action-deletefield $tiddler=<<createTiddler-title>> $timestamp="no" created modified />
</$action-createtiddler>
</$list>
\end

\procedure delete() <$action-deletetiddler $filter=<<title-filter>> />
\procedure reset() <$action-deletefield temp />

<$edit-text field=temp tag=textarea class="tc-max-width" />
<$button actions=<<create>>>Create as title-only tiddlers</$button>
<$button actions=<<delete>>>Delete the above tiddlers</$button>
<$button actions=<<reset>>>Reset list</$button>

<$list filter=<<title-filter>>>

<$link />
<$list filter="[<currentTiddler>fields[]] -title" variable=field>

* <<field>>
</$list>

</$list>

EDIT: It wasn’t an issue in my testing on TW-com, but if you’ve filled in a username in your personal wiki, you’ll want to add creator modifier to the list of fields to be deleted — so it should read:

<$action-deletefield $tiddler=<<createTiddler-title>> $timestamp="no" creator modifier created modified />
1 Like

Eek!

Q: If I keep the username empty would that prevent creation of that field?

Ah. The particular data needs no tracking at all.
And my motivation is to reduce the number of not needed bytes, not time.

FWIW, the final aim is I want to do Burroughs style cut-ups in TW with thousands of titles.
This is a step towards that, I hope.

Thanks for your comments. I will look at @Mohammad’s Commander to see if it can help.
TT

1 Like

I’ve done a lot of tiddler-generation from external sources. For no obvious reason, I used to include created and modified timestamps and ensured that they were generated in an appropriate order. It took @pmario pointing out that I was simply wasting space for no reason for me to stop this behavior.

Over a lot of tiddlers, these things do add up.

2 Likes

Right. For the purpose the editor is totally unnecessary.

Thank you for the tool! Very much!

I tested it. It did a test job no problem with very long titles …

TT

1 Like

is this not an ideal case for missing tiddlers?, importing the titles then making a small process to format them as titles and join them with a space placing them in a tiddler, eg “short phrases”?

  • then use them as missing tiddlers
  • delete the imported titles
  • use [[short phrases]links[]] to list or search.

this way the short phrases tiddler stores the metadata for all the titles imported, it’s created modified details, tags etc…

the link editor toolbar button could be modified to include missing tiddler titles so you can insert any of these titles in your tiddler with ctrl-L

coincidentally if they were imported as tiddlers they would be my recently discussed “nascent tiddlers”.

1 Like

Yes indeed. I was thinking the desire was to make the new-tiddler creation process (as it presents in the core with buttons and keyboard shortcut) default to title-only tiddlers.

The only use-case I can think of for title-minimalist tiddlers — pretty different from OP’s interest in efficiency — would be to make a settings tiddler that “wears its content on its sleeve” just by existing as it were, like $:/custom/subtitle/off or something like that, where an override state is toggled by creating this tiddler, and toggled back to normal by deleting it (or bypassing it on save).

Still, this doesn’t scale in a way that makes it worth putting extra effort into bypassing timestamps. And of course the efficiency in such a tiddler is largely offset by configuring other tiddlers to check for such a tiddler.

In general, I’m inclined to think that a tiddler is powerful because it’s multidimensional, and the set of minimalist tiddlers (of this title-only form) might as well just be a list of some kind. (Or, as you say, a data tiddler.)

3 Likes

Ah, a lovely project.

I have a very modest gesture toward random phrases that has a bit of that spirit, making use of a randomizer @Mohammad shared at some point…

For this purpose I suspect @etardiff is right that a dictionary tiddler — or even just a tiddler whose lines can be extracted at random — would serve well and would be mighty efficient.

Still: Certainly there’s something nice about having your lines of text render separately like tiddlers for tinkering and reordering. I think I might actually run with virtual tiddlers for this.

“Generator button, please add to $:/StoryList … 8 lines of text randomly chosen from $:/data/Phrasebook tiddler.”

Delete the “missing tiddler” message, and you’ve got strips of text in the story river, and you can rearrange their order in the open sidebar…

If you want to give me a bunch of phrases, I’ll mock up a proof of concept! :slight_smile:

1 Like

Ah, without having read the whole thread — not having seen your post — it’s something like this very idea that I also landed on.

The lines of can be put in the story river in various ways (drawn from some source list or data tiddler), and once they’re in the story river, they’ll function as title-only “missing” (virtual) nodes.

EDIT TO ADD: For what it’s worth, I still suspect the single “repository” tiddler is a bit more efficient than one-real-tiddler-per-potential-line… since the repository tiddler avoids even the JSON wrapping (naming the title field) for each line. STILL, If the whole wiki were dedicated to this kind of experiment, it might be worth setting it up as minimalist title-only tiddlers; there’s the impression of egalitarian non-orderedness about tidlders (though of course they exist in the html file in an order). Certainly I think the experiment could be intriguing. How (from where) would you be curating the thousands of lines of text?

1 Like

Researching the writing of my book and in my days to day organiser I often create tiddlers whose title is more like prose, and the date fields help me know when they were created and modified. I have felt that one implementation of tiddlywiki would be for creating content for form letters. Where most of the content, standard phrases, salutations or demands would be tiddler titles and in this case we may not even need to know the creation/modified dates (An often valuable piece of info).

One would just edit a tiddler and search for text (as titles) an insert that title using the insert link ctrl-L Editor toolbar button. Leave the titles intact and before printing place \rules except wikilink in the tiddler (or another mechanism)

  • Updating a phrase will be reflected in every tiddler that uses it with relink installed.

The one piece of information missing tiddlers have is the tiddler in which they are “[[named]]”, this is the only mechanism we have to categorise them and should make use of it, such as one full of salutations, and other full of closing statements etc…

Certainly backlinks are one thing we can check for…

But there may also be important filter-discernable facts based on whatever’s in their title, as well as (for missing links) the question whether or not they currently exist in the story river (or any other field-based list). All these things can be leveraged by filters for the sake of cascades and templating… (I’m thinking of the number wiki that both @Scott_Sauyet and I have envisioned, which can offer a story river full of what looks like tiddlers representing numbers — with interesting facts about each one made possible by templates that check for various mathematical properties — but without ever having a single tiddler that is a number tiddler — and without there being any hard-coded backlinks to them either!)

In particular, the OP’s project here might want to parse virtual node titles (or missing tiddler titles — a bit more specific but all in the same conversation) to handle them differently based on various regexp facts about them, as strings. (Does this string have punctuation of this or that kind? What is its length? Does it contain the temp string that user added to the search box? Does it seem to contain a [[link]], and if so, a link to a real tiddler…? etc.)

2 Likes

I like it. This is somewhat related to the no-touch flags solution I experimented with as it interrogates any flag-list field (which I add to relink) but it should be easy enough to do so for the story, A little harder for the history.

Another valuable condition even for a missing tiddler may be is it the “currentTiddler” in the wiki found in the $:/HistoryList!!current-tiddler field (confirmed), think of it as last opened in the story. Others have identified the tiddler at the top of the current view.

Great fun cliched phrase making tool! Thanks! …

I looked at how it works. Something like that could be useful in a “Burrough’s Machine” for what I call the “second pass”—after the first pass which selects phrases to work with from the massive number of input titles.

I’ll ask for further advice if I get stuck.

@etardiff’'s point is maybe true. Not sure yet. I’ll need be further on to get to that.

TT

1 Like