Here is the plugin (I downloaded the copy in my wiki)
Download and then import into your wiki.
$__plugins_ebalster_formula.json (124.3 KB)
Here is the plugin (I downloaded the copy in my wiki)
Download and then import into your wiki.
$__plugins_ebalster_formula.json (124.3 KB)
I want this in TiddlyWiki.
It is a demo how dievelopped the obsidian Excalidraw plugin of Obsidian is by now. Many concepts are very similar to TW like transclusion.
Look at how a prezi-like slideshow can be built at Minute 9
That’s nice, but already on min 1:20 it tells us that we need a server where we need to send our private data, to get a main feature working. — Sorry I did not look any further.
I would be happy to have a mind-map feature as a library but not megabytes of libraries and probably links to tracking code.
I would be happy to have a mind-map feature as a library
do you mean something different than TiddlyMap?
Hi @pmario I think this is a misunderstanding: The OCR-support he mentions there is an obsidian-feature that needs a server in the background, not Excalidraw itself. OCR-Support is a difficult task and I do not know any Web-Canvasses that mange to do this without a backend.
I think, here the autor of the video wants to show how deep Excalidraw is integrated into Obsidian which makes it possible to run obsidian scripts and even obsidian-webservices excalidraw.
In the last months Excalidraw also implemented AI features which need a server but I think taht is optional.
Obsidian either runs as a programm on the PC or in the obsidian cloud - it is their buiseness model to enable people to share parts the things stored in the cloud on the web.
Tiddlymap alas has had no development for at least 5 years. The Developper is not even in this forum.
To be up to date it would definitely need more features. But it is a terrible drag to tinker in Tiddlymap because there only is a minified .js version.
And the worst point: We discussed here that the library it depends on - vis.js is no longer in active development.
The source code is at: GitHub - felixhayashi/TW5-TiddlyMap: Map drawing and topic visualization for your wiki
Okay, I was wrong. And if @Flibbles is commiting to the repo as I just saw, that would help me gain confidence. Sorry, for me tiddlymap is a sort of disappointed love.
So good news on that front. I’ve gotten tired of supporting tiddlymap for people, so I’ve started development of a replacement using memgraph’s orb visualization library. It will have learned from all the mistakes.
I’ll probably post an early version of it when I can to get feedback on how it’s design should be, but it will be weeks or months, since I’ve got other stuff going on.
I did think about jsmind with some TW specific plugins, that would have been to be written. The code base is relatively straight forward. I would like the data structure to be 2-way compatible with the streams plugin.
And TiddlyMap really hasn’t been maintained. I’ve just been doing bug fixes here and there that mostly concern myself than others. As you said, the code base is an absolute pain to work with, being pre-minified and having no testing framework.
This is lovely news, Flibbles! I’ve been using TiddlyMap somewhat reluctantly for a couple of years, as there weren’t really any good alternatives, but it might as well be a black box from a user perspective — and I’ve never figured out how to change or fix anything myself, since it’s basically all JS. I’m sure it was a nightmare to maintain, and I do appreciate that you took on what must have been a pretty thankless task.
Will the replacement have more or less the same features? Any chance it will be easier to modify/configure with wikitext?
My current idea is to make graphs based on templates or \widgets
rather than the heavy-weight views. So instead of having a fixed number of graphs based on views, you would define graphs as you need them like:
<$graph>
<$list filter="[prefix[Nodes/]]">
<$node>
<$action-do-something-on-click />
<$appearance settings here somehow?/>
</$node>
<$list variable=toTiddler filter="[all[current]each:list-item[tags]]">
<$edge to=<<toTiddler>> />
</$list>
</$list>
</$graph>
This would make a graph of nodes based on tiddlers with the “Nodes/” prefix, and edges based on their “tags” field.
The idea is then you could put this is a template tiddler or a procedure, and create graphs based on whatever you want, like different prefixes or something.
I would also have premade \widgets
which would have common configurations of settings or actions. So you could do stuff like, <$graph.editable filter="[prefix[Nodes]]" />
or maybe even <$graph.tiddlymap view=OldTiddlyMapView />
to help with migration.
There are a lot of design choices still in the air, like how to handle actions or visual customization, but I think this general approach is much more flexible. It’s lighter weight, more “Tiddlywiki” in its approach, and should be far more hackable. No more “tmap.id”! Once my hackathon gets further along, I’ll probably ask people what it would need to be able to do for them to migrate. Like do people actually manually add edges and nodes like tiddlymap allows?
I like the syntax, very legible. I’m not currently using TM for any dynamic graphs, but I might start if it were this simple.
I don’t particularly care about manual edges, but manually added nodes are crucial to me. I’ve been using TM mostly to annotate maps and static diagrams; I like that I can hover over a node to see a preview, and double-click to open a tiddler in the story river.
Hi @Flibbles, that is great news! I immediately hat to look what the “memograph orb” is:
This was the image i liked best. (source:GitHub - memgraph/orbicon: Explore your userbase and community with the power of graph analytics.)
This looks like a neat visualiation for structured data.
As you wanted feedback - what I fear I could be missing in this approach
As a teacher I want the possibility to do untidy jams on my board.
So I would like to
And I would like to
Do you think some of this could be possible?
That is good to hear, I really liked TiddlyMap, though I often cusotmized it to look more akin to how obsidianMD looks.
In addition to making modules more decoupled from each other like @Mohammad said, making each module a common js library that can be used by all js software developers. Also want tiddlywiki to be front-end headless software. Only provide api on the browser side. no involvement in interface design.
Except ui stripped out. I also hope that the data storage side is also stripped out.
BTW, a question. Why is the data storage format [] and why not {}? Is it because of performance reasons?
One thing I miss in TiddlyWiki, that is now a standard on most text-based interfaces (eg in Notion, but also in Github and Gitlab for example), is the ability to quickly make todolists by simply adding [ ]
/ [x]
to the start of a bulleted list, e.g.
* [x] A checked item
* [ ] An unchecked item
producing an interface with a clickable checkbox that, when checked, edits the actual line and not a field somewhere else.
(both in TW5 and markdown tiddlers)
Yes, it is not part of core, but it exists as plugin feature. Examples are kara, table-csv shiraz
There are other plugins around.