This could be life changing for me, very glad I found this.
But for people who want a use case: See that entry called svgbob? That’s pretty incredible. If you need a logo for a minor project, just throw some characters in a file and let it do its thing.
From the results there, it’ll generate something professional quality, and because it’s an SVG, as a logo, it will be superior to any jpg or png file, able to scale to any size without degradation.
I wish the author added a brief description for each entry and how they differ. I’m particularly interested to find out which is best for making networking topography diagrams. This is still useful nevertheless.
I've been experimenting with many of them and PlantUML is the best one I've used so far.
It has a great feature set, and a lot of ways to define different alignment and styles of the diagrams.
I'd love on with a basic live collaboration feature. In Covid I'm really noticing the lack of a whiteboard and so long as you know what type of diagram you're doing from the outset I'm found doing something in Mermaid/similar to be much quicker than online whiteboards.
So, I never tried to make text diagrams and the thought of doing it (by hand, I assume) haunts me:
How is it considered easy to build a simple table (or diagram) by having to hit keyboard keys and navigate the cursors to seemingly random places to place characters, while keeping the spacing, proportions, etc correct?
I don't like word processors but at least they recognize that if I drawing of building a table I need a different interaction model.
Maybe someone in this thread could highlight use cases where making text-diagrams (that may be later converted using these tools) is better than picking a richer text editor.
Markdown tables are bad. Apart from that, maintaining/updating a textual description is way more easier. Because its text, you can have simple version control over it
Making plain-text diagrams isn't as hard as you might think, especially using "overwrite mode" (that weird thing that happens when you fat-finger the insert key!)
But many of these tools work by allowing you to express the _intent_ of the diagram without having to draw it out. They're basically markup languages that get "compiled" into diagrams. PlantUML is a good example: http://www.plantuml.com/plantuml/uml/SyfFKj2rKt3CoKnELR1Io4Z.... With these tools, you can make diagrams much, much faster (but usually with less nuanced control over presentation)
I used MermaidJS for some diagram I wanted to add to some of my notes in Markdown, and it was simple enough and Joplin (markdown-based note-taking tool) added support not too long ago.
And you can also add Mermaid preview support to VSCode if that's something you want.
31 comments
[ 2.8 ms ] story [ 70.1 ms ] threadBut for people who want a use case: See that entry called svgbob? That’s pretty incredible. If you need a logo for a minor project, just throw some characters in a file and let it do its thing.
From the results there, it’ll generate something professional quality, and because it’s an SVG, as a logo, it will be superior to any jpg or png file, able to scale to any size without degradation.
http://www.webgraphviz.com/
There's also some interesting projects on GitHub for generated PlantUML diagrams, including this one: https://github.com/thibaultmarin/hpp2plantuml
Without a description, I have no idea, what is the input, outcome, are these popular, are these in active development, are these open source, etc.
While more elaborate interactive lists (e.g. https://explorabl.es/ and https://p.migdal.pl/interactive-machine-learning-list/) might be more time consuming, putting screenshots isn't.
This is doing to vice-versa, it allows you to create a text-based diagrams.
How is it considered easy to build a simple table (or diagram) by having to hit keyboard keys and navigate the cursors to seemingly random places to place characters, while keeping the spacing, proportions, etc correct?
I don't like word processors but at least they recognize that if I drawing of building a table I need a different interaction model.
Maybe someone in this thread could highlight use cases where making text-diagrams (that may be later converted using these tools) is better than picking a richer text editor.
But many of these tools work by allowing you to express the _intent_ of the diagram without having to draw it out. They're basically markup languages that get "compiled" into diagrams. PlantUML is a good example: http://www.plantuml.com/plantuml/uml/SyfFKj2rKt3CoKnELR1Io4Z.... With these tools, you can make diagrams much, much faster (but usually with less nuanced control over presentation)
And you can also add Mermaid preview support to VSCode if that's something you want.
https://github.com/mjbvz/vscode-markdown-mermaid
https://marketplace.visualstudio.com/items?itemName=bierner....
[1]: https://www.diagram.codes/d/graph