Markup for note taking

Overview

Subtext: markup for note-taking

Subtext is a text-based, block-oriented hypertext format. It is designed with note-taking in mind. It has a simple, people-friendly syntax with a passing resemblance to Markdown.

See the Speculative Specification.

We're experimenting with Subtext as part of Subconscious, a new tool for thought.

Warning to implementors: Subtext is currently experimental status. We'll be spending some time living with Subtext and building experimental tools on top of it before committing to anything. The language design is just a hypothesis! It might undergo radical breaking changes! This is work in progress, and shared in the spirit of working with the garage door open.

A bit of Subtext

Here’s an example:

# Heading

Plain text

- List item
- List item

> Quoted text

& example.csv
& https://example.com

Subtext is line-oriented. Each line in the file is treated as a discrete block of content. The type of a line is determined by a sigil character, like #, &, >, at the front of the line. If a line doesn’t have a sigil character, it is treated as plain text. This makes Subtext very easy to parse, and very easy to write. It is currently impossible to write broken Subtext, which is nice!

Subtext is for notes

Today the book is already… an outdated mediation between two different filing systems. For everything that matters is to be found in the card box of the researcher who wrote it, and the scholar studying it assimilates it into his own card index. (Walter Benjamin)

HTML comes in web pages. The analogy for an HTML document is quite literally a page. The image that springs to mind is of an 8.5x11” sheet, carefully typeset, with multiple fonts, headings, complex formatting, perhaps laid out across many columns. HTML is a publication format, designed to produce complete, indivisible artifacts, called pages.

The right mental analogy for Subtext is not the page. It is the the index card.

Subtext deliberately avoids the kind of complex presentation features offered by publishing formats like HTML, PDF, and LaTex. It has no opinions about fonts, colors, sizes.

Like a stack of index cards, there are many ways to use Subtext, beyond simple linear layout. It isn’t just for narrative. It’s hypertext montage.

Subtext is block-oriented

Subtext represents block-oriented documents as line-oriented markup.

A block-oriented document is made up of a list of blocks of different types (or occasionally, a tree of blocks). Each block type may be displayed differently. For example, a quote block may render as quote-formatted text, while an image block may render an image in-place.

Some of the earliest hypertext proposals were block-oriented, including Ted Nelson's ELF (Nelson, 1965). Block-oriented documents have also independently evolved within many contemporary tools-for-thought, including Notion, Roam, and Ward Cunningham's Federated Wiki.

Why does this pattern keep re-emerging? One reason might be that block-oriented editing is an easy way to express rich formatting. But more importantly…

Blocks are composable

Blocks are thought legos. A block-oriented document is composable (and decomposable). You can break it apart into component blocks, filter it down to blocks of a particular type, merge documents, pluck out blocks, link to specific blocks, etc.

In theory, this is true of any tree-based markup language, such as HTML. But try meaningfully merging two HTML files in practice... Yikes! Tag soup!

A linear block-oriented format resolves the problem by radically simplifying it. With a linear data model, the range of meaningful document structures is narrowed, and this means you can make complex, yet meaningful programmatic decisions, without much context about the specific document:

  • Excerpt a document by taking the first text block
  • Select all quotes from a collection of documents
  • Select all links, and generate a link graph for a collection of documents
  • Find all backlinks and append them to the document as links

Linear block-oriented documents are like shipping containers for discrete thoughts. Because blocks are structurally uniform, they can be automatically moved around and reorganized. Software can split, join, and merge documents easily and effectively, because the document structure is simple.

Subtext is hypertext

Link blocks (&) are the most important feature in Subtext. They let you reference other files, and URLs. You can link to any kind of file, including other Subtext files!

The plan is to have Subconscious display these links as transclusions. Rather than linked words in text, imagine something more like a quote tweet… Links to images display as literal images, links to videos display as playable videos with playback controls, links to documents display some or all of the content inside of the linked document. This lets you compose hypertext documents from many smaller documents.

This keeps Subtext simple. Rather than extending the syntax of Subtext to include a complex feature like tables, we might, for example, link to a .csv file, which then gets rendered as a table. This also means the data stays in its native file type, and can be used in other applications.

One of the many attempts of nature to evolve a Xanadu

By an accident of convergent evolution, Subtext happens to have some similarities to Ted Nelson's ELF format (Nelson, 1965).

Ted Nelson “A File Structure for the Complex, the Changing, and the Indeterminate”, 1965

Like ELF, Subtext documents are made up of a list of small blocks. Also like ELF, links are transcluded. Big documents can be composed by linking to small documents.

I discovered Ted Nelson’s ELF paper after writing up my first draft of Subtext. Uncovering this bit of convergent evolution was encouraging! It suggests I’m pulling on a worthwhile thread. Xanadu by way of Markdown? Something like that.

Why not Markdown?

I took a deep breath before thinking about the jump from Markdown. If you’re a programmer, Markdown is a de-facto standard for formatted text. For many, it is the first obvious choice for this kind of thing. So why Subtext?

Subtext has evolved out personal experiments with plain-text note-taking, spanning 10 years and 12k notes. Many of these notes are written in Markdown. However, over time, I noticed that my markup needs for note-taking were different from my markup needs for publishing. My note-taking style organically converged on a tiny subset of Markdown's features: text, links, lists, quotes, and one level of heading. To have more may be useful for publishing, but is often overkill for note-taking.

At the same time, I began to write small generative programs that worked with this collection of notes, little scripts that would combine ideas, remix notes, algorithmically generate new notes… these were the seeds that would later become Subconscious.

Here, I started to run into limitations with Markdown and HTML. As a complex publishing format, it is unclear how to meaningfully decompose or merge Markdown/HTML documents. When you combine documents, heading levels may need to be changed, lists may need to be flattened or nested. Because the document format is complex, foreknowledge of the meaning of the document is necessary to make meaningful changes. That limits what you can do with software.

Subtext is an attempt to resolve the problem by radically simplifying it. Paradoxically, by limiting the format to a flat list of blocks, we radically expand what software can usefully do with it. Blocks are easy to parse, easy to work with, and you can do all sorts of interesting generative algorithmic things with them.

The syntax is also simple, and hard to mess up, and I’m happy about that, too.

Project links

Owner
Gordon Brander
Building something new (prev @google, @mitmedialab, @mozilla).
Gordon Brander
Implementation of our paper 'PixelLink: Detecting Scene Text via Instance Segmentation' in AAAI2018

Code for the AAAI18 paper PixelLink: Detecting Scene Text via Instance Segmentation, by Dan Deng, Haifeng Liu, Xuelong Li, and Deng Cai. Contributions

758 Dec 22, 2022
[EMNLP 2021] Improving and Simplifying Pattern Exploiting Training

ADAPET This repository contains the official code for the paper: "Improving and Simplifying Pattern Exploiting Training". The model improves and simpl

Rakesh R Menon 138 Dec 26, 2022
Textboxes_plusplus implementation with Tensorflow (python)

TextBoxes++-TensorFlow TextBoxes++ re-implementation using tensorflow. This project is greatly inspired by slim project And many functions are modifie

81 Dec 07, 2022
OCR, Object Detection, Number Plate, Real Time

README.md PrePareded anaconda env requirements.txt clova AI → deep text recognition → trained weights (ex, .pth) wpod-net weights (ex, .h5 , .json) ht

Kaven Lee 7 Dec 06, 2022
Image processing in Python

scikit-image: Image processing in Python Website (including documentation): https://scikit-image.org/ Mailing list: https://mail.python.org/mailman3/l

Image Processing Toolbox for SciPy 5.2k Dec 30, 2022
Convolutional Recurrent Neural Networks(CRNN) for Scene Text Recognition

CRNN_Tensorflow This is a TensorFlow implementation of a Deep Neural Network for scene text recognition. It is mainly based on the paper "An End-to-En

MaybeShewill-CV 1000 Dec 27, 2022
Document blur detection based on Laplacian operator and text detection.

Document Blur Detection For general blurred image, using the variance of Laplacian operator is a good solution. But as for the blur detection of docum

JoeyLr 5 Oct 20, 2022
This pyhton script converts a pdf to Image then using tesseract as OCR engine converts Image to Text

Script_Convertir_PDF_IMG_TXT Este script de pyhton convierte un pdf en Imagen luego utilizando tesseract como motor OCR convierte la Imagen a Texto. p

alebogado 1 Jan 27, 2022
A python screen recorder for low-end computers, provides high quality video output.

RecorderX - v1.0 A screen recorder made in Python with the help of OpenCv, it has ability to record your screen in high quality. No matter what your P

Priyanshu Jindal 4 Nov 10, 2021
Hand gesture detection project with aweome UI implementation.

an awesome hand gesture detection project for you to be creative! Imagination is the limit to do with this project.

AR Ashraf 39 Sep 26, 2022
Code release for our paper, "SimNet: Enabling Robust Unknown Object Manipulation from Pure Synthetic Data via Stereo"

SimNet: Enabling Robust Unknown Object Manipulation from Pure Synthetic Data via Stereo Thomas Kollar, Michael Laskey, Kevin Stone, Brijen Thananjeyan

68 Dec 14, 2022
Tensorflow-based CNN+LSTM trained with CTC-loss for OCR

Overview This collection demonstrates how to construct and train a deep, bidirectional stacked LSTM using CNN features as input with CTC loss to perfo

Jerod Weinman 489 Dec 21, 2022
Document Image Dewarping

Document image dewarping using text-lines and line Segments Abstract Conventional text-line based document dewarping methods have problems when handli

Taeho Kil 268 Dec 23, 2022
(CVPR 2021) ST3D: Self-training for Unsupervised Domain Adaptation on 3D Object Detection

ST3D Code release for the paper ST3D: Self-training for Unsupervised Domain Adaptation on 3D Object Detection, CVPR 2021 Authors: Jihan Yang*, Shaoshu

CVMI Lab 224 Dec 28, 2022
Just a script for detecting the lanes in any car game (not just gta 5) with specific resolution and road design ( very basic and limited )

GTA-5-Lane-detection Just a script for detecting the lanes in any car game (not just gta 5) with specific resolution and road design ( very basic and

Danciu Georgian 4 Aug 01, 2021
Polaris is a Face recognition attendance system .

Support Me 🚀 About Polaris 📄 Polaris is a system based on facial recognition with a futuristic GUI design, Can easily find people informations store

XN3UR0N 215 Dec 26, 2022
Table Extraction Tool

Tree Structure - Table Extraction Fonduer has been successfully extended to perform information extraction from richly formatted data such as tables.

HazyResearch 88 Jun 02, 2022
Detect textlines in document images

Textline Detection Detect textlines in document images Introduction This tool performs border, region and textline detection from document image data

QURATOR-SPK 70 Jun 30, 2022
TableBank: A Benchmark Dataset for Table Detection and Recognition

TableBank TableBank is a new image-based table detection and recognition dataset built with novel weak supervision from Word and Latex documents on th

844 Jan 04, 2023
Provides OCR (Optical Character Recognition) services through web applications

OCR4all As suggested by the name one of the main goals of OCR4all is to allow basically any given user to independently perform OCR on a wide variety

174 Dec 31, 2022