DEV Community: Ian Effendi

Quicklook: Discovering Logseq

Ian Effendi — Thu, 01 Aug 2024 22:42:32 +0000

Photo by Viridiana Rivera

LogSeq is a personal knowledge management system application that assists those interested in working with lots of networked information.

Discovering LogSeq

Screenshot of the LogSeq homepage.

What is this software?

LogSeq is suited for knowledge workers. It's the perfect tool for students, academics, writers, project managers, and developers (among other professions) to manage, store, and retrieve notes that they create.

LogSeq is provided under the AGPL-3.0 License.

Who maintains this software?

LogSeq is open-source software with a relatively healthy community.

At the time of writing, the last commit was from 2 weeks ago, with over 1,524 open issues and 53 open pull requests.

Who created this software?

The project was started by Tienson Qin (@tiensonqin) four years ago on May 23rd, 2020. Qin (and four others) went on to create the privately funded LogSeq, Inc..

LogSeq, Inc. raised $4.1M in May 2022, with some very notable investors:

Patrick Collision
Nat Friedman
Tobias Lütke, Founder of Shopify
Sriram Krishnan, GP at A16Z
Charlie Cheever, Founder of Expo/Quora
Dave Winer, the forefather of outliners, scripting, weblogs, RSS, podcasting, etc.

Where do you go to download new releases?

LogSeq provides a number of built binaries for different platforms on their downloads page.

What is this software powered by?

According to the breakdown on GitHub, LogSeq is written with the following languages:

Clojure 59.3%
TypeScript 20.5%
CSS 13.4%
JavaScript 4.7%
HTML 0.9%
Swift 0.5%
Other 0.7%

Logseq is also made possible by the following projects:

Clojure & ClojureScript - A dynamic, functional, general-purpose programming language
DataScript - An immutable database and Datalog query-engine for Clojure, ClojureScript and JS
OCaml & Angstrom, for the document parser mldoc
isomorphic-git - A pure JavaScript implementation of Git for NodeJS and web browsers
SCI - A Small Clojure Interpreter

Reception

LogSeq itself has been well-received, although it hasn't reached the same levels of market utilization as similar PKMS giant Obsidian.

The good news, is that LogSeq has a pretty decent following with 30K+ followers on Twitter and 30k+ stars on GitHub at the time of writing.

What's it do well?

This probably has a lot to do with LogSeq's primary philosophy diverging from known tools:

Logseq is challenging legacy knowledge management systems that are based on skeuomorphic design (which mimics the systems used for storing physical papers, files, and folders). Why? Because humans don't think linearly in pages and folders, but rather link inter-connected concepts together non-linearly. Logseq is designed to capture this type of non-linear, thinking creating intuitive and interconnected data, and allowing users to unlock their maximum potential.

The pure emphasis on backlinking and graph discoverability for ones notes is a novel idea and takes some getting used to for people who are purely familiar with traditional notetaking applications (eg. physical notebooks and their similar digital counterparts).

For users who get past this hurdle, they tend to love it!

What can it improve?

LogSeq is technically early access - there isn't a version 1.0.0 available yet. The maintainers themselves maintain a roadmap of what they need to do to get there. One particular sticking point is performance of large "graphs" and the lack of real-time collaboration, which they talk about here.

Any competitors on the market?

There are plenty of similar competitors currently on the market:

Further resources

These resources may be valuable:

Crosspost! Publishing to Dev.to From My Personal Blog

Ian Effendi — Thu, 01 Aug 2024 22:41:32 +0000

Photo by Shahram Hafezi

Would you look at that? You’re now seeing crossposts to my dev.to blog. Check out the canonical link at the top to link back to the original post!

How'd I do it?

There were two main parts:

Generate an RSS feed for the original blog at effendiian.github.io
Connect it using the RSS extension provided by Dev.to's settings.

Generating the RSS feed

My website is a static site built with Hexo and served through GitHub Pages. Hexo's documentation isn't the best, but with a little digging, I found that, in the years since I last used it, they've provided a pretty robust first-party plugin for generating RSS and ATOM feeds.

You can read more about atom.xml feeds here.

Using the RSS feed

Thanks to the excellent writeup from Tessa Kriesel on, Connecting your RSS feed URL for easy publishing to dev.to, it was quite simple to get it all working.

You simply add the feed to your Dev.to site via the RSS extension under settings.

You can read more about rss2.xml feeds here.

Connecting to other third-party platforms

This is good for now, but I'd also like to see if I can connect to other platforms, like LinkedIn and the site formerly known as Twitter.

There may even be a pretty reasonable solution using GitHub actions, given this project from @lwojcik/github-action-feed-to-social-media, which allows for posting to platforms like Mastodon and Discord as well.

Why do this now?

Three driving reasons for why I want to get back into blogging:

Vague interests in technical writing.
I love organizing information - similar to personal knowledge management systems.
I should be documenting my learning.

The old site hadn't been touched in years since I was busy getting degrees and working full-time. Now that I've finished the official rebranding of my site and @effendiian GitHub account (formerly known as @rimij405), it's time to start posting, making things, and sharing on my different social media platforms.

With that, I'm glad to be writing again!

What's next? What should you expect?

Expect more. Well, more of this stuff:

Frequent, shorter posts that get straight to the point
Occasional post-mortems and retrospectives
Musings on system design, every once in a while
Observations about note-taking process and knowledge management
Quicklooks into useful open-source projects

There are also some other site-related improvements:

[ ] Updating the footer so it's not so cramped and has a better sitemap layout.
[ ] Embed this GitHub contributions tracker on the site.
[ ] Adding support for ::-style emojis (eg. :smile:) with this hexo-filter-emoji package.
[ ] Polishing the website's landing page (maybe with some 3D illustrations... looking at you Three.js)
[ ] Adding some analytics (which is simple, but also requires I add cookie consent trackers and global policy control listeners... so not so straightforward).

I'm also considering rewriting the application with a framework like Nuxt or adding UnoCSS styles, just to play around with the technology and see what I can do. Hexo feels like it's not quite ready for production use in an enterprise context, but it's really straightforward to get up and running with for smaller groups and solo developers.

Feature: chromium-apt-get Issue 626 (PR 627)

Ian Effendi — Wed, 31 Jul 2024 03:32:26 +0000

Photo by Deepanker Verma

Put together a quick Dev Container feature that I’m hoping to share with the community. This article was originally published on my personal site.

Background

I needed a headless browser instance for running the Lighthouse CLI from within a Docker container. I thought I could get a quick one setup for Dev Containers, but it turns out there wasn't an easy way to get headless chromium installed onto a Dev Container.

So I made a feature request on the devcontainers-contrib/features repository.
[Feature Request]: chromium-apt-get #626.

I wasn't even the first person to ask for something like this, at least, in the broader community: [Feature] Headless browser (chrome/chromium/firefox) devcontainers/features#189

The Feature

DevContainers work by installing features in layers through their devcontainers.json specification. We just needed to get it in there.

I added a pull request that describes the whole thing here: Pull Request #627.

You can use this feature like so:

"features": {
 "ghcr.io/devcontainers-contrib/features/chromium-apt-get:1": {},
}

📖 Context

There are some historical reasons for why it's being done this way:

The official Chromium team doesn't supply "stable" binaries of Chromium; they are usually compiled into stable versions by third-parties.
Chromium installation breaks when installed with snap packages on Ubuntu.
apt-get install -y chromium on Ubuntu 20.04+ is just a wrapper for a snap installations.
Referenced this guide to find a reliable PPA: How to Install Chromium Browser via PPA in Ubuntu 20.04, 20.10 - credit to UbuntuHandbook.

⚖️ Caveats

The maintainers may wish to scrutinize the default PPA used here, both for security and stability concerns. I agree!

I think future improvement to this feature could include:

Parameterizing the selected PPA to allow sourcing chromium binaries from public and private PPAs.

Next Steps?

Now, all that's left is to wait for the pull request to be reviewed and then I can go ahead and mark this one as done.

DS1 - Journey into Python

Ian Effendi — Mon, 22 Apr 2019 20:40:35 +0000

Data Science 001 - Journey into Python

This was originally posted on my blog here.

I am practicing being concise. I want to simultaneously explore the field of data science and document my journey.

As part of enrollment in a seminar covering humanitarian free (libre) and open software (HFOSS), I was introduced to Python for the sake of creating an educational application for the Sugar desktop environment. Sugar is not the focus of this post; instead, I want to focus more on learning Python.

The Past Informs the Present

I was born in December of 1997. I was first introduced to Python when I was seven years old. My father sat me down in front of an old computer tower that had a floppy disk drive in it - I don't remember the model name - and put me in front of an old IDE. I don't remember the name of that, either.

The syntax didn't stick, but, the concepts remained. I would go on to enroll at the Rochester Institute of Technology in 2015 and pursue my undegraduate degree in Game Design & Development.

Data Science

Recently, I became enamoured with the idea of actually using my knowledge of computer science to explore datasets in a meaningful way. Many years ago, while enrolled in high school, I was privileged to be in a school that was fell-funded and held encouraging programs for those interested in the sciences. At the time, I was still learning the basics of Java and C#, so data never gripped me as much as the game development industry did.

I went on to watch my peers do interesting things with data - I recall one group mapped geography to crime data and won a competition for their findings. My memory fails me, and, I am unsure of the specific name of that competition - but the essence of what inspired me still influences me to pursue this field in my leisure time.

Why Python?

Python is only one option when dealing with data analysis and visualization. There are libraries for data visualization in JavaScript, such as D3.js and R is a free software environment created for statistical computing and graphics.

I simply chose Python since I was already using it for Sugar development and felt that the barrier to entry would be lower.

Learning Python

So that's the goal: learn python for the sake of getting better with Data Science and chronicle the journey along the way.