As you start developing an AI/ML based solution, you quickly figure out that you need to run workflows. Not only that, you might need to run those workflows across various kinds of infrastructure (including GPUs) at scale. Ville Tuulos developed MetaFlow while working at Netflix to help data scientists scale their work. In this episode, Ville tells us a bit more about MetaFlow, his new book on data science infrastructure, and his approach to helping scale ML/AI work.

Encore now supports the two most highly requested features from our users!

• Native support for building REST APIs
• Support for ORMs and other database query helpers

Encore v0.17 also comes with improvements to authentication, and lots of minor improvements and bug fixes!

Alexey Milovidov, announcing the formation of a (VC funded) corporation around ClickHouse, an open source analytics DBMS:

Today I’m happy to announce ClickHouse Inc., the new home of ClickHouse. The development team has moved from Yandex and joined ClickHouse Inc. to continue building the fastest (and the greatest) analytical database management system. The company has received nearly $50M in Series A funding led by Index Ventures and Benchmark with participation by Yandex N.V. and others. I created ClickHouse, Inc. with two co-founders, Yury Izrailevsky and Aaron Katz. I will continue to lead the development of ClickHouse as Chief Technology Officer (CTO), Yury will run product and engineering, and Aaron will be CEO.

ClickHouse wasn’t always a business. It also wasn’t always open source.

Making ClickHouse open source was also not an easy decision, but now I see: doing open source is hard, but it is a big win. While it takes a tremendous effort and responsibility to maintain a popular open-source product, for us, the benefits outweigh all the costs. Since we published ClickHouse, it has been deployed in production in thousands of companies across the globe for a wide range of use cases, from agriculture to self-driving cars.

Seed is a Rust framework that uses an Elm-like architecture to help you build fast and reliable web apps that run on WebAssembly. Here’s why you might want to use it:

Seed allows you to develop the front-end with all the benefits of Rust, meaning speed, safety, and too many more things to count.

The Seed templating system uses a macro syntax that makes Rustaceans feel right at home. This means linting, formatting, and commenting will work, and it’s all in Rust. This is opposed to a JSX-like syntax that relies on IDE extensions to improve the developer experience.

Seed has a batteries-included approach. This means less time writing boilerplate and less time installing dependencies.

And a few reasons why you might not:

• It’s newer. It’s harder to find support outside of Discord.
• WebAssembly is newer with less support. Browser compatibility is at 92.9%.
• Pre-built components are rare. You will likely have to roll your own components such as date pickers.
• No server-side rendering yet
• You may prefer other Rust frameworks like MoonZoon or Yew.

Following thanks to all contributors, the blog notes:

Most applications that worked with OpenSSL 1.1.1 will still work unchanged and will simply need to be recompiled (although you may see numerous compilation warnings about using deprecated APIs). Some applications may need to make changes to compile and work correctly, and many applications will need to be changed to avoid the deprecations warnings.

And points out a couple of new features:

OpenSSL 3.0 introduces a number of new concepts that application developers and users of OpenSSL should be aware of. An overview of the key concepts in libcrypto is available in the libcrypto manual page.

A key feature of OpenSSL 3.0 is the new FIPS module. Our lab is testing the module and pulling together the paperwork for our FIPS 140-2 validation now. We expect that to be submitted later this month. The final certificate is not expected to be issued until next year.

And finally, LWN notes on the license change:

OpenSSL has also been relicensed to Apache 2.0, which should end the era of “special exceptions” needed to use OpenSSL in GPL-licensed applications.

This post is a confession of an “egocentric maniac” (his words) and how damaging code review can be:

This review I kicked off the article with? I didn’t send it. Instead I gave the guy a couple of comments and politely asked to fix a couple of things. No big deal if the code’s not good, I can fix it myself it I need to. But I can’t fix the psyche of a guy broken by dozens of harsh reviews.

My personality today isn’t my disease. It’s a disease of the whole industry, at least in Russia. Our mentality is predicated on the cult of power and superiority. And that’s what we need to fix: just stop being that. It’s quite easy, actually.

This week we’re joined by Adam Jacob, CEO of System Initiative and Co-Founder of Chef, about open source business models and the model he thinks is the right one to choose, his graceful exit from Chef and some of the details behind Chef’s acquisition in 2020 for $220 million…in cash, and how his perspective on open source has or has not changed as a result. Adam also shared as much stealth mode details as he could about System Initiative.

Russel Goldenberg & Caitlyn Ralph from The Pudding join Amelia & Nick to talk about how they create data-driven, interactive articles, how the team works on both The Pudding’s data journalism articles and Polygraph’s client work. We also dive into how the team works with contractors and how the company manages itself using a Holocratic method.

Chris Fidao:

I’ve been using AWS “professionally” since about 2015. In that time, I’ve made lots of mistakes.

Other than occasionally deleting production data, the mistakes all arose from ignorance - there’s so much to know about AWS that it’s easy to miss something important.

Here’s a collection of the most commonly missed things when using AWS with Laravel Forge!

Learning from your mistakes is powerful. Learning from other people’s mistakes can be just as powerful without the major drawback of, you know, feeling all that pain!

This week Emile Vauge, founder & CEO of Traefik, joins Gerhard to share a story that started as a solution to a 2000 microservices challenge, the real-world implications of shipping many times a day for years, and the difficulties of sustaining an inclusive and healthy open-source community while building a product company.

Working every day on keeping the open-source community in sync with the core team was an important lesson. The second learning was around big changes between major versions.

The journey from Travis CI to Circle CI, then to Semaphore CI and eventually GitHub Actions is an interesting one. The automation tools inspired by the Mymirca ant colony is a fascinating idea, executed well. There is more to discover in the episode.

At the moment, there is only a dozen emoji sets, most of them from big tech companies. These emoji are visually adapted to the respective appearance of their software platforms. In addition, the usage rights are often very restrictive (e.g. the terms of use of Apple’s emoji).

That is why we have developed OpenMoji as the first open source and independent emoji system to date. When designing the OpenMoji system, we have developed visual guidelines that are not linked to a specific branding. In addition, our goal was to design emojis that integrate well in combination with text.

A collaboration of 60+ students and 3 professors from HfG Schwäbisch Gmünd.

Natalie sits down with Go book authors Bill Kennedy & Sau Sheong Chang to discuss the ins and outs of writing (and reading) books about Go!

Wise words from Rach Smith:

What I’ve learnt through experience is that the number of languages I’ve learned or the specific frameworks I’ve gained experience with matters very little. What actually matters is my ability to up-skill quickly and effectively. My success so far has nothing to do with the fact I know React instead of Vue, or have experience with AWS and not Azure. What has contributed to my success is the willingness to learn new tools as the need arises.

The author of this site handed a neural network brief text descriptions of a bunch of movies and let it generate image that represent each. Read how he did it or just have fun trying to guess the movie titles from the images. Harder than I thought it’d be!

Here’s how Tom Payne describes his project:

chezmoi is a popular dotfile manager (currently over 4.5K stars on GitHub and increasing quickly). chezmoi helps you get your prefered environment synchronized across multiple machines (e.g. your home desktop, your work laptop, and a temporary development container in the cloud) while easily coping with differences from machine to machine and keeping all your secrets safe either with your password manager or encryption. Using chezmoi feels very much like using git (and indeed it builds on git). chezmoi is easy to install, quick to start with, runs everywhere, and scales from managing a handful of files on one machine to complex multi-machine set-ups with hundreds of dotfiles and plugins.

Getting a new machine set up looks like:

$ sh -c "$(curl -fsLS git.io/chezmoi)" -- init --apply <github-username>

My dotfiles “manager” is just a combination of git clone and setup.sh, but if I used many machines I’d probably reach for something more robust like this. If you’re already using a manager for yours, here’s a comparison guide of how chezmoi stacks up to other popular options.

In this article, three experts discuss some of the key findings of the State of Technical Debt 2021 report including the impact of technical debt on engineering teams, the pros and cons of dealing with maintenance work continuously, the future of technical debt, and what each engineering teams can do to communicate the importance of dealing with technical debt to leadership.

Remember, not all tech debt is bad, but even good tech debt has its ramifications. Solid insights here.

Any AI play that lacks an underlying data strategy is doomed to fail, and a big part of any data strategy is labeling. Michael, from Label Studio, joins us in this episode to discuss how the industry’s perception of data labeling is shifting. We cover open source tooling, validating labels, and integrating ML/AI models in the labeling loop.

A (short) must-read piece from rachelbythebay:

“Everyone knows” that code is something you type into a computer, that gets interpreted by a computer, and is run by a computer. But that’s not really the end of it. Before that, it’s being “run” on whoever’s working on it. After that, it’s “running” on whoever’s digging into it to fix a bug or add some feature.

You can throw all kinds of wicked, nasty, complicated, Klein-bottle-wannabe tricks into code and the computer will shrug and slog on through it.

Try feeding that same mess to a human and you will have a variety of problems. We see them every day, and, unfortunately, we /create/ them every day.

This is phoboslab’s entry in this year’s JS 13k Games competition. Stunning what you can accomplish with so little. View the source right here and play the game (desktop only) right here.

I read about KDL (pronounced “cuddle”) over the weekend. Color me impressed!

Kat (and others) put a lot of thought into this.

KDL is a document language with xml-like semantics that looks like you’re invoking a bunch of CLI commands! It’s meant to be used both as a serialization format and a configuration language, much like JSON, YAML, or XML.

Check out the FAQ for all the common objections (like why not YAML or TOML or ETC) as well as a note about the XKCD comic you’re probably thinking about. Here’s hoping it catches on. 🤞