Search the archive

Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.

Results

14,500 matches

ZERO DOWNTIME POSTGRES UPGRADES

Knock’s team successfully upgraded from PostgreSQL 11.9 to 15.3 with zero downtime using logical replication and various tools. They were able to de-risk the upgrade process by monitoring it closely and having carefully planned processes in place. This article explains different options for upgrading PostgreSQL, including in-place upgrades (not suitable for zero downtime), and replication-based upgrades, which is what they chose. A replication-based upgrade involves setting up publications and subscriptions, adding tables incrementally, and carefully configuring source and destination databases.

2023.12.13
knock.app29 min readTLDR Web Dev

HOW TO CREATE AN AUTOMATED PROFILE README USING NODE.JS AND GITHUB ACTIONS

GitHub gives you the capability to build a profile README, a Markdown file that lives on your profile that can be used as a quick portfolio to highlight achievements. This article covers how you can take your profile README to the next level with scripting and CI/CR automation. It shows you how to do this using a custom script built with Node.js, the markdown-it library, the GitHub API, RSS Feeds, and GitHub Actions.

2023.12.13
blog.bolajiayodeji.com16 min readTLDR Web Dev

BUILDING COMPONENTS FOR CONSUMPTION, NOT COMPLEXITY

Creating and maintaining a design system is challenging but necessary for a great user experience. This article discusses considerations for responsive design within component libraries and suggests structuring components based on the design and development team's specific needs and platforms. The author provides practical insights and recommendations for structuring design systems within design tools like Figma.

2023.12.13
smashingmagazine.com23 min readTLDR Web Dev

THE VALUE OF REPAYING GOOD TECHNICAL DEBT

Good technical debt, as opposed to bad technical debt, is intentional, provides initial benefits, and is controlled through tripwires to prevent excessive interest. Teams often struggle to articulate the business value of repaying technical debt, leading to the misconception that a total system rewrite is the only solution. A simpler approach is proposed: identify hindrances to progress, capture them on sticky notes with dates, and add dots each time they recur.

2023.12.13
infoq.com4 min readTLDR Web Dev

THE SURPRISING CONNECTION BETWEEN AFTER-HOURS WORK AND DECREASED PRODUCTIVITY

With the results of a survey taken by 10,000 desk workers, Slack’s Workforce Index uncovers some interesting key learnings about workday productivity. The survey found that employees who step away from their computers at the end of the workday register a 20% increase in productivity compared to those who work after hours. Only 25% of workers feel productive in the afternoon and workers who regularly take breaks have 13% higher productivity.

2023.12.13
slack.com7 min readTLDR Web Dev

MACOSPILOT

macOSpilot is an AI assistant that can answer your questions about anything, in any application without having to reach for a new window. A keyboard shortcut activates the assistant, to which you say your question aloud. The assistant will give the answer in text and audio as a response.

2023.12.13
github.comGitHub repositoryTLDR Web Dev

NAME CHECKER

An online tool that checks package registries to see if a name is used by another project in various package managers.

2023.12.13
namechecker.vercel.appToolTLDR Web Dev

PERSONALIZING THE DOORDASH RETAIL STORE PAGE EXPERIENCE

DoorDash built a personalized shopping experience for its new business vertical stores using machine learning. Its recommendation framework has three main categories of collections: operator-generated, rules-based personalized, and ML-based personalized. The framework includes components like collection generation, collection retrieval, horizontal item ranking, item post-processing, collection ranking, and collection post-processing to offer personalized recommendations effectively.

2023.12.13
doordash.engineering9 min readTLDR Web Dev

GITHUB PRODUCTIVITY AT 30 BILLION MESSAGES PER DAY

GitHub is used by 85% of developers and serves nearly 400 million repositories with a guaranteed uptime of 99.9%. GitHub supports 30 billion messages per day and therefore, increasing productivity within the company increases overall productivity for a vast majority of software developers. This article takes a look at the work done by Akshaya Aradhya, senior director of engineering at GitHub, and how she helps her team increase their productivity across the company.

2023.12.13
thenewstack.io10 min readTLDR Web Dev

Sponsored

RAMPANT CLOUD ACTIVITY?

Cloud risk can grow faster than your AWS bill (true story). That’s why Wiz partnered with Wiley to create the AWS SECURITY FOR DUMMIES EBOOK. This free pdf contains 46 pages of expert tips to harden your AWS environment, including:

2023.12.12
wiz.ioSponsorTLDR

IOS 17.2 ARRIVES WITH NEW JOURNAL APP AND SPATIAL VIDEO CAPTURE SUPPORT

iOS 17.2 is now available for compatible devices. The update includes a new Journal app designed to get users writing about events in their lives with prompts that draw from data on their phones. The update also adds the option to record spatial videos. Other features include the ability to set the iPhone 15 Pro's Action Button to translate phrases, new Weather widgets, new features for the Messages app, Qi2 support for the iPhone 13 and 14, and more. Apple also recently released watchOS 10.2 and tvOS 17.2.

2023.12.12
theverge.com3 min readTLDR

MISTRAL AI, A PARIS-BASED OPENAI RIVAL, CLOSED ITS $415 MILLION FUNDING ROUND

Mistral AI has closed its Series A funding round, raising around $415 million. The company, which is valued at roughly $2 billion, is opening up its commercial platform today. The platform will allow developers to access the Mistral 7B model, Mistral-medium, and the new Mixtral 8x7B model, which uses a router network to process input tokens and choose the most apt group of parameters to give answers. Mixtral 8x7B has been released under the Apache 2.0 license and is available as a free download.

2023.12.12
techcrunch.com3 min readTLDR

THE RACE BETWEEN INTEL, SAMSUNG, AND TSMC TO SHIP THE FIRST 2 NM CHIP

The world's leading semiconductor companies are racing to make 2-nanometer processor chips to power the next generation of technologies. Smaller transistors mean less energy consumption and higher speeds. The chip industry pulled in well over $500 billion in global chip sales last year and demand is projected to grow further. Several companies say they will be ready to ship 2nm ships in 2025.

2023.12.12
arstechnica.com7 min readTLDR

GM’S HYDROGEN ‘POWER CUBES’ WILL BE USED IN CEMENT MIXERS AND TERMINAL TRACTORS

General Motors will use Autocar's hydrogen fuel-cell technology to power a range of its heavy-duty work vehicles. Hydrotec power cubes are lightweight, enabling large payloads, excellent range, and quick refueling. The first vehicles with the power cubes will go into production in 2026. Hydrotec technology is currently being sold to commercial and military customers, but there are plans to offer residential versions of the technology in the future.

2023.12.12
theverge.com2 min readTLDR

BIOME

Biome is a performant toolchain for web projects that aims to provide developers with tools for maintaining the health of their projects. It features a fast formatter and performant linter for JavaScript, TypeScript, and JSX. Biome is designed to be used interactively within an editor. It can format and lint malformed code as it is being written.

2023.12.12
github.comGitHub repositoryTLDR

HUH?

Huh? is a simple library for building interactive forms and prompts in the terminal. It can be used as a standalone application or be integrated into Bubble Tea applications. Huh? contains a first-class accessible mode for screen readers. Examples of how to use the library are available.

2023.12.12
github.comGitHub repositoryTLDR

WHY ARE WE SURPRISED THAT STARTUPS ARE SO FREAKING HARD?

Nvidia's CEO Jensen Huang recently said that if he could be 30 again, he wouldn't start a company as the work he put into Nvidia wasn't worth it. This article explores Huang's statement from a founder's perspective with a focus on figuring out why it's such a surprise to people that building a startup would be hard. No matter how successful a company is, nearly every founder says that the work was much harder than they thought it would be. This doesn't happen in fields like sports, where athletes work hard for decades but they are never surprised about how hard it was.

2023.12.12
every.to10 min readTLDR

GOOGLE PROVIDES INSIGHT INTO THE TOP SEARCH TRENDS OF 2023, AND THE LAST 25 YEARS

Google's latest 'Year in Search' overview highlights the most searched topics over the past year. It provides global and regional insights as well as interactive displays, games, and more. Google added a new element dedicated to historic Search trends to celebrate 25 years of Search. It allows users to see how things have evolved over the years and how that's reflected in the top trends.

2023.12.12
socialmediatoday.com3 min readTLDR

Sponsored

WHAT’S NEW IN AWS INNOVATION?

The AWS innovation Ambassadors Podcast showcases use cases such as generative AI, cloud cost optimization, scaling workloads, and more.

2023.12.12
aws.amazon.comSponsorTLDR

MARMOT

Marmot is a distributed SQLite replicator that creates robust replications between nodes by building on top of fault-tolerant NATS JetStream.

2023.12.12
github.comGitHub repositoryTLDR

THE HIDDEN DEPTHS OF THE INPUT ELEMENT

This article looks beyond the various types that HTML's input element can embody to lesser-known attributes that make the element more usable, accessible, and applicable to more situations.

2023.12.12
htmhell.dev11 min readTLDR

Sponsored

OPPORTUNITY

Just last month, an iconic Picasso painting shattered expectations when it sold for $139 million at auction. Impressive, considering it was purchased for _around $1 million _in the late 1960’s. But there’s a surprising group of investors also celebrating this sale: 61,000 everyday users of one investment platform [5].

2023.12.12
masterworks.comSponsorTLDR AI

META’S AUDIO FOUNDATION MODEL

Meta recently teased a foundation AI model for audio. It has released the paper along with more samples and compelling demos. The main goal of the project was to generate controllable audio content with styles all from the same model.

2023.12.12
ai.meta.com8 min readTLDR AI

MICROSOFT AND LABOR UNIONS FORM ‘HISTORIC’ ALLIANCE ON AI

Microsoft is partnering with the American Federation of Labor and Congress of Industrial Organizations to discuss AI's impact on workers and to provide AI training from winter 2024. This groundbreaking alliance aims to involve workers in AI development and policy shaping and includes a neutrality agreement to ease unionization at Microsoft.

2023.12.12
finance.yahoo.com3 min readTLDR AI

EU REACHES DEAL ON THE WORLD’S FIRST COMPREHENSIVE AI RULES

The European Union has reached a tentative agreement on the world's first comprehensive artificial intelligence legislation, known as the Artificial Intelligence Act. The deal, achieved after lengthy negotiations, addresses various aspects of AI, including generative AI and law enforcement's use of facial recognition, and sets a legal framework for AI usage, with potential global implications for AI regulation and development.

2023.12.12
apnews.com3 min readTLDR AI

BIOCLIP

A vision model designed for biology applications. It outperforms OpenAI's clip by almost 20% on specific bio tasks. A 10m paired images and text training set is also available.

2023.12.12
imageomics.github.io8 min readTLDR AI

ENHANCING RL ALGORITHMS WITH STATE-ADAPTIVE BALANCING

Researchers introduce FamO2O, a framework designed to boost the capabilities of current offline-to-online RL algorithms by determining the best balance for improvement and constraints based on the state.

2023.12.12
shenzhi-wang.github.io3 min readTLDR AI

BRICKSLLM

BricksLLM is a cloud-native AI gateway written in Go that serves as a proxy to OpenAI and lets you create API keys that have rate limits, cost limits, and TTLs.

2023.12.12
github.comGitHub repositoryTLDR AI

PYTORCH 2 INTERNALS

A great recent talk about all the new things in Pytorch 2 like Dynamo, Instructor, and ExecuTorch.

2023.12.12
slideshare.net73 min readTLDR AI

HUGGINGFACE GETS AMD SUPPORT

Transformers 4.36.0 has been released with the new Mistral model, AMD support, safetensors as default, and more!

2023.12.12
github.com5 min readTLDR AI

AI TAMAGO

A local LLM pet with a web front end.

2023.12.12
github.comGitHub repositoryTLDR AI