[Dev Catch Up # 71] - Grok 4 Controversy, OpenAI-Windsurf Deal Off, Storage Types, Forward Vs Reverse Proxy, Backlog.md, DB Connection Pools, Comet AI browser, How Discord handles Load, and much more!
Bringing devs up to speed on the latest dev news from the trends including, a bunch of exciting developments and articles
Welcome to the 71st edition of DevShorts, Dev Catch Up!
For those who joined recently or are reading Dev Catch Up for the first time, I write about developer stories and open source, partly based on my work and experience interacting with people all over the globe.
Thanks for reading Dev Shorts! Subscribe for free to receive new posts and support my work.
Some recent issues from Dev Catch up:
Join 8000+ developers to hear stories from Open source and technology.
Must Read
Grok 4 was released last week amid ongoing controversy as xAI Groq bot started making extreme and sometimes offensive statements on X. xAI quickly stepped in, and removed the offending posts, and said they’re working to fix the Grok model. Read the post for details on Grok 4 release.
A month back, we posted about the OpenAI-Windsurf deal. Now, that deal appears to be off, and Google has hired Windsurf’s CEO and most of the team. Read the full details here.
Every product is built with different storage options based on the data it handles. This post explains file, object, and block storage with clear examples. It helps you choose what best fits your needs. Worth checking the substack post before you decide on storage for your next build.
If you’re building an app, it’s important to know how to manage traffic and secure your application. This post explains forward and reverse proxies with simple diagrams and examples making the concepts easy to understand. Check the substack post for more details
OSS Highlight of the Week
This week we are featuring Backlog.md. Backlog.md turns any folder with a Git repo into a project board. It lets you manage tasks for any Git project using just the command line. You can also view your tasks as a kanban board and use AI to help organize them. Check out Backlog.md on GitHub.
Good to know
Handling database connection is important, as leaving them open or mismanaging them can slow down your app and overload the server. This post explains why connection pools are important, how max open connections affect performance. Check this post to keep your database healthy.
MCP is the growing standard for connecting LLMs to external tools, but it also opens up new risks. This post shows how Supabase’s MCP integration can be exploited. Prompt injection can leak your private SQL data. Read this post if you are using database MCP server with LLM.
OpenAI is getting ready to release their Open Weight model. This means we will be able to access and use the model weights directly. It could make it easier for developers and researchers to fine-tune and build on top of it. Check Sam Altman’s Twitter post for details.
Cloudflare has launched a new feature that allows you to block AI bots, scrapers, and crawlers with just one click. It helps sites stop AI bots from taking their content. We need to wait and see how this is going to affect existing AI services and the broader web. Read the Cloudflare announcement for more details.
Notable FYIs
Perplexity announced Comet, an AI-powered web browser, a few months back. It’s now released and available for Perplexity Max subscribers. Read the official announcement here. There’s also buzz that OpenAI may be launching its own web browser soon. Read about the OpenAI browser news here.
When your app updates the database, you expect the data to be saved. But crashes can happen before data is written to DB server. This post explains how Write-Ahead Logs (WAL) in PostgreSQL make sure your data isn’t lost after a failure. Check this post to learn more about WAL and data reliability.
When working with LLMs, you might still get invalid JSON as response even with strict prompting. jsonrepair is a tool that fixes broken JSON so you can keep your pipeline running smoothly. Useful for cleaning up API responses or broken JSON from LLM. Check out jsonrepair on GitHub.
We all know Discord servers, but most of us don’t know how they handle trillions of messages at scale. This post details how Discord managed growing traffic and performance, covering their migration and technical choices. A good post to read if you’re interested in backend scaling.