This was such a satisfying process... especially finding ways to explain Donna without any jargon. I love the image of a hand holding the elevator. Can't wait to realease the implementations across the product. Thanks to all our users who helped in the process!!!

Is software engineering daily podcast hacked or something wrong with the people behind it? Seems like a sudden collapse of reality 🤔😕🤷‍♀️

Am I a bad person for giving my name as #10 at #Starbucks just to screw up their fake "yelling people's names sounds better" corporate policy? They would have gotten my name wrong anyways 🤷‍♂️

Now THIS is #RadicalCandor ❤️ wish people practiced this more. Although fictional…still my new goto example when giving someone “permission” to be brutally honest with me. youtube.com/shorts/DKm_OA-…

Reason 23 I love #rustlang…no nulls. Most code isn’t written. Most code isn’t read. Most code is re-written by someone who barely read it. Compared to Go,Java,TS explicit options help trade fewer LoC and some readability for confidence when refactoring old code.

Köpte han strumporna hos #IKEA också?

Did you know VS @Code added floating windows back in November 2023? That means you can move your chats to another screen!

Announcing The LLM Litmus Test! 📊 This is the 1st AI experiments in the new Sourcegraph Labs! We believe greatly in LLM choice, but we often get the question: Which LLM is right for me? With this new tool, you can: 🔴 Compare+contrast different LLMs 🔵 Try GPT-4 Turbo vs

Wow. We just enabled GPT-4o in Cody and the first zero-shot code generation in a big existing codebase just...works? No red squigglies! It is *really* good at learning from the context Cody provides from our specific codebase.

If you want a visceral comparison of how much faster and better GPT-4o is, here it side-by-side with GPT-4 Turbo, using @sourcegraph-provided code context. On the right, Turbo is slower and fails to incorporate the context, giving a generic—and incorrect—answer. On the right, 4o

It can be annoying to switch between tools when coding: open the issue in Linear or Atlassian Jira, jump into Figma, look up the error on Sentry, and so on. With OpenCtx, an open standard for bringing info from all these tools into anywhere you read code, you don't have to.

OpenCtx is available in the pre-release branch for Cody AI for VS Code. Learn more about the release here: sourcegraph.com/blog/openctx-a… See it in action 👇

Cody in JetBrains lets you tag multiple repositories to include in context for questions that may span multiple codebases:

Honestly, the 6.0 version is such an improvement across the board. If you tried Cody in JB before and were (in all honesty rightfully) disappointed, I'd give it another shot. The updates have also set the foundation for a lot more cool stuff to come down the pipeline very soon!

Reading feedback like this makes my Sunday!

Thanks YK aka CS Dojo 📺🐦, the amazing Sourcegraph team who put on the AI dev tools event, and everyone who came out on a Tuesday night! Had a great time and met so many awesome builders. SF is back!

I fucking love #Typescript! A bit of help from type-fest and template literals...🪄 you've built a string format validation with helpful errors, at compile time 🤯 PR for github.com/sindresorhus/t… in 3...2...1... Any way you can benchmark types for testing performance?

We're excited to be partnering with Cloudflare, Jam.dev, Warp, @Clerkdev, LogRocket, Prisma Postgres, Intercom and Mixpanel on the Dev Starter Pack, a bundle that makes first-class dev tools available at steep discounts for startups! devstarterpack.io

Just heard from a customer that soon after rolling out Sourcegraph to all their devs, their issue-to-PR time got 15% faster. Love to hear it!

Don’t know why, but this tickled my brain 😂