othmanosx

Not just VSCode, but other pieces of software I always use, not sure if that's related to AI, but this genuinely upsets me.

I would expect things to be better as AI helps accelerate development and bug reports and whatnot, but it looks like the opposite is actually happening... which is concerning.

Show HN: Pyor - A better alternative to Github for reviewing PRs

since the introduction of AI, the bottleneck for shipping code shifted from producing code to reviewing it, and the PR reviewing experience on Github is lackluster, leaving a lot to be desired.

my team now ships too much code and it sits in review for too long, review AI code is not easy and we need a better tool to review code, especially on large PRs.

So I built Pyor, a website and a performant desktop app to review code at least 3x faster compared to Github (from experience)

As a web dev, I compiled all of my painpoints into this project, what would make a review easy (and even fun) for me, I integrated tools to group noise from actual important bits to focus on reviewing, that speeds up reviewing time immensely, especially on migration PRs where most is mechanical name changes.

Try it out at https://www.pyor.review and let me know what you think!

it’s free!

For the past year, I’ve been building GM Pro — a Chrome extension that upgrades the chat experience inside Google Meet.

It started simple: reactions, replies, mentions, dark mode for chat. Then I added auto-join, auto-mute, transcription tools, lobby notifications, attendee shuffling. Basically all the things you wish Meet chat had by default.

People loved it. 5-star reviews. Steady installs. Real usage.

And then, after many years of lackluster chat, Google announced they’re integrating Meet chat directly with Google Chat — persistent conversations, reactions, file sharing, the works.

Which means… the exact surface area I built on top of is becoming a first-party feature.

On one hand, this validates the idea. The direction was right. The need was real.

On the other hand, platform risk just punched me in the face.

When you build on top of a giant platform, you’re effectively prototyping features for them. If the feature works, they absorb it. If it doesn’t, you disappear quietly. Either way, they win.

Now I’m thinking through:

Do I pivot to power-user tools Google won’t prioritize?

Double down on automation and workflow features?

Move away from chat and toward meeting intelligence?

Or accept that consumer Chrome extensions sitting on core UI are inherently fragile?

Curious how other builders here think about platform dependency.

Have you ever had your product “Sherlocked” by the platform you’re building on? What did you do next?

Starting a startup is tough, and I learned that the hard way. My first attempt, Brieflytics, was born out of a problem I saw in my own development team—so many repetitive daily tasks spread across different tools like GitHub, Jira, and Slack. I envisioned a platform that would aggregate this information and generate useful analytics and summaries for teams.

Excited by the idea, I quickly put together a landing page to showcase the product and its planned features. I started reaching out to people online, gathering feedback, and trying to gauge interest.

That’s when reality hit me: only one person signed up for the newsletter. Just one. It was a painful but clear signal—I hadn’t validated the idea properly. I was building something I thought people needed, not something they actually wanted. That was my first major lesson.

Fast forward a bit, and during our usual Google Meet calls at work, I noticed something interesting. People weren’t happy with Google Meet. They were missing features they loved from Zoom, better chat UI, reactions, replies, persisted messages, transcriptions, and more. That was my cue. This time, instead of rushing to build a landing page, I built a simple prototype and got my team to use it. They loved it.

That prototype evolved into my new startup, GM-Pro. a Chrome Extension that enhances Google Meet with features like a better chat UI, reactions, persisted messages, transcriptions, and many more improvements. If you want to try it, you can install the extension from the Chrome Store.

This experience was a complete shift from my first startup. I had validation before investing too much time. I had real users giving me real feedback. Now, I’m focusing on refining and expanding this product, knowing there’s actual demand for it.

Lesson learned: don’t just assume a problem exists—make sure people actually care about solving it before you build. Validation first, then execution.

Has anyone else gone through a similar startup failure? How did you pivot?