Urgent.News

What's breaking now, across thousands of outlets.

AI

This one weird trick made Claude Code so much more reliable, and nobody tells you to do it

Set a goalpost for the bot

This one weird trick made Claude Code so much more reliable, and nobody tells you to do it

A developer named Mayuki has created a modern recreation of the Windows 3.1 Program Manager using Claude, an artificial intelligence tool. This recreation, named ReProgman, offers a nostalgic experience for users who want to relive the Windows 3.1 era on their current Windows 11 or macOS systems. The project, which is named after the original Windows 3.1 shell progman.exe, was built by Mayuki using the artificial intelligence tool Claude.

The initial version of ReProgman was completed in under an hour and was described as "pretty decent," but quickly became boring to use. The final releases of ReProgman weigh between 12 and 16MB, making them easy to download and use without the need for installation. To build and run ReProgman, users need to have the .NET SDK version 10.0 or higher installed on their systems.

The code for ReProgman has been released on GitHub under the MIT license. While ReProgman provides a quick trip down memory lane for those who remember using Windows 3.1, others may prefer to experience the original MS-DOS/Windows 3.1 combination on real hardware. In April, another enthusiast demonstrated Windows 3.1X running on a modern system with a Ryzen 9 9900X and RTX 5060 Ti graphics card.

Written by urgent.news from Tom's Hardware's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Read the original at xda-developers.com →

More in AI

You Thought Training an AI Model Was Enough. Then There's Post-Training

Hello, I'm Rijul, and I'm building LiveReview — a blast-radius aware AI code review built for your business-critical systems.

  • Post-training techniques essential for AI model's helpfulness
  • Supervised Fine-Tuning teaches model response format and instructions
  • Direct Preference Optimization and Proximal Policy Optimization optimize response preference

More from Thursday 10 September →