Notices by Ingo Paschke (ipaschke@cyberplace.social)

@dalias @mos_8502 It's his project, he is the only contributor, he hasn't changed the license terms.

Also: "It's now bad software because there's a Claude.md in the directory" is not a really convincing argument unless you can point to specific problems with the codebase.

Sorry, this is all just excuses for "let's all dogpile on this person so he feels bad about his passion project".

It's just shitty behaviour. Hate on AI companies, not individuals.

@mos_8502 General rule for giving feedback for what is clearly a passion project: If you don't like if for any reason, just ignore it.

Nethack for Atari now actually runs on real Atari hardware after some struggles.

In other news, I have bought an Atari Falcon!

#atari #retrocomputing #nethack #roguelike

@root42 I'm totally prepared for the metric minute!

Nethack 3.7 development snapshot fully playable on an Amiga 1000 with HDD + 8 MB RAM. #nethack #retroComputing #amiga

The second installment in "Update all the orphaned ports!": Frotz for the Amiga.

This great Amiga port from 2007 now uses the newest Z-Code engine from Frotz 2.5.5 (2025).

Should be available soon on aminet.

#Amiga #interactivefiction #textadventure #retroComputing #infocom

If you have an Amiga with at least 4 MB of RAM, you can help me test this early build of Nethack-3.6.7

https://drive.proton.me/urls/2SGX26WPFC#k1Wa2HUjn62T

Unpack the archive, assign nethack: to that directory, make sure to have a resonably large stack size (like "stack 65535") and run NetHack.

Users with RTG graphics can set "amitile" as the window type in Nethack.cnf.

#nethack #amiga #retroComputing

@GossiTheDog Was his apartment also prompted into existence or what is going on there!?

@GossiTheDog Next up:

https://en.wikipedia.org/wiki/Tank_biathlon

@GossiTheDog Most of the neat Copilot use cases can probably be run locally now or in the near future with NPUs becoming more common in client PCs. Deepseek and - I hate to admit - Grok have demonstrated that you can train state of the art models without ultramassive datacenters. So I can't really see how MS' massive datacenter investment can even pay off.