The Future

So, I fired up the KHWar internal bits and it's firing on all cylinders. No idea how it still works despite having been migrated between 2 different servers with 0 updates, but it works, albeit slowly. 

I plan on migrating it out of KH and into its own thing. From there, I can restructure, improve, modernize, etc. After that, probably build out a mobile app that talks to the backend, push it to Android, get some whales to fund things, and keep us going for a while longer.

Anyone out there know mobile dev? :D

I'm actually pretty curious why this website costs like 100/month

That's after scaling down as much stuff as possible. Keep in mind, this is just raw VM hosting cost, not include S3 storage and domains.

KH was designed on a distributed platform with multiple redundancies. Gutting all of that so that it could be dropped down to a single instance would require making code changes, and considering the code here hasn't been touched in over a decade, that's not doable without a hefty time investment that I just don't have anymore.

Back in our prime, we were loading up a stack of servers faily well. Since the death of flash, I've paired back as much redundency as possible without requiring a rewrite. It's not like we're seeing millions of hits a month a month anymore :P

If you put some work into a good setup and spec for Claude Code, it should be able to rewrite or refactor everything. $20 and 30 mins of work every day for a month could cut the costs to almost nothing and might be a fun experiment :P

> Anyone out there know mobile dev? :D

Just gonna mention Claude again. He's definitely better at mobile dev than me.

AI coding is a joke and a half. I've used it, and am forced to use it for part of my job.

Refactoring a decade+ old code is just something that isn't going to happen any time soon. If we were a bunch of bullshit nodejs, who cares, it'd probably work out the box, but we're old school PHP.

Getting good results with agentic development requires quite a bit of configuration and experimentation at the moment. It's a lot faster if you have good documentation/specs, including all the knowledge you'd share through onboarding a new developer - one who's a genius at some things but get other, rather basic things completely wrong.

Almost no one writes all that stuff down so there's a lot of upfront cost to get the most out of AI. That said, I work on a big Scala codebase and our AI config reached a point a month a two ago where agentic development started to give significant productivity boosts.

The big thing with KH is that we have 3 entirely different generations of code running.

1. Flat php files with includes
2. Auto-loaded PHP via a custom auto-loader
3. Properly routed auto-loads with Composer backing.

More so, it's at least a decade behind on the newest end.

Hello everyone!   I only recently discovered the full history behind KongHack, and I wanted to say that what you built here is worth preserving. This is more than a collection of tools and old forum posts—it is part of the history of Flash games, Kongregate and the old internet.   Even if rebuilding everything is unrealistic, perhaps the community could help preserve the site as a static archive, document the old code, organize important downloads and keep a small space alive through Discord or GitHub. Small contributions from different people may be more manageable than placing the entire burden on one developer.   Thank you to everyone who created tutorials, researched games, maintained the servers and kept KongHack online for all these years. Projects do not truly disappear when their knowledge and community are preserved.   Greetings from Brazil. I hope this is not the end of KongHack, but the beginning of a new chapter—even if it is a smaller one.

Are the currently newest versions of AI up to the task now?

Potentially. The problem I'm facing now is cost. But I'm working on it.

Consider KH 3.0 in the works.

Phase 1: Modernize the code. Bring everything inline. Eliminate the old bullshit access sytems. Gut the junk that doesn't work and hasn't worked in years.
Phase 2: Modernize infrastructure. We don't need Kubernetes, but having a new server with docker and compose so we can make things a bit more durable and portable is what I'm shooting for.
Phase 3: Clean the ever loving shit out of the database, then migrate to the newest MariaDB.
Phase 4: Clean/remove legacy applications. The KHUT is most likely completely dead. I think the last time I had it installed was on Windows 7. Kolonel disappeared years ago and hasn't been seen since. Hell, on steam, I'm seeing "Last online 10 years ago".
Phase 5: Overhaul the email system. Clean out all of the old cruft, get something clean and straight forward working.
Phase 6: Do something with ADR
Phase 7: Do something with KHWar.
Phase 8: Official 3.0 launch and announcement.
 

Testing a new post