Raph was the lead game designer on SWTOR a game that was way ahead of it's time and one of the most enjoyable sandbox mmorpg's I've ever played. I'm working on a new game that will take inspiration from lessons learned there.
I remember when Raph was working on Metaplace[1], which was a kid-targeting, programmable (Lua dialect), virtual world/user generated content factory that was contemporary to the launch of Roblox ca. 2006-2007. I wonder quite often what things might be like if Metaplace had gotten to the scale and scope that Roblox wound up achieving.
What was interesting/worked about it's design (and why did the players care?[1])
Was it resilient to the, uh, many, many well-documented problems that the genre pushes players/itself into?
---
[1] There's a lot of ideas in this space that sound interesting on paper to nerds bikeshedding, but often fall flat in actual implementation. I'm curious as to what were the ones that worked.
Game was SWG, not SWTOR. Launched in 2003 and was sunset in 2011 when SWTOR launched.
SWG set out to be something like Dwarf Fortress in terms of depth to the worlds physics; for example, gunsmiths could tinker on all parts of a gun and maybe get a lucky roll to unlock +N more damage or -N recoil. Same with land vehicles and bioengineered animals, droids. Parameters to noodle all the way down. Some under user control, others random to foster sense of a chaotic physical world.
As the in game object economy was entirely propped up by crafters this fostered economic PVP.
Lucasarts of 2000-2003, when the game was developed, did not understand MMO, and 3D games take much longer than 2D adventure games and shoved it out the door 2 years too early.
It also suffered from 90s OOP heavy software development patterns. Devs had difficulty managing it and updating over the years.
Ultimately it failed at being a Star Wars game. PVE was just "kill a nest of bugs" and failed to leverage storylines and characters. Players with nothing else to do ended up ruling the economy or whatever. Could have made them compete against Star Wars power brokers, IMO. Jabba sabotaged your factory, or something. Once a player was kitted out they had nothing to do.
Some have spent the last 10+ years implementing a server emulator, various tools and mods. An emulator built around the original release is here: https://github.com/swgemu
I tinker on a modded private server now and then. Initially added in random world events, to generate things to go do and replacing odd design decisions like mission terminals with NPC models to talk to in that seedy back alley, to foster more in world RP vibe.
When WOW launched SWG was redesigned to play more like that. Typical MBA "copy paste what they are doing" project management.
It truly was ahead of its time, I don't think any one game has come close to implementing such a rewarding group of systems and economy in an MMORPG, except maybe EVE but that is a very different game and admittedly I did not find EVE fun.
The most exciting systems to me had very little to do with combat, but especially as it pertains to this article, also couldn't be as rewarding without it. It was all the player run economies, homsteads, towns and cities, player shops, craftsman and markets. The fact that materials mined had quality which impacted item stats, on and on.
To get good gear, you had to know a guy who made it, they had to know a guy who'd mined good quality minerals, and that person may have found the minerals through another player who had prospected it.
It made sense to be part of a player city, so you could put your house in a known market area for people to visit.
It all mattered because people needed the equipment to go do the quests, and so it was a really symbiotic set of systems that made crafting and economy matter.
To me I really liked the fact that when you made your character in SWG (1 per server too), you are just a civilian. There's no light/dark side or rebellion/imperial choice to make, you're just a regular person in the galaxy. You are NOT the hero.
The skill tree system was so nice compared to the rigid class systems of other MMORPGs, too.
The fact that player towns just emerged was really cool.
It was such a shame the space expansion was so ... flat. Neither space nor ground had a storyline to follow, but space wasn't an open world, and had no real element of choice in skill paths.
I enjoyed the new aspect ships brought to crafting, and there's something special about walking around your own ship while it's in transit. But otherwise totally agree, it was kind of just space combat arenas and not much more.
I had a collector's edition 3-man transport ship, but IIRC the novelty of standing on the ship while in transit wore off before the beta ended. Cool, but too shallow on its own.
I can't figure out if the open world game was fun enough just on its own that an open space game would've been chef's kiss, or if it did need some kind of story telling too. It's too long ago to remember well enough, for me.
PVE was indeed awful. Especially given the back drop; it should have been full of adventure across the galaxy, established characters messing with players, but was merely "run here and kill 6 kobolds". NPC AI sucked.
Would love to strip from my private server, NPC generation as-is as implementation is static and does not allow dynamic responses. Replace it with modern agents to connect like players and train them to build out the world like players can.
Also started a project to make a new client using video and segmentation, gen AI to recreate initial game engine entities as Godot scenes to have full control.
Too little time for either, initial code has sat untouched for years.
> Also started a project to make a new client using video and segmentation, gen AI to recreate initial game engine entities as Godot scenes to have full control.
That sounds fascinating, I've been working in godot for a few projects now. I'd be interested to know how you would integrate the Godot scenes into the current engine, or if it would be an entirely new client.
My plan was/is entirely new client, mapped client state to SWG emulator server.
Godot is a pain given my workflow is pretty cli heavy though. Since I last touched that project I looked into switching to Wicked Engine. Just include C/C++ headers rather than Godot.
But job got interesting (am an EE in hardware development land) and I have to spend free time diving into AI model architecture to keep up. Both SWG projects have sat idle for 10-12 months now. shrug
I think the headline is a bit vague, it includes passwords as well. Does anyone know if Troy's HIBP'd site reveals the passwords to verified users? I'd like to know if my current or what generation of passwords has been breached to evaluate if I have a current or past problem with my devices.
That's where I really appreciated Android's ability to sub-out its entire UI to something totally customizable, accessible and able to be locked down. Drag my face contacts of myself and siblings onto the home screen so Mom can just click a face and it calls her contact. On the desktop, Windows is terrible for seniors, MacOS too. iPad can be decent but there are alot of areas where seniors get confuddled and have to call someone in.
I had a dozen domains and projects on the shelf for years and now 8 of them have significant active development. I've already deployed 2 sites to production. My github activity is lighting up like a Christmas tree.
In the mid-2000s I was doing some consulting in middle America and noticed a bunch of Indian IT workers staying at the same hotel I was staying at. Since, I was Indian-origin, we engaged in some small-chat. They were working for a now defunct telecom giant (but contracted by one of the big Indian consultancies) on various software projects. There were there on not, H1Bs, but another class of short-term business-meeting only visas. They told me every 3 months a whole new team would cycle onto the project with a fresh set of business visas and the rest would cycle back home, rinse-repeat. They weren't allowed to "work" per say, and some of their colleagues got rejected at the border by more savvy border agents. But for every 1 rejection, 20 of them would get through.
The pivot is likely because there's more VC dollars there.
It is a handy AI-cli for any terminal. I've been using the "terminal" app for a few months and found it was a very competent coding tool. I kept giving feedback to the team that they should "beef up" the coding side because until Claude Code this was my daily driver for writing code until Opus 4. The interface still is a bit janky because i think it's trying to predict whether you're typing a console command or talking to it for new prompt (it tries to dynamical assess that but often enough it crosses the streams). Regardless, I highly recommend checking it out, I've had some great success with it.
Wonderful, timely article. It sounds like a hybrid approach might produce good results: Using ChatGPT-5 for planning/analysis and using Claude for execution.
Another option would be to add something to your AGENTS.md or whatever giving examples of the kind of code organization you want. You could ask Claude to explain its approach in terms explicitly that GPT-5 can understand. GPT-5 seems much more sensitive in its responsiveness to instructions. My sense is that in the long run this will be really nice, but that the current prompts in these mainstream LLM coding tools are designed for models with a different style of responsiveness to instructions.
It's morbidly impressive how much Sam Altman is a sociopath's sociopath, knowing the right things to say to ensnare his fellow sociopaths into his trap.
I just had the same thing happen. Some comprehensive tests were failing, and it decide to write a simple test instead rather than investigate why these more complicated tests were failing. I wonder if the team is trying to save compute by urging it to complete tasks more quickly! Claude seems to be under a compute crunch as often I get API timeouts/errors.
reply