HN user

danschuller

469 karma

British Software Developer - UK South Coast. https://mastodon.gamedev.place/@danschuller

Current side-project: http://www.howtomakeanrpg.com/

Posts7
Comments115
View on HN

A few things that have worked well for me:

  - Like normal software engineering - building small parts in isolation, verifying they work and then bringing them together to make the whole.

  - Asking Claude to do things using sub-agents to keep the parent context clean. i.e. reviewing plans, research, code reviews
  
  - If there's ever something you can have verification on - a test suite, a screenshot with pixel reading, a certain output to match. Then with `/goal` it tends to work very well with little guidance (in my experience)
  
  - If it keeps getting things wrong, then I make it write a skill. i.e. for some reason it always had errors running playwright but would get there in the end, and it would always have trouble with auth on my NAS but with the skill it can follow the path that's worked for it in the past and not burn tokens
Accelerando (2005) 2 months ago

I'm reading Accelerando at the moment and I kept thinking about The Quantum Thief. I enjoyed the Quantum Thief more, but Accelerando feels more relevant to the current times.

Your solution is for everyone in the world to foster a degree of self-discipline they've never shown any inclination for in the past?

That doesn't sound like a scalable solution. These apps are so pervasive and their use adopted at such a young age (with a large degree of social pressure) it seems like legislation is the only way to curb these dark patterns.

Like other commentators have said these are extremely well resourced companies looking for ways to exploit human psychology at scale. It's in a walled garden, compiled app so it's you have very limited ability to modify what you're shown. End-users of apps need to be given more power over what they can be shown and if that kills companies then to me that's an acceptable tradeoff.

It's telling this sentence has the student as one to blame, when it's a structural issue and the weight of the blame rests more heavily on the shoulders of the universities.

I also find this an interesting thought. Maybe an inability to monetize ideas in the space?

I've read many many anecdotes over the years of Europeans learning or perfecting their English language skills playing classic point and click adventure games that were either unavailable localised or the localisation was bad. It feels like that's an interesting language learning entry point.

I also like the idea of an AI plugged into a glasses video camera just narrating what you're doing in your target language.

For me Mastodon misses some common workflows I have. When I find someone interesting I like to easily see who they follow with bios and then I may follow some of those. In Twitter and Bluesky this is pretty easy, in Mastodon it's hard and has a lot of friction points.

I have been trawling through the Wizardry 1 decompiled pascal source code for the Apple II. Uncovering neat little bits like:

  PROCEDURE PAUSE1;   (* P010011 *)

  BEGIN
   FOR LLBASE04 := 0 TO TIMEDLAY DO
     BEGIN
     END;
  END;
The perks of having standard hardware and a compiler that doesn't optimise away empty loops.

C++ is not a sane language for anyone with a passing familiarity. At least Rusts macros aren't literal text substitutions, a move towards the light.

The low-code, no-code niche that I see is writing shaders using visual graphs - blender, unity, unreal etc. I don't see that being replaced by LLMs in the near term. The context where you're expressing what you want is represented quite nicely by the graph. If you're saying to an LLM make me an animated, iridescent, force field like shader seems a way off to be able to get to what you really want.

There was a game on the PLATO network (circa '79 or '80) called "Oubliette" that nearly caused me to flunk out of law school. [...] Wizardry was in many ways our attempt to see if we could write a single-player game as cool as the PLATO dungeon games and cram it into a tiny machine like the Apple II.

https://games.slashdot.org/comments.pl?sid=23752&cid=2567054

Oubliette (https://howtomakeanrpg.com/r/l/g/oubliette.html) was a precursor to Avatar, so it goes a little further back than even that!

imagine a non-arrogant programmer that actually does what you want

I don't think this is going to be reality anytime soon. In order for the LLM or agent to do what you want, you'd need to be able to precisely specify what you want and that's a hard problem all on it's own. And if you were able to do that precise specification you would be the programmer.

Not say the software developer paradigm won't change but it seems very unlikely to become "make me a better google ads system" anytime soon. I could see getting to something were you are given a result by an agent and then can iterate on it, towards some solution.

I think you could do style transfer on a framebuffer before pushing it out and that might be quite interesting, especially if you can have it look correct over time.

The larger possibility is that video has shown the AI has some capacity to preserve structure overtime. The objects in the video maintain their dimensions and appearance as we would expect. This suggests there's some representation of objects or an equivalent to it. If that could be controlled more then potentially you have the ability to hold something like a game in working memory and iterate on parts of it.

A history of western philosophy? GEB and that book don't have much overlap in content and GEB is a more fun read. History of Western Philosophy is occasionally witty but is otherwise pretty dry.

GEB isn't much to do with philosophy at all, it's more focused on computing and touches on philosophy where that seems relevant.

I'm cheered to see any wider adoption of mastodon. For my very niche game engine developer follows Mastodon is now almost as good as twitter was but for mainstream or other niche interests it's got a ways to go.

One thing I find it really great for is language learning. It can create examples, explain grammar, contrast to languages you already know etc. Also I'm not a sys admin but occasionally have to do sys admin things and I ask for the shell commands I need in English, it is far far faster than using stackoverflow etc. (i.e. pulling logs from multiple servers and extracting bits) Also I don't want if I want to use a library I ask for what how I want to use it and get a small example entirely tailored to me.