antidamage

This is my exact setup as well and dear lord gemma is absolutely batshit insane. I'm trying to get a self-reflection and confidence loop going now, but it does feel like it's not the local resources, it's the limits of the training. Dedicated coding or dedicated real-world task models would be a good optimisation.

The roadblock to this is you seem to have to build it yourself. I've noted that none of the current cloud models are very good at building a replacement for themselves, and there's significant work that needs to be done to make a local LLM reliable in any way. I haven't found a single standalone package that makes setting them up easy. Sure, I can run Hermes Agent and a model, but getting the self-reflection loop in and all of the other stuff the need to actually be good? I'm still at it, trying to get anything to work reliably and factually.

Runescape has a similar thing going on. What everyone's not understanding is not that improvements are bad, lots of the player base also want to see games evolve. It's just that a significant proportion of the audience also like The Old Thing and they're often more engaged as an audience in the first place.

You hear this story over and over about every kind of software.

There are two audiences every successful developer needs to cater to: the "I wish this did X, I want new features" side and the "I liked it the way it was" crowd and they're distinctly different groups.

For a long time I produced a popular technical art asset for video games and even I realised I needed to include every single version of the tool with every single installation. If a developer has to go and find out how to get "the right version" at all then I'm 90% likely about to lose a user.

Focusing on the "we did this right just keep going like that" and really understanding WHAT you did right and WHAT people like is really important. It's really hard to be impartial when you made the world rather than consumed it, but always take the win.

It's very familiar. My father was abusive towards me and so I tried to disturb him as little as possible. He was extremely supportive of my younger brother, however, and we both turned out very differently. He turns to my parents constantly for help (and receives it) and the few times I have I've only been brushed off, and so I've always done everything myself, even if it meant spending long periods living in poverty with no footholds upward.

Years later I'm nearly 50 and have transitioned to female. My now divorced mother decided she likes having a daughter and so suddenly all of this support has materialised and with presumably less time to live the life I missed out on I'm making a conscious choice to start being a little dependent on it and shortcut some things that would have taken longer otherwise.

It's hard to describe but it feels like I'm living a more normal, less marginalised life now. I definitely have more friends, where previously I chose to avoid having any because I felt like having friends meant sharing burdens that were mine alone to carry.

I also see a lot of very similar behaviour in the wider trans community where most people's axis of behaviour revolves around some facet of not having support. We don't all handle it in a healthy way, but I now do my best to help my community find stability and adjust their expectations to having better outcomes.

Different people engage in different ways. I think OP is telling us what got them over that initial habit adoption hurdle. Sometimes spending money on something makes you use it.

My example case is I paid for a year of gym membership up front so I could feel the total I'd spent more tangibly. That got me to go regularly, even if I was just showing up to recovery stretching at first. Now I enjoy going and I look forward to it, and I didn't need a PT to motivate me this time.

It looks like it has a 3d draping view, but it didn't seem very good yet. Check out Marvellous Designer, that's what anyone doing digital-first uses.

Drapey clothing is probably the easiest to freehand without a pattern though. It's accurate fits that need more measuring, planning, temporary stitching, test garments, etc.

It looks like it's missing so much that you'd need even to hand-sew a pattern at home. There's no mention of interfaces or bindings.

This looks more like something for making clothing as digital content - e.g. Marvellous Designer. Possibly more straightforward even.

Edit: found interfacing. It calls it "interlining".

It's not an accurate recollection at all. In 1990 a couple of us 12 year olds snuck into the university library to use the web to look at the Marathon website. It took 5 minutes to load some trivially-sized gifs and a tiny amount of HTML. They had a pretty decent connection for the day.

Web pages took a minute to load, now we're optimising them for instant response.

I'm imagining that this is the work of the Boy Genuis

I loved System 7 for its simplicity yet all of the potential it had for individual developers.

Hypercard was absolutely dope as an entry-level programming environment.

Teletext existed in the 80s and was widely in use, so we'd have some kind of information network.

BBSes existed at the same time and if you were into BBSes you were obsessive about it.

It's not that different from how some creative Mac games were doing 3d lighting on 2d textures prior to 3d accelerated hardware being available. The neat part here is that it runs on a Gameboy Colour.

Check out Rustler Grand Theft Horse on the Epic Games Store, it's the same top-down format, very GTA-like but set in medieval times, yet has all of the modern banter. It's so great.

A couple of months ago I had to write a CDL-based triangulator to solve a use case where ear-clipping doesn't support the shapes we had.

We had no AI policy at the time so I had to read up on CDL and implement it by hand. The concept is straight-forward and I also targeted regularity as acceptance criteria for the mesher, but making it optimal was hard.

I ended up having to park it after the ticket ran out of time, but now we have an AI policy this was the first problem I gave it. What it put out was similar but better structured and more informed.

I worry a little that AI will stunt our problem-solving in 20-30 years, we still need new algorithms, even when ML is capable of producing a model that can do the same thing. But right now it's much better at the things we've already done than we are.