{@thread.name}
zachdaniel:
.terris:
This is fantastic, especially the parts where Zach is improvising on the fly. I would watch a video where the project is built from scratch. I also don’t think, if I was new to Ash, I would appreciate the meta coding. It just looks too easy and also too foreign.
zachdaniel:
What do you mean by meta coding?
.terris:
The macros / DSL. You added an action on the fly, and it’s suddenly callable. There’s hardly any code. Skeptical developers don’t like magic - we are used to seeing a lot of imperative code to get small amounts of functionality. Ash is similar to something that Joe Armstrong only talked about. He didn’t live long enough to get to experiment with his biggest ideas. Who knows what he would have come up with but I think Ash is pretty close, maybe even better since Erlang and Elixir are so different.
zachdaniel:
I think there is still a long way to go. Probably a lifetimes worth of work to make it what I think it can be really
zachdaniel:
Just need to keep going. I appreciate the kind words 🙇
.terris:
Ash is going in the right direction on the right albeit long road. That’s not ego stroking. Ash has real users.
.terris:
By far, the bumps in the road today mostly involve authentication. That is to be expected. It’s a new and complex feature with endless permutations. It’s what everyone wants. It’s literally what gets used first. The team will get past this - it’s not a “nice to have.” Authentication out of the box is the main reason why folks (like me) open their wallets for closed source starter kits. This will finally stop. When we spend money, especially for software licenses, it should be for extra-ordinary functionality (Stripe payments? Shopify integration?), not something that every nontrivial app needs. However, I send Zach a cup of coffee each month, as should anyone else who benefits from the Ash Framework. Free code, tutoring, and ultra-responsive support bordering on consulting is a deal you won’t find anywhere else.