Adopting such a niche UI framework sounds like spending a lot of innovation tokens for questionable gain.
> Ideally we should be able to interact with the SPA router and the various reactivity states of the dashboard from an external script. And this unfortunately is very difficult to be implemented when using a compiler framework like Svelte because we'll have to force on everyone an additional UI Node.js build step when extending PocketBase and I want to avoid that.
My initial reaction was “huh, modern frontend code is always compiled” (bundling, TypeScript transpilation). But I think they mean that with Svelte they will require a recompilation of the entire app when adding a plugin, which indeed seems a bit cumbersome, but honestly also potentially not that bad since as I understand it Pocketbase is a development tool anyway.
I don’t have a stake in or intimate knowledge of the project, but from the outside it seems a bit weird to spend so much effort on this constraint.
Now, instead of users running a Svelte build step every once in a while when extending the Pocketbase UI, extension authors will have to learn a whole non-standard framework.
In a perfect world this would mean that higher developer load is traded for better user experience, but I think it will probably result in less extension developer interest instead - potentially a bad tradeoff, especially when it’ll often be the same people.
> Ideally we should be able to interact with the SPA router and the various reactivity states of the dashboard from an external script. And this unfortunately is very difficult to be implemented when using a compiler framework like Svelte because we'll have to force on everyone an additional UI Node.js build step when extending PocketBase and I want to avoid that.
Whew, that is a very daunting intimidating constraint, if that's true as it sounds. Feels incredibly unfortunate to have to leave the grid and go back pioneering in the forest, making a new framework, to work around this. I really want to imagine Svelte has some options it might be able to offer here to help this situation.
Adopting such a niche UI framework sounds like spending a lot of innovation tokens for questionable gain.
> Ideally we should be able to interact with the SPA router and the various reactivity states of the dashboard from an external script. And this unfortunately is very difficult to be implemented when using a compiler framework like Svelte because we'll have to force on everyone an additional UI Node.js build step when extending PocketBase and I want to avoid that.
My initial reaction was “huh, modern frontend code is always compiled” (bundling, TypeScript transpilation). But I think they mean that with Svelte they will require a recompilation of the entire app when adding a plugin, which indeed seems a bit cumbersome, but honestly also potentially not that bad since as I understand it Pocketbase is a development tool anyway.
I don’t have a stake in or intimate knowledge of the project, but from the outside it seems a bit weird to spend so much effort on this constraint.
Now, instead of users running a Svelte build step every once in a while when extending the Pocketbase UI, extension authors will have to learn a whole non-standard framework.
In a perfect world this would mean that higher developer load is traded for better user experience, but I think it will probably result in less extension developer interest instead - potentially a bad tradeoff, especially when it’ll often be the same people.
> Ideally we should be able to interact with the SPA router and the various reactivity states of the dashboard from an external script. And this unfortunately is very difficult to be implemented when using a compiler framework like Svelte because we'll have to force on everyone an additional UI Node.js build step when extending PocketBase and I want to avoid that.
Whew, that is a very daunting intimidating constraint, if that's true as it sounds. Feels incredibly unfortunate to have to leave the grid and go back pioneering in the forest, making a new framework, to work around this. I really want to imagine Svelte has some options it might be able to offer here to help this situation.