[manmal]

If you are using TCA only for dependency management, I’d suggest to reconsider and put all business logic in it. In my experience, most serious efforts of wrangling SwiftUI into a complex app wind up with a solution that’s similar to TCA.

[jamil7]

Ah! We're putting all our business logic in it, I guess I misunderstood what OP was asking and assumed they meant something else, that's why I mentioned we built our own version of it internally.

[ChrisMarshallNY]

Nah. I meant that it is a library, that needs to be included as a dependency.

That's a nonstarter, for me. I'm pretty picky about dependencies. If I use external ones, I like them to be something that can be encapsulated/injected, as opposed to being totally interwoven.

But I'm not trying to make money, I just want to write really good apps.

[manmal]

I share your sentiment, and try to avoid dependencies like the pest. But, I wouldn't write my own crypto library, and TCA falls into a similar category - it's a really good solution for a hard problem. Those guys invest a big part of their time into just those few hundred lines of code. There's word that even Apple uses their stuff internally.

[ChrisMarshallNY]

I understand, but I have been burned -badly- by relying on critical path dependencies that failed to stay up to date.

Even if only by a week or two, it can be enough to wreak havoc (but my experience is that they are often many months behind).

Like I said, though. I am not competing with anyone, or trying to make money. I write software for free, for those that can’t afford it. I pursue the vocation as a craft, and deliver Quality that would be totally unacceptable in a commercial endeavor (by which, I mean, high quality. Most commercial efforts go for “good enough” quality).

Most folks here would not find working the way I do, to be very satisfactory.