lager VS immer

observer: I changed how observer<> works, this is basically a template-class which observes a value and then calls a registered callback on modification. It can now create a sub-observer<> from a member variable of the value it observers. This makes it possible for widgets to be composable. This is based loosely on lager's lenses and cursors.

By using arximboldi/lager

My preferred approach is to abstract things so the boundary between the objects is not a function call, but a value being passed. This was shown to me the first time via Boundaries (from Gary Bernhadt). In C++ world, the talks from Juan Pedro Bolívar Puente are great, as he also is the author of a framework to structure applications that way, the framework is open source, and it's been used in production applications, probably using Qt as well. I would start by watching The Most Valuable Values and Squaring the Circle. If you prefer text, go read Lager's excellent documentation.

I like working in C++, after a decade of working in Java, Python, Javascript and Clojure, I find working in C++ (which I learned before these other languages) to be quite fun and pleasant, at least with relatively modern C++.

I've been, on and off, working on a little toy game engine, for a few years. Its a mix of keeping up with C++ advancements, learning various concepts like physically based rendering, and just the fun of crafting a big project, with no constraints other than my time and ability, no deadlines, no expectation of releasing anything. Its cathartic and enjoyable. I really do enjoy it.

Last September, I got frustrated with something I was working on in a more serious capacity. It was some server software, it responded to HTTP requests, it accessed third party services over HTTP and Websockets, it talked to a Postgres database. Overall it was an event driven system that transformed data and generated actions that would be applied by talking to third party services. The "real" version was written in Clojure and it worked pretty well. I really like Clojure, so all good.

But because I was frustrated with some things about how it ran and the resources it took up, I wondered what it would be like if I developed a little lean-and-mean version in C++. So I gave it a try as a side project for a few weeks. I used doctest[1] for testing, immer[2] for Clojure-like immutable data structures, [3] lager for Elm-like application state and logic management, Crow[4] for my HTTP server, ASIO[5] and websocketpp[6] for Websockets, cpp-httplib[7] as a HTTP client and PGFE[8] for Postgres, amongst some other little utility libraries. I also wrote it in a Literate Programming style using Entangled[9], which helped me keep everything well documented and explained.

For the most part, it worked pretty well. Using immer and lager helped keep the logic safe and to the point. The application started and ran very quickly and used very little cpu or memory. However, as the complexity grew, especially when using template heavy libraries like lager, or dealing with complex things like ASIO, it became very frustrating to deal with errors. Template errors even on clang became incomprehensible and segmentation faults when something wasn't quite right became pretty hard to diagnose. I had neither of these problems working on my game engine, but both became issues on this experiment. After a few weeks, I gave up on it. I do think I could have made it work and definitely could go back and simplify some of the decisions I made to make it more manageable, but ultimately, it was more work than I had free time to dedicate to it.

So my experience was that, yes, you can write high level application logic for HTTP web backends in C++. You can even use tools like immer or lager to make it feel very functional-programming in style and make the application logic really clean. Its not hard to make it run efficiently both in terms of running time and memory usage, certainly when comparing to Clojure or Python. However, I found that over all, it just wasn't as easy or productive as either of those languages and I spent more time fighting the language deficiencies, even with modern C++, than I do when using Clojure or Python.

I think I would think very long and hard before seriously considering writing a web backend in C++. If I had the time, I'd love to retry the experiment but using Rust, to see how it compares.

[1] https://github.com/doctest/doctest

[2] https://github.com/arximboldi/immer

[3] https://github.com/arximboldi/lager

[4] https://github.com/CrowCpp/crow

[5] https://think-async.com/Asio/

[6] https://www.zaphoyd.com/projects/websocketpp/

[7] https://github.com/yhirose/cpp-httplib

[8] https://github.com/dmitigr/pgfe

[9] https://entangled.github.io/

I've been working on an editor (not text) in C++ and pretty early got into undo/redo. I went down the route of doIt/undoIt for commands but that quickly got old. There was both the extra work needed to implement undo separately for every operation, but also the nagging feeling that the undo operation for some operation wasn't implemented correctly.

In the end, I switched to representing the entire document state using persistent data structures (using the immer library). This vastly simplified things and implementing undo/redo becomes absolutely trivial when using persistent data structures. It's probably not something that is suitable for all domains, but worth checking out.

https://github.com/arximboldi/immer

How does this compare to https://github.com/arximboldi/immer (other than the C/C++ difference)?

Also, it's my understanding that, in practice, persistent data structures require a garbage collector in order to handle deallocation when used in a general-purpose way. How does your implementation handle that?

You might be interested in ewig and immer by Juan Pedro Bolivar Puente:

https://github.com/arximboldi/ewig

https://github.com/arximboldi/immer

See the author instantly opening a ~1GB text file with async loading, paging through, copying/pasting, and undoing/redoing in their prototype “ewig” text editor about 27 minutes into their talk here:

https://m.youtube.com/watch?v=sPhpelUfu8Q

It’s backed by a “vector of vectors” data structure called a relaxed radix balanced tree:

https://infoscience.epfl.ch/record/169879/files/RMTrees.pdf

That original paper has seen lots of attention and attempts at performance improvements, such as:

https://hypirion.com/musings/thesis

https://github.com/hyPiRion/c-rrb

You’re absolutely right, however people have been putting in the “extra efforts” required for efficiency. Check out immer if you’re interested.

The C++ immer library: https://github.com/arximboldi/immer

For C++ check this one out - https://github.com/arximboldi/immer

Yeah I agree that I should link that wikipedia page in the docs, I'll do that as soon as I get a chance. immer (https://github.com/arximboldi/immer) also links that page in its docs, for the exact same reason I'm sure. Interestingly, there is a lot of overlap between persistent data structures in the functional programming sense and persistent data structures in the persisted-to-disk sense because persistent data structures in the FP sense are one of the best ways to guarantee atomic updates and safe failure recovery in a persisted-to-disk system! Btrfs and ZFS, as well as many databases, are at their core basically just copy-on-write B+ trees.

I’ve never tried it, but I love the idea of implementing editor scene state using immutable data structures like https://github.com/arximboldi/immer With that, every edit would append a new node to a list of scene states. Undo/redo becomes iterating your view of the scene up and down through that list. Can’t screw up an undo function if there’s never any work to do :P

I have! I think it's related to the C++ immer library which I used several years ago in Vortex. It's kinda like the previous generation of ValueScript. 🍻