As I said, my app is write heavy. So there are several separate processes that constantly write to the database, but of course, often, before writing, they need to read in order to decide what/where to write. Currently they need to have their own read cache in order to not clog the database.
The "web service" is only the user facing part which bears the least load. Read caching is useful there too as users look at statistics, so calculating them once every 5-10 minutes and caching them is needed, as that requires scanning the whole database.
A CDN is something I don't even have. It's not needed for the amount of users I have.
If I was using Postgres, these writer processes + the web service would share the same read cache for free (coming from Posgres itself). The difference wouldn't be huge if I would migrate right now, but now I already have the custom caching.
After 2 years in production with a small (but write heavy) web service... it's a mixed bag. It definitely does the job, but not having a DB server does have not only benefits, but also drawbacks. The biggest being (lack of) caching the file/DB in RAM. As a result I have to do my own read caching, which is fine in Rust using the mokka caching library, but it's still something you have to do yourself, which would otherwise come for free with Postgres.
This of course also makes it impossible to share the cache between instances, doing so would require employing redis/memcached at which point it would be better to use Postgres.
It has been OK so far, but definitely I will have to migrate to Postgres at one point, rather sooner than later.
It depends very much on the language/server you are using. In Rust, IMO GraphQL is still the best, easiest and fastest way to have my Rust types propagated to the frontend(s) and making sure that I will have strict and maitainable contracts throughout the whole system. This is achieved via the "async_graphql" crate which allows you to define/generate the GraphQL schema in code, by implementing the field handlers.
If you are using something which requires you to write the GraphQL schema manually and then adapt both the server and the client... it's a completely different experience and not that pleasant at all.
The answer to the final question in the article is: "Mostly, because the government forbids us from solving this problem". Start a company? That will be 2k for incorporation and 3k per year for auditing/bookkeeping. You want to build something physical in Europe? What a bummer, come here and pay the exorbitant carbon tax. You still think you are very smart? Your will pay 70% effective tax rate if you want to hire someone else to help you.
I have a Rust web project written ~5 years ago. Zero updates until a month ago, also zero problems. A month ago I upgrade the docker file to latest rust, zero problems compiling. Could have left it to run like this for 5 more years probably but I decided to experiment...
I issue a `cargo update` to upgrade the leaf dependencies and do automatic minor version updates. Issue cargo check. Some new warnings from `clippy`, but it still compiles and runs without problems. Could have deployed for 5 more years, but I decided to experiment more...
I upgrade some of the libraries to major new versions - I am experienced and I know which ones will upgrade without problem. They do upgrade without problem. Could have deployed for 5 more years but decided to walk the extra mile...
I upgrade the more problematic ones, especially actix_web, the web framework, which had a massive rewrite and a huge new release with almost completely different API surface... It's a bit difficult to understand the changes, especially some parts of the old code written for the old version (which I no longer remember), but in an hour I'm done. Afterwards `cargo outdated` does not report any outdated libraries. I deploy for the next 5 years. Zero problems since then.
Well, it's not decades yet, but I can imagine similar effort to maintain it over the next decade.
Having written a moderately big project in OCaml which I tried porting to F# and later did port to Rust, to me Rust feels much faster to develop than either OCaml or F#, especially once you figure out the "core", adding more features is a breeze. Refactoring is also easier. Not to mention that reading Rust is much easier than reading OCaml and coming back to the project after a year feels very easy. I think that I have less bugs with Rust than with OCaml. And the end product's core ended up being ~3-4 times faster to execute in Rust.
I tried F# some years back when I was searching for a language to port my OCaml project in... It felt too much .NET-y and too much MicroSoft-y. And back then .net for linux had just been released and was somewhat unpolished.
It seemed that I had to learn C# in order to use F# properly and it seemed that porting it to C# was the saner option. I went with Rust after all and it seems to have been the right choice.
I don't think Zoom is the problem. It's just as bad in the classroom. The only difference is that in the classroom there is a century old system in place, designed and tested to crush children's emotions and frustrations and also designed to transform them into obedient tax cows. And in the classroom you are not witness to this process.
That's exactly what they did to us in communist Bulgaria. Forced education in the politically correct curriculum so that we can be "free" to follow the communist guidelines.
And then it collapsed. Currently the western "liberalism" is going in the exactly same direction
The "web service" is only the user facing part which bears the least load. Read caching is useful there too as users look at statistics, so calculating them once every 5-10 minutes and caching them is needed, as that requires scanning the whole database.
A CDN is something I don't even have. It's not needed for the amount of users I have.
If I was using Postgres, these writer processes + the web service would share the same read cache for free (coming from Posgres itself). The difference wouldn't be huge if I would migrate right now, but now I already have the custom caching.