gdcbe

Who had known that their Mac Minis would be the servers they never wanted to sell again :')

Seems not to be only an issue in the development world though.

At least in Belgium there’s also a lot of unregistered solar panels to the point that the energy regulator is using satellite imagery to catch and find them…

Starting from the upcoming 0.5 release of Rama (~ end of September '26), Rama's CLI `serve proxy` command will have a Man-In-The-Middle (MITM) mode `--mitm` that will serve a (protected) web GUI to monitor and engage with the network flowing through the proxy. Happy to also anwer any questions here regarding rama the network framework, or our CLI tool built with it.

You might want to read the “prior art” section of their article…

That is a question I can answer in many ways. But the prime answer is that it's a weird question to ask.

When you mention tower you might mean two things:

1. the `Service` and `Layer` concept. That is in the end just a "tool" to provide composition. Just like you also use records (`structs`), procedures (functions) etc... We also lean on that concept. And also note that Tower did not invent that concept either. So in that sense Rama is just another child down the lane of this lineage.

2. The tower ecosystem. Rama can bridge into it, but it's true that we are not itself part of it. Six years ago Rama was still built directly with Tower. Later we created `tower-async`, and eventually we went our own road. This because it was clear at that point (and still is) that the tower ecosystem is not really gonna change any time soon despite rust having evolved a lot since its initial conception.

So if you like the concept of Tower, with its services and how it allows you to compose than you will also like Rama. That's however kinda where the comparison ends. Rama is a protocol agnostic network framework to build clients, servers and proxies. That is not what Tower is or its goal. Tower its only goal is provide an impl of the concept of "services". Rama uses that same concept, but its just one of many ways how we allow composition and is in a way just one of the concepts that make it so nice to build network services / network stacks.

Now different than what you might be used to in tower, is that in rama the service concept is everywhere (https://ramaproxy.org/book/intro/services_all_the_way_down.h...). You might also like to read https://plabayo.tech/blog/rama-101-2-core-concepts. With rama your services go through the different OSI layers, and easier to use and implement.

Rama is there as the foundation to build network services. Tower is there to provide you with a way to impl the concept of Services. Tower is also a ecosystem and we keep a very close eye on it. I suggest you explore Rama a bit for yourself. A lot more will be clear to you than whatever I might say here.

That’s an unrelated project indeed, both in terms of ownership as well as what it is about.

You can learn more about our project name origin at https://ramaproxy.org/book/faq.html#why-the-name-rama

I hope as you learn more about our project that you get excited once again! :)

Info for those who do not know Rama yet: it is a modular rust framework to build network services of any kind (client, server or proxies). For projects where you need to combine multiple protocols, be in control of the stack, build advanced stacks or any kind of proxy, rama is probably for you. And if not, do reach out to us.

Earlier this week we released Rama 0.4, and with it we are bringing system configured proxy support to clients (rama-net), this includes PAC support (rama-pac) driven by Javascript within a WASM runtime (rama-js). It also expands our gRPC support and adds support for ttRPC, a lightweight alternative to gRPC running directly on top of TCP.

Read more in the blog post and discover also the other new features and improvements that Rama brings in this new release, that is 0.4.

As always we are very much open to feedback on this Release, rama itself or how it is presented. It's a project 6 in the making, with everything written for production purposes since day 0. Feedback is always greatly appreciated especially if it allows us to improve Rama. Contributions welcome of course as well.

I have yet to see an actual part of that policy which requires a cookie banner though. Seems more to me that it's a combination of (a) websites allowing all kind of fcked up use cases of cookies on their site (most sites do not even need cookies for real) and (b) not respecting http headers that ask to not be tracked... They much rather have a very confusing popup that kinda forces you to accept all :) How convenient.

Just like websites also give zero fcks about accept-language header... sure do geoip lookup, so much easier... not

Yes we use it for rama [1]. You can check its justfile and CI workflow file how we use it. Those run thousands and thousands of tests thx to nextest and what feels like instantly (once compiled).

Large projects build with rama use it as well. But those are proprietary from partners so sadly cannot share those.

[1]: https://github.com/plabayo/rama

Thank you very much for developing nextest. It is what allows our projects like rama [1] to run thousands and thousands of tests in a blink of an eye! Keep it up!

[1] https://ramaproxy.org

Thank you very much. We will see, I have been on this for 5 years, if it takes another 10 years for the wider global community to discover rama, so be it. We have patience.

No clue how an AI would write it, but that is sadly how I write, sentences that are perhaps ending a bit too... late. I like my comma's and my em dashes... always have... Used to blog a lot more, trying to pick it up again, with this blog article being hopefully the first of many more to come.

And yeah mostly want to highlight the fact that is ready for production because sometimes people are worried for < 1.0 projects, but want to make sure that people know that we and plenty of other companies are already using it in production for serious projects, not just a hobby or R&D thing.

Our aim is to solve a real world problem with this, so that we can all stop reinventing the same duct-tape mess and instead have a shared foundation that we can all build upon. Such that if there is something to improve or a mistake it has to be fixed once and all rama users benefit from it. While still ahving the flexibility to use and stack your different services however you wish.

OP Here, sorry for not being clear. And happy for feedback on our presentation, docs, my article and ofc the code itself. It has been under developed since over 5 years, and was used in production right from the start. Now we are at a point that it is used by several different companies and even more individual users.

Here is some context and info about the framework that I am quoting from our README to provide you with context:

Rama is a modular service framework for the Rust (programming) language.

The framework is intentionally explicit. Your network stack is built from services, layers, transports, protocols, and state that you compose yourself. That makes the shape of the system visible in the code, instead of hidden behind framework magic or configuration.

This makes Rama a good fit not only for proxies, but for network services where the stack itself matters: how traffic enters, how it is decoded, where state lives, what gets inspected, what gets transformed, and where it goes next.

Whether you're inspecting traffic for security analysis, writing a web service, emulating clients with custom user agents, controlling connection behavior for advanced testing, or building high-performance proxies, Rama provides a clean and composable Tokio-native foundation for network services in Rust.

Rama is used in production for network security, data extraction, API gateways, routing, and other networked systems.

Yes the blog post was written completely by me. Not entirely certain what triplet you are referring to, or what that might be, but your suggestion is spot on. Going to add a comment with more context of what rama is as a comment myself. Appologies. Not yet used to promoting projects myself...

The nicest thing of Elm is how much it feels like Haskell. Have built some fun things with Elm years ago.

The second nicest thing with Elm is the philosophy of if it compiles it works. And to be honest you can get that same feeling with most of Rust as well. Sadly not as much of a haskell feeling but at least it has a warm shadow of some of its functional ancestors.

Seems to flag any project related to networking — regardless if it is a network framework or a podcast website — as unsafe... oh well... let's see how it is once they losen up...

Maybe in USA in big tech where companies give absurd wages to engineers anyway in some states, that might be acceptable. But to make their ROI they need that (and more) to be spend world wide... no way that is gonna be a budget that is gonna fly in the long term...

Companies love to cut costs, and just like they axe employee numbers at will, they will just as well make that kind of budget quickly dissapear the moment they realize they can go a different path for same or better value... Or simply because share holder short-term value demands it...

What if you phrase the question from "will AI ever be useful" (a term as utterly vague as "IT") to "will it ever be able to promise the financial gains these companies are hoping? Especially with local models eating their lunch :shrug:

I do not disagree with what you are saying, but I honestly still believe that most of the utility we experience are honestly gonna become very boring very soon that we can just run local... Even if it's a bit more slow who cares, can just run in background while you work on other stuff yourself, read up on things, review other work...

It's not that the utility of it put in question. What is however a giant question mark is how the heck any of the big AI companies are ever gonna get that ROI? Given how many of us are becoming more and more fine with local models that run just fine especially on a good enough computer which most developers have anyway...

Congratulations terts and team. In November’25 we had pleasure to interview you about roto and Nlnetlabs [1]. Happy how far it has come already. Before the summer ends We’ll implement support for it in rama [2] to support scripted services and anything else you might want! Looking forward to that day. Until then, keep it up!

[1]: https://netstack.fm/#episode-14

[2]: https://github.com/plabayo/rama

… is that even legal to do for microsoft? Are there no requirements to adhere to certain standards? Would have thought that is part of it.

Father of three here as well, amen to that. Soo hard, but oh so amazing.

Do you or someone reading this know who would be the best person that would be willing to come on a guest on a podcast and has the correct knowledge (ideally the person who implemented in WC3, or something similar enough).

Asking as I'm the host of netstack.fm, a podcast about networking and rust, but some episodes are just about networking alone.

Would love to devote an episode to the Kali TCP/IP IPX bridge as there's a lot to unpack there and that can be learned from. Any tips for a guest for such an episode are more than welcome!

Htmx got me into hypermedia heaven, but it lead me to datastar for sure. Recently we also had an interview with the creator of datastar, where he also talked a bit about darkstar (something he wants to built on top of webtransport for the few things where datastar is no well suited for now)

https://netstack.fm/#episode-4

Maybe they should focus less on "agentic" and more on just keeping their core product solid... I suppose that doesn't rhyme with growth at all cost... zzz sad... it is

Yeah such clear cut cases I've done in the past. E.g. there was a period some months ago when random AI people (so not humans "operating AI's) did such PRs. Those were easy to filter out.

But sometimes you have people that do seem to really do put in the time and do their best, and I suspect that without LLM they couldn't do what they can now do.

It's those cases that I find less clear cut. Like how to politely ask them to have certain criteria before they submit.

That’s vaguely what https://blog.cloudflare.com/ai-labyrinth/ is about

For me it only worked after I added their form text to my hn account description on my actual hn account page…

Invalid pointers and related bugs are older then 20 years…

Languages like Rust don’t solve all problems and nobody is hoping that the research and progress towards better tooling stops there. Even many people working on Rust give plenty of ideas on what future languages can do better or different.

You don’t need to wait 20 years, those discussions are already happening…

So uh… I’m confused on what exactly you are trying to communicate here?

Why do people talk about hallucinations? Pretty deceptive word of you ask me.

Not an expert though, but isn’t that behaviour inherent to how it works? Bit of a misnomer and giving people the wrong idea of what is going on here.