timokoesters

Hi Neil, fun to see you on HN. I agree with your points and I you summarized it very well as "Ultimately, open source is built on trust".

AI is destroying trust in open source and many other areas and I think this will discourage teams from publishing their source code in the future.

On the other hand, personal connections are becoming even more important, which is unfair to the younger generation and people who don't live near tech hubs.

This is in response to https://news.ycombinator.com/item?id=47789060

Here is an official response from the Bundeswehr (German):

https://www.bundeswehr.de/de/organisation/zahlen-daten-fakte...

This document is an Internet-Draft (I-D). Anyone may submit an I-D to the IETF. This I-D is not endorsed by the IETF and has no formal standing in the IETF standards process.

https://datatracker.ietf.org/doc/draft-thain-ipv8/

I thought this is how it works for most software. What are the exceptions to this rule?

Markdown has a lot of weird choices and works best for longer documents.

Check out my "Advent of Markdown" where I go through surprising markdown behavior: https://mastodon.social/@timokoesters/115643467322561173

Do you have advice for building up this network for graphics development? I'm a Master's student building a custom rendering stack with wgpu and it's difficult to meet people interested in specific skills like rendering.

Markdown has a lot of weird choices and works best for longer documents.

Check out my "Advent of Markdown" where I go through surprising markdown behavior: https://mastodon.social/@timokoesters/115643467322561173

I'm the author of the spec issue this blog post is based on: https://github.com/matrix-org/matrix-spec/issues/852

In my implementation for the Conduit Matrix server, the /sync order is used for everything. The timeline is just one list that grows on one end for incoming events and on the other end for backfilled events.

I think it's important that the message order does not change, because that's very difficult to communicate to the user.

For TLS certificates, the certificate authority has to look up IPs to verify the domain. So the security is still based on DNS, right?

I often hear complaints about DNS. How secure is it in practice and why are there little efforts to fix it?

You can find many professional fonts in https://news.ycombinator.com/item?id=36793029, but almost all of them require expensive licenses or even monthly fees. swisstypefaces.com licenses include all formats with unlimited usage for a one-time fee.

I hope that MIMI (More Instant Messaging Interoperability) will get more attention when it's ready.

https://datatracker.ietf.org/group/mimi/about/

Legally, you are allowed to sell the license (at least in Germany). We need to force distributors to give us tools to do it.

You can find more information here: https://www.dndbeyond.com/posts/1717-2024-core-rulebooks-to-...

Location: Germany

Remote: Preferred

Willing to relocate: Yes

Technologies: Rust

Résumé/CV: https://www.linkedin.com/in/timo-k%C3%B6sters-44329522a/

Email: hn@koesters.xyz

I'm the creator of the Conduit Matrix Server open source project https://conduit.rs . I also like working with computer graphics (e.g. https://plotf.xyz ) and am interested in building optimized software with Rust.

This looks like a great project!

I have a question: Why is it necessary to specify the device for every Tensor? Wouldn't it be possible to set the device once and then all allocations are made to that device?

I made a proof of concept drawing program with Rust and wgpu, compiled to wasm to run in your browser: https://plotf.xyz/drawing

The idea is that you can zoom in infinitely with your mousewheel (or two-finger drag).

Location: Germany

Remote: Preferred

Willing to relocate: Yes

Technologies: Rust

Résumé/CV: https://www.linkedin.com/in/timo-k%C3%B6sters-44329522a/

Email: hn@koesters.xyz

I'm the creator of the Conduit Matrix Server open source project https://conduit.rs . I also like working with computer graphics (e.g. https://plotf.xyz ) and am interested in building optimized software with Rust.

I think the model of having all events form a giant graph is overengineered. They tried to cope with this by introducing things like "fast room joins", but I think they are flawed and treat the symptom rather than the cause of the problem. Servers should only replicate the very core part of a room and fetch remaining information on demand. But I'm working together with the Matrix team to improve this in the future.

I think most other things in the spec are necessary complexity. It's annoying to work on logic for threads, spaces, read receipts, read receipts in threads and so on, but they allow Matrix to have a lot of great features.

What problems did you encounter writing bots for Matrix?

Fluffychat supports it too. Check out https://matrix.org/ecosystem/clients/, you can click the clients for a feature list.

Hi, the goal is to create a fully-featured server that works seamlessly with all Matrix clients. The one notable exception is that we will not implement support for "outdated" room versions, which means that there are some rooms that are unjoinable because they are too old, those should be upgraded to a more recent room version.

I'm also prioritizing features needed for smaller instances, instead of things like user management or horizontal scaling.

Thank you for the kind words!

1: I can't say how much time I will have for Conduit, but I think the project is in a good shape and I think it can reach a stable release without me working full time on it, but of course it will take longer.

2: I think a good way to start is to hang around in the Conduit Matrix room and see if any issues pop up. Often these are relatively simple things like "these logs should have more details" and are a good way to get started.

I will also use this opportunity to link my LinkedIn profile: https://www.linkedin.com/in/timokoesters/

Most bridges should just work, some require messing around a bit, also see https://news.ycombinator.com/item?id=36939480, I haven't debugged this yet.

You can find some bridge documentation here: https://gitlab.com/famedly/conduit/-/blob/next/APPSERVICES.m...

P2P sounds great and I'd love to implement it when it is ready. I think Conduit is pretty well suited for this. But it will still take quite some time until Matrix is ready for p2p clients, it needs some significant protocol changes: https://arewep2pyet.com/

No, currently you have to start a completely new server. At some point after reaching a stable Conduit release we will hopefully be able to create migration paths.

Yes, I use my Conduit instance for all my development work and also host a big Discord bridge on it (for https://veloren.net, which was mentioned in another comment).

Hello, I'm Timo and I started the Conduit project a few years ago. Feel free to ask some questions.

In Conduit (https://conduit.rs) there is no default config path. In order to start Conduit, you need to specify the CONDUIT_CONFIG environment variable, which is the path to your config. This will typically be done in a systemd service.

This has multiple benefits:

- You can't accidentally start Conduit with a wrong config (e.g. the config in the current working directory)

- You can have multiple Conduits running at the same time with different config

- It's easier to understand for the system administrator because there is no hidden information