This is the annual curl meetup. Developers, users, fans, enthusiasts. People who want to talk about curl and listened to presentations about curl over a whole weekend.

Insights from the project. Stories from users. Planning for the future. Learning from the past. Related technologies. Coming protocol developments - and everything that can be considered related or adjacent to this. There is no better place to learn about curl and curl internals than curl up.

We are an open and inclusive set of friendly people. Past curl ups have typically consisted of 20-30 persons.

May 3-4, 2025

Prague, Czechia
Vlkova 36
Praha 3 – Žižkov, 130 00

The venue is called Pracovna. Find it on Google maps / Mapy.cz

Go here. Registration is mandatory to attend.

The event is entirely free of charge, thanks to our generous project sponsors.

Dinner event for those who said yes before April 27.

Social event: Dinner at a restaurant. TBD

Tell us what you want to speak up about.

Tell us what you want to listen to.

60 - 90 minutes walk-through of the curl project right now, and what's happening 2025.

A short presentation about the current setup for contingency and an open discussion about the future, recruiting new maintainers, getting younger people involved in the project, financing and how to make curl survive another hundred years.

Past year, where we are, what to do, how to do better. Security is important.

With participants adopting the roles of 'The Curl team', 'Others' and 'Note takers' - I will facilitate running a series of plausible security scenarios against curl with the goal of identifying better approaches to secure development, identify how we might respond to emerging security issues as well as have a bit of fun. The session would be between 1-2 hours, interactive and establish ground rules of engagement as well as being mindful that running in a public forum naturally constrains.

A look at some things we worked on during last year

• Abstractions help: blocking connect code squeeze
• Controlled connection shutdowns (and why they are good)
• TLSv1.3 Early Data, SSL session caching and Im-/Export

I have an idea. It's not really new. But maybe its time has come?

A curl-agent and a curl-add, very similar to what OpenSSH has, that acts as a key=value store. Starting a curl command line would contact the agent over a unix domain socket and lookup variable values there when needed.

A session for some light Sunday morning brainstorming...

See also https://github.com/curl/curl/wiki/curl-tool-master-client

It might be nice (maybe at the beginning of first or 2nd day) to have a general 'meetup with curl' open for general public ... its something we can try to attract interested people ... enticements such as stickers and tshirts might go down well, we might even consider this for a general event at the pub. The idea being that we would have small workshop for people to ask questions, give feedback and engage.

A short presentation on how curl interoperates with common security data formats and tooling - does it matter for an open source project like Curl ?

The curlx_ "system" could be turned into a proper internal (static) library that can be used by libcurl, curl and test servers. Could reduce build times and simplify things?

• cookie update
• QUERY method
• resumable uploads
• ...

We can spend some time going through Issue / PR.

Which formatting options, if we should do this at all ?

Lets have a discussion about the many kinds of performance testsuites that we might consider for curl/libcurl

If live, has to be on the 3rd between and 12:00 and 15:00 (CET) - Best between 13:30 and 14:30. This session doesn't need the full hour, 30m would be fine.

What can we glean from the curl regression test data analyzed by Test Clutch?

We intend to record all presentations on video and make them available after the event.

If the infrastructure permits, we will try to live-stream them as well.