2025-07-24 – 15:00-16:40 (Africa/Abidjan), Main Room 5
The Julia I/O minisymposium will discuss the low-level details of Julia's input and output ecosystem and its interactions with the Julia Data ecosystem. The ability for Julia to read in and then write out data efficiently is critical to doing data science and scientific computing. Julia is unique in terms of input and output in that the language combines both dynamism and a parametric type system. In order for high-level data storage to succeed, a strong I/O foundation is required.
The Julia I/O minisymposium will gather experts on Julia's input and output capabilities to discuss how Julia's standard library can imporove support for both low-level packages such as TranscodingStreams.jl, HDF5.jl, Zarr.jl, Arrow.jl, and HTTP.jl and high level packages such as a DataFrames.jl and DataFramesMeta.jl.
Proposed speakers include
1) Stefan Karpinski to discuss the design and improvements for the IO interface in base Julia.
2) Nathan Zimmerberg to discuss TranscodingStreams.jl, compression codecs, and plans for version 2.
3) Jonas Insensee to discuss JLD2.jl and its interactions with transcoding streams codecs
4) Jakob Nybo Nissen to discuss improvements to base Julia IO for bioinformatics and its comparison to the Rust standard library.
5) Jacob Quinn to discuss HTTP.jl and OpenSSL.jl and the transition of Julia from mbedTLS to OpenSSL as well as data packages for which is the primary author.
6) Bogumił Kamiński to discuss DataFrames.jl and the related tables ecosystem.
An important topic for discussion will be how to improve data schemas to take advantage of Julia's types to grow beyond Python-like dynamic type interfaces.
The minisymposium will also accept outside proposals. The minisymposium call for proposals will be mainly advertised to the Julia community via Discourse, Zulip, Discord, and Slack.
Julia co-creator and JuliaHub co-founder.
I am a Software Engineer at the Howard Hughes Medical Institute Janelia Research Campus.