Readiness of RNTuple

[weak_physicist]

We have an old codebase that was used for our previous project, which makes use of TTree extensively, that we want to use for a new upcoming project planning to production on mid 2025. We are exploring the use of RNTuple instead of TTree in the codebase for the upcoming project.

I looked at some references about the readiness of RNTuple.

• RNTuple: Where are we now and what’s next? - ROOT
• https://www.epj-conferences.org/articles/epjconf/pdf/2020/21/epjconf_chep2020_02030.pdf
• https://arxiv.org/pdf/2204.09043
• ROOT 7 road map? - #5 by Axel

I gather that its planned for production within 2024. Given we are well into 2024, is the stable release of RNTuple still on schedule this year? Is it advisable to use RNTuple for a project that is planning to go production mid 2025?

[Danilo]

Hi,

Thanks for your interest in RNTuple, which is the currently the flagship development project of ROOT. If this can be helpful, I would like to invite you to write to me privately, so that we can arrange a meeting about your project and its needs at your convenience.

The answer to your question depends of course on the nature of the project you are starting.

At the moment, in collaboration with LHC experiments and Dune, as well as several power users, we are working to make RNTuple functional in offline and online data processing environments as well as high performance data analysis setups. During 2024, we will fix the on-disk file format. API changes will have to be expected also after 2024: the precise timeline for those will be communicated in due time.

In terms of releases for 2024 and H1 2025, the timeline is the following:

• 6.32.00: End of May 2024 - 2025 Data taking release for LHC experiments
• 6.33.00: November 2024 - Development Release, not for production.
• 6.34.00: May 2025 (tbc according to the LHC schedule)

Hopefully that can be of help in planning which one to pick for your project.

Best,
Danilo