zopen community: 2024 in Review & 2025 Goals

[IgorTodorovskiIBM]

Zopen Community: 2024 in Review & 2025 Goals

Below are some initial thoughts. But we also want to hear from you - what are some challenges and goals you want us to consider?

2024 Highlights:

• 200+ Projects: Reached over 200 projects, including tmux, emacs, cppcheck, meson, and more.
• Joined Open Mainframe Project: Read about it here. A major milestone for visibility and collaboration.
• New Contributors: Welcomed contributors from Georgia Tech (who helped build zopen audit), several more from IBM, and beyond!
• Adoption Growth: Installs grew from 3k to 5k per month.
• Key zopen package manager Improvements:
  • GPG signing for enhanced security.
  • Pre-requisite support for better dependency management.
  • Improved handling of conflicting tools.
  • zopen audit tool for quality and security checks.
  • Beta dataset I/O support in Vim, Nano and Emacs.

2024 Challenges:

• Package Manager Quality: Inconsistent quality and error handling issues.
• Community Engagement Dip: A dip in activity in the second half of 2024.
• Documentation: Improvement for usage, configuration, and troubleshooting.
• z/OS HW Access: Continues to be a barrier for new contributors.
• Compiler Migration: Need clearer guidance/plan on the xlclang to clang transition.

2025 Goals:

• zopen package manager:
  • Air-Gapped Support: Enable registry mirroring and local pax installs.
  • Enhanced Testing: More unit and end-to-end tests.
  • RPM Support: Add support for RPM package format.
  • Consistent Error Handling: Standardize and improve error messages.
• Community:
  • More Events: Increase frequency and variety of guild meetings + OMP events.
  • Better Onboarding: Improve documentation for new contributors (videos)
  • Incentives: Explore ways to recognize contributions.
  • Public z/OS Access
  • University/Internship Outreach: Expand partnerships with universities / OMP to attract new contributors
• Upstreaming:
  • Track and improve upstreaming success, including CICD upstreaming.
• Continue to simplify porting on z/OS
  • AI:
    • Explore using AI to improve zopen and assist in porting.
  • ZOSLIB
    • Continue to enhance zoslib to bridge the gap between z/OS and Linux
• Documentation:
  • Additional tutorials, how-to guides, reference. See https://diataxis.fr/ (thanks @ijmitch )
• Tools:
  • Focus on High-Priority Tools: Prioritize based on community feedback (Issue #922).
  • Quality over Quantity: Improve existing tools, add missing z/OS extensions.
  • Coreutils: Address outstanding issues (Issue #87).
  • Hardware Optimization: Leverage IBM Z features (e.g., zEDC).
  • Dataset I/O: Expand support in more tools.
• Security:
  • Work towards OpenSSF Best Practices Badge.

Share Your Thoughts!

• What are your thoughts on these reflections and goals?
• What did we miss?
• What are your priorities for zopen community in 2025?
• Should we target specific open-source projects for porting?

Let us know in the comments!

[v1gnesh]

Forgot to bring this up - it'd be useful to add rsync to OEF, as it'll allow tools that work better with rsync (for synchronization), to work better. Ex: Jetbrains IDEs' Deployment feature.

[IgorTodorovskiIBM]

Thanks, the main item missing in rsync is it not preserving file tags (RsyncProject/rsync#447).

[ccw-1]

Also other extended attributes are not transferred.

[v1gnesh]

@ccw-1
Looks like rsync supports xattrs?
https://unix.stackexchange.com/a/707035
https://www.man7.org/linux/man-pages/man1/rsync.1.html
https://access.redhat.com/solutions/3628891

[ccw-1]

@v1gnesh it doesn't I am talking about

https://www.ibm.com/docs/en/zos/3.1.0?topic=descriptions-extattr-set-reset-display-extended-attributes-files

[v1gnesh]

Ahh, okay. Thanks for clarifying.

[ijmitch]

Documentation is third in the list of 2024 challenges but isn't explicitly listed as the topic of a goal for 2025 - let's make it explicit.

I'd like to suggest that there are (at least) 2 aspects to this:

1. Reference and How-Tos - the obvious essentials
2. Tutorials and Explanations - good for easing the more unfamiliar folks in the audience into the fold

If you've come across already Diátaxis (which I did over the holidays) then you'll recognise those 4 terms:

I coincidentally noticed the adoption of the Diátaxis principles in the Ubuntu Server documentation.

[IgorTodorovskiIBM]

That's helpful, thank you!

[ijmitch]

Of course, we need to scope the breadth of the potential need which our audience has for the different types of supporting content - it would be all too easy to take on the responsibility for converting the last TSO/ISPF holdout (I can't help but think cold dead hands at this point!), but there is a set of knowledge we're assuming that NOT everyone is 100% confident/competent/enthusiastic about.

Purely as a example, https://carpentries.org/ has grown out of the need to teach Data Scientists the basics of shell, git and other technologies and working practices to be productive. An immediately relevant topic here is the tutorial for Data Scientists to use shell - https://swcarpentry.github.io/shell-novice/ - adapting that for USS would be interesting.

[v1gnesh]

IMO, drawing parallel to IBM docs (user guide & reference guide), how-to and reference might be enough.
The more thorough, the better, however, tutorials aren't standalone. Instead, in the user guide, we can cover a bunch of common dev workflows.
Also wanted to ask Igor, how important is having man pages, or having command-line args rendered in the docs site reference section?