The meeting discussed updates on the graph library, specifically Graph v3,
which uses descriptors to improve abstraction and eliminate memory
references. Key features include vertex and edge descriptors, forward
iterators, and support for non-integral vertex IDs. The team also discussed
the need for a data frame and CSV file reading capabilities to enhance
C++'s AI infrastructure. Michael Wong presented a strategic document on
C++'s role in AI, emphasizing the importance of governance, ecosystem
improvement, and strategic priorities like data frames and MD arrays. The
team agreed to address feedback on special value detection and insufficient
values in future proposals.Action Items

   - MW Write and submit the 'C++ directions for AI' paper for the
   directions group, prepare it for the February mailing (deadline ~Feb 23),
   and share it with the group for feedback before publication
   - PR Implement the remaining graph algorithms (level propagation,
   articulation points, clustering coefficient, and any other outstanding
   algorithms), roughly ~1 hour per algorithm, then submit for group review
   - PR Run benchmarks comparing graph v3 results to graph v2 and Boost
   Graph (performance and correctness) and publish the benchmark results in
   the repository for review

Outline
Introduction to Graph Library v3

   - PR explains the background of the graph library, mentioning its
   evolution from Boost Graph based on C++98 to a modern graph library based
   on C++21.
   - The new version, Graph v3, uses descriptors to improve abstraction and
   eliminate the need for memory references.
   - PR details the changes in Graph v3, including the use of descriptors
   for vertices and edges, and the benefits of this approach.
   - The discussion includes the implementation of forward iterators and
   the impact on the library's design and algorithms.

Detailed Explanation of Graph v3 Features

   - PR provides a detailed explanation of the descriptor record for
   vertices, including storage types and member functions.
   - The discussion covers the edge descriptor, which includes the source
   vertex descriptor and the significance of this feature.
   - PR mentions the use of agile coding to accelerate development and the
   minimal refactoring required for the new version.
   - The improvements in the library, including the reduction of concepts
   in the undirected adjacency list, are highlighted.

Future Plans and Documentation

   - PR outlines the remaining work, including implementing additional
   algorithms and updating the documentation.
   - The importance of having the documentation ready for the next meeting
   is emphasized.
   - Michael Wong discusses the need for a public repository and the
   potential for benchmarking results.
   - The conversation shifts to the use of GitHub Copilot for coding
   assistance.

Discussion on AI and C++ Standards

   - Michael Wong introduces a paper on C++ directions for AI, focusing on
   governance, ecosystem improvement, and strategic priorities.
   - The paper addresses the use of AI in generating normative wordings and
   the importance of human authors taking responsibility for the accuracy and
   logic of their work.
   - The need for a curated, high-quality dataset of modern idiomatic C++
   code for AI tooling is discussed.
   - The conversation includes the potential for a call to action for the
   C++ community to create this dataset.

Strategic Priorities for C++ and AI

   - Michael Wong outlines the strategic priorities for C++, including the
   need for a data frame and the ability to read CSV files.
   - The discussion includes the importance of MD array for homogeneous
   multi-dimensional numeric data and its role in AI workflows.
   - The potential for using compilers and modules to provide modern usage
   information to AI tools is explored.
   - The conversation highlights the need for a comprehensive approach to
   support AI in C++, including data ingestion, generation, and infrastructure.

Feedback and Next Steps

   - Speaker 1 and Michael Wong discuss the feedback received on the paper,
   including suggestions for branchless special value detection and handling
   insufficient values.
   - The importance of consistency in error handling and the potential for
   future enhancements is emphasized.
   - The conversation includes the need for a separate proposal for
   branchless optimizations and the potential for integrating these
   enhancements in future versions.

On Wed, Feb 11, 2026 at 4:13 PM Michael Wong <fraggamuffin_at_[hidden]> wrote:

> Hi all,
> there will be an SG19 meeting 2 PM ET .
>
> Phil Ratzloff is inviting you to a scheduled Zoom meeting.
>
> Topic: SG19 Machine Learning Monthly Meeting
> Time: Feb 12, 2026 02:00 PM Eastern Time (US and Canada)
> Every month on the Second Thu
>
> Please download and import the following iCalendar (.ics) files to
> your calendar system.
> Monthly:
> https://iso.zoom.us/meeting/tJ0kcOGvpzMvGNPqhE6xz-l6prC_nKP0lUcS/ics?icsToken=DDZ5JNBb6PYhUdSZxwAALAAAAIMaV47wjekDB9bQPB0nZe9SOxhl78u3FrFlu0iVNoURVCMFGd8tyN_XHIf77yhRwOWY9n_8HmTldADWlzAwMDAwMQ&meetingMasterEventId=gtPct5vfSpSXMe1GqOkyhQ
>
> Join from PC, Mac, Linux, iOS or Android:
> https://iso.zoom.us/j/99978798527?pwd=nXOqfOzroZgFwPkUprywbXMQrfwycr.1
> Password: 933427
>
> Or iPhone one-tap :
> US: +13017158592,,99978798527# or +13126266799,,99978798527#
> Or Telephone:
> Dial(for higher quality, dial a number based on your current location):
> US: +1 301 715 8592 or +1 312 626 6799 or +1 470 250 9358 or +1
> 470 381 2552 or +1 646 518 9805 or +1 646 876 9923 or +1 651 372 8299
> or +1 786 635 1003 or +1 267 831 0333 or +1 346 248 7799 or +1 408 638
> 0968 or +1 602 753 0140 or +1 669 219 2599 or +1 669 900 6833 or +1 720
> 928 9299 or +1 971 247 1195 or +1 206 337 9723 or +1 213 338 8477 or +1
> 253 215 8782 or 888 788 0099 (Toll Free) or 877 853 5247 (Toll Free)
> Meeting ID: 999 7879 8527
> Password: 933427
> International numbers available: https://iso.zoom.us/u/acqgszVKqM
>