Anuket Project

2023-03-06 - [Anuket RA2] - Meeting Agenda and Minutes

Attendees:

Agenda and Minutes:

  • Antitrust notices
  • Orinoco
  • AOB & Project review
  • Permanent FYI
  • Actions/Next steps
  • Reference Architecture 2
    • GSMA
      • next Weds' technical meeting - 1st mar - https://zoom.us/j/97220708371?pwd=TnBZMTByTXdPb0NYK3lzRWxidEl3UT09 - confirmed
      • what is clear:
        • rtd with tox pipeline - like ra1
        • GSMA sent a doc template - NG.139 CR1001 (3).docx
        • have issues for:
          • add principles and terminology to ch1
          • add .bib bibliography
          • change references to :cite: the .bib
          • remove links to cntt:common
      • what to clarify:
        • what are the GSMA document guidelines and where are they hosted
        • how to validate document structure - possibly automatically
        • how to validate document content - template mentions a "Document Quality Team" review
        • script for building docx
          • incl. merging rc2 into ra2+rc2 single document
      • notes -
        • http://testresults.opnfv.org/functest/gsma.html#/title-slide
        • http://testresults.opnfv.org/functest/gsma2.0.html

           
        • follow RA1 - output must be docx - copy tox.ini and togsma.py
        • GSMA will have an internal process to apply template and headers
          • they re-generate the docx
        • MD to GSMA - old way
          • used pandoc - issues with links, html code & no backporting fixes
          • switched to RST w/ intersphinx & number
          • proposal to use PDF with GSMA sphinx/latex template
            • GSMA refused due to internal process needing to work docx
          • GSMA rules are not written anywhere - best to reverse-engineer them from RA1
            • references at top/bottom of table, etc
          • final proposal: modified RST to DOCX via Pandoc
            • sphinx features used:
              • intersphinx - can be removed by using plaintext refs
              • autosectionlabel - 
              • numref - can be removed by using plaintext refs 
              • bibliography - 
              • glossary - already removed
            • merge RC2 into RA2
            • switch to list-table everywhere
            • write refs.bib bibliography and use cite - use howpublished field in bib if it's not a website, and just url for websites
            • clean and include terminology from common - must be included in the document itself - copy common, clean up
            • remove all intersphinx entries 
            • remove all :numref: s
          • togsma.py - makes rst usable by docutils/pandoc
            • creates both references and bibliography tables by reading refs.bib
            • dump all rst into a single temp rst file
            • rc2 to be part of ra2, add it as a separate chapter
            • replace all :cite: by numerical index+target
            • replace all :ref: to implicit hyperlink targets by reading build/objects.inv
          • tips
            • no duplicated titles across chapters!
            • no basic tables
            • fix all rst and line wrapping first
            • no multiline refs and URLs! (they break & togsma does not manage them)
          • use default title name in links, not the label (e.g. :ref:'label<link>') - link is ok, label is not.
          • recommend to just build pdf first, even if not used, to see warnings - e.g. unsupported unicode chars
  • new project for RA2 dashboard -https://github.com/orgs/anuket-project/projects/1/views/1


Meeting Recording