TODO: go through this, make sure everything is in the minutes of the steering committee meeting or in the report, and then delete.

Video-conference https://hubl.in/odk-steering

Discussion of changes to budget

• addition of Ghent as ODK partner
• updates of previously accepted due dates for tasks and deliverables
• changes in budget:
  • minor reshuffling
  • transfers from Paris Sud to UGENt

Postion of Paul-Oliver DEHAYE in UZH/Michael KOHLHASE in Bremen

Things got better thanks to the SC intervention. A contract is on track.

Michael Kohlhase announces that he has received a job offer from FAU Erlangen Nuremberg. There is a high probability that he will accept this offer, effective September 1. 2016. Most of the research group will move with him, Florian Rabe (Co-PI) is not formally decided yet. The steering committee congratulates Michael and will support him in moving his group and the OpenDreamKit project to the new university.

(internal) financal evaluation

percentage of money spent, only 20.69% Pre-financing month 1 = 2 134k EUR (covers 18 months) Funds spent month 9 = 440k EUR (not everyone supposed to be hired was hired)

Brexit

Wait and see. Keep calm and carry on.

Advisory board

News from Steve Linton: - we have a list of potential memebers - nobody has been contacted yet. Maybe one person was. - POD is volounteering t owrite a template email - We want to have 7 people in the AB - attention to gender equality

• 2-3 people should come together and decide to pick the first 7 people asked to be part of the Board

Link in the slides, and below:

https://github.com/OpenDreamKit/Participants/blob/master/AdvisoryBoard.md

Next meetings

January 2017 (small meeting)

• workshop PARI/GP Jan 9th-13th 2016 Lyon
• Bilefeld (group theory)
• USTAN and UJF come together to decide which of their workshops can host ODK meeting
• Review month 18 (March): located in Brussels and/or Gent?

June 2017 (preparation for month 24 deliveries)

• Keep an eye out the other conferences, notably FOCM (Barcelona 10-19 July)

Conferences

• 2017 in ICMS: timeline not so tricky => USTAN in charge

  Aiming fo January?

• Florian Rabe takes in charge application to Dagstuhl (+ add Wolfram Decker in the process)

Workshops:

- Hans: planning workshop in US
- Min: jupyter at SciPy US july / EuroSciPy august
- Min: 3D visualization at Simula in winter/spring
- Vincent Delecroix: planning a Julia/flint/Sage meeting about algebra (from polynomials to number fields), 1st semester 2016-2017 Bordeaux (?)
- Michael Kohlhase: Workshop in Edinburgh as CICM satellite, july 2017
- Olexandr: joint event with Codima's training school (Codima = extra booster for dissemination activities)
- Loic Gouarin: training school
- Software Carpentry instructor training around Paris

Communication problem

The website does not advertise enough what we are doing despite the fact that the procedure is described on the readme

https://github.com/OpenDreamKit/OpenDreamKit/

Progress reports, 2016-07-01 10:21 +0200

Michael: - running around starting new discussion group - only done one single line of code, and that was a deletion - lots of fruitful discussions that hopefully pushed things in the right direction - Finalized the job offer negotiations to move to …

Luca:

• lots of stuff added at the bottom, per project
• in parallel, I was working on the incomplete CyPari deliverable

Florian

• talked to Nicolas? about Sage, my impression: there’s a huge bottleneck
• protocol is negligible because it’s too hard to export anything
• there are many operators in Sage which will be very hard to export because they are Python objects and it will be very hard to get a handle on them
• constructors tend to be concrete things and are harder to export
• Nicolas’s categories …
• we might know that there is xxx of Groups, but presently we don’t know what the constructors for groups are
• Nicolas: or more precisely, generic constructor such as cartesian product…
• Dima: permutation groups in Sage have a GAP implementation
• it seems that there are several ways to get an object, none of which has a complete coverage

Olivier

• starting a page on “who is working on what”

OlexandrK:

• in GAP, it took a while to get to this idea;
• the initial implementation that was not updated for years already gives you a working solution
• in the next step you can xxx
• in Sage you can extract generators of a permutation group, you can encode this in OpenMath

• Markus: GAP-system should find a way to tell package authors to do things in a way that they can be automatically xxx, without requiring big effort from them

• Michael: Nicolas and Sam should learn from Olexandr and Markus

DONT TAKE NOTES PER PERSON

SCSCP GAP/OpenMath

• Olexandr: setup for a generic GAP SCSCP server to accept OpenMath object or GAP input as a string, evaluate it and return resulting OpenMath object or GAP output as a string: see https://github.com/olexandr-konovalov/scscp-evaluator

OpenMath in python

• participants: Luca, Tom
• the code is on github in a repo that belongs to the OpenMath project https://github.com/OpenMath/py-openmath
• once this is done, we can use it to get pretty quickly and fast scscp support for PARI (should be doable through CyPari)

SCSCP Sage/OpenMath

• Luca: working on SCSCP in Sage (demonstrator soon)
• Michael: it is very good to have scscp protocol, because as soon as you are able to export something, you are also able to transport it
• Michael: another potentially interesting thing:
  • as soon as something is OpenMath or MathML capable, which for now is the same, we can also harvest it into MathWebSearch;
  • you just need some url that you can point us to and the rest is automatic
  • anything you can search with a pair of eyes we can search with mathwebsearch
• Michael: Olexandr, is there a list of projects that implement SCSCP
• yes: www.symbolic-computing.org/

Sage -> MMT export (Nicolas, Sam, Dennis)

• https://github.com/nthiery/sage-gap-semantic-interface
• git@gl.mathhub.info:ODK/Sage.git

• browsable at https://odk.mathhub.info/mh/mmt/?http://www.sagemath.org

• Explained the inner working of categories, axioms, …
• Expanded the Sage->MMT export with:
  • semantic annotations
  • all methods in categories
  • stuff toward exporting all methods of all classes
• Further annotation work: lie algebras, …

GAP -> MMT export (Markus, Olexandr, Dennis,…)

• git@gl.mathhub.info:ODK/GAP.git
• browsable at https://odk.mathhub.info/mh/mmt/?http://www.gap-system.org

Preparing for curating effort (POD, Michael )

• discussions with Sage -> MMT group
• https://github.com/OpenDreamKit/OpenDreamKit/wiki/Curation-Effort

## Towards a web UI for MMT

• Markus has a student working on a WebIDE for GAP, based on CodeMirror. Might be used as starting point for MMT.
• Michael: it would be much easier if there was a standard browser-based editor. We also lack manpower.
• Tom has looked into JupyterJS, it might be used for this

Dynamic documentation (Florent, Vincent, Nicolas, …)

(toward D4.5)

Finalization and review of http://trac.sagemath.org/ticket/20690

Legacy Sage notebook -> Jupyter conversion

(toward D4.5)

Playing around with Volker’s implementation, fixing some trivial bugs:

https://trac.sagemath.org/ticket/19877

Explore how one could reimplement Sage-Explorer using Jupyter (Min, Nicolas)

(toward …)

Redesign of the website (Luca, Nicolas, …)

• Luca and Nicolas had a discussion yesterday on redesigning the website
  • idea: do the design on a temporary version

Future possibilities

• Olexandr: contacted ICMS (Edinburgh) regarding the possibility of running dissemination workshop in January 2017

Upcoming workshops

• Dissemination workshop in ICMS, jointly with CoDiMa
  • Olexandr wrote to ICMS yesterday, they confirmed that the week of ??16-23?? Jan 2016 is available
  • theme: xxx
  • hope that Peter Cameron could come
• Olexandr and Dima will talk about the upcoming WP7 workshop to set dates and some concept of the workshops
  • tentative list of speakers
  • tentative list of participants to invite

WP7 (Dima)

• Dima has been working on understanding a bit more about WP7
• I hope to work more on the Sage and GAP interface

Future funding discussion: Cost network, … (Markus, Nicolas, Michael)

Rambling about what the consortium could be, who to encourage to lead, …

Docker

• some work by Erik

• we have a wiki page listing docker containers for the various projects in OpenDreamKit https://github.com/OpenDreamKit/OpenDreamKit/wiki/D3.1-Virtual-images-and-containers

• Dima: floating-point libraries, polymake, xxx

• Nicolas and Erik had a discussion

• OpenDreamKit container (for branding)
  • will be huge
  • reviewers seem to want a product
  • then they have something tangible
  • it will also be useful in some contexts

Windows

• some work by Erik

WP overviews

• Florian: would be nice to have WP overviews
• 2017: WP5 high performance computing
• 2016-04 Cernay WP3 mostly
• WP7 workshop next fall
• it’s kind of natural that WP6 which is more overarching needs more face-to-face time

Sage infrastructure

• Erik been working past few weeks on Sage infrastructure: migrating trac.sagemath.org and git.sagemath.org
• Erik and Dima discussed this a bit this week, and will discuss more today

Organisation of curation effort

• Paul + Michael, everyone else
• figuring out what needs to be done
• assessment of where data is stored, which hooks are ran

• https://github.com/OpenDreamKit/OpenDreamKit/wiki/Curation-Effort

• Olexandr: read the new readme, found it illuminating
• not clear when deliverable is ready
• I would in principle be interested in looking at deliverables when they are ready
• the readme says: when deliverable is ready, send email to participants mailing list
• Min did that once, Serge does it and writes a blog post
• maybe send a pre-email shortly written the submission, to ask for feedback

Yesterday Florian suggested that we need to have xxx for GAP scscp server, (Markus opened a ticket) which accepts anything and returns anything, where anything can be of two types

• I set it up, it’s available on the repository
• if you wanted to run it locally, you can start xxx
• nobody from outside will be able to run it
• Markus: would be better to have it in the Docker container
• don’t think security is so much of a concern

• if somebody starts computing bitcoins, you can just kill the machine

• Olexandr: in our GAP docker container pipeline, we have
  • on top of that, third container which contains all packages configured
• but there is no gap development version there
• i need gap development container
• also want to have the master version of the branch for another purpose: Jupyter
• my next plan is container with gap-master, automate its generation on the grounds of using one of the base containers
• and then offer a public service
• then I will not be so concerned about security