In February 2019, a delegation from Red Hat along with Professor Paul Watson of Newcastle University delivered a two-day workshop on Open Source for The Turing Institute. This page summarizes the agenda, provides the slide decks for the presentation portions of the workshop and the next steps gathered from the whiteboard sessions on day 2.

Mark Little’s closing lecture on the first day was a public event. The Turing’s public event page is here: https://www.turing.ac.uk/events/open-source-challenges-enterprise

Day 1

09:30 – 10:00

Registration

10:00 – 11:00

Introduction: Why open source for data science at the Turing? – Mark Little (Red Hat), Paul Watson (University of Newcastle), James Hetherington (The Alan Turing Institute)

11:00 – 12:00

Building a successful open source community – Red Hat (Simon Woodman and Jonathan Dowland) Turing Inst July 2019_ OSS Community.pdf

12:00 – 13:00

Lunch

13:00 – 13:30

Choosing the right open source licence – Red Hat (Mark Little) Turing Institute Open Source Licence.pdf

13:30 – 14:30

Open source: a view from another company – Quantum Black

14:30 – 15:00

Coffee break

15:00 – 16:00

Building a business around open source software – Red Hat (Mark Little) Turing Building a Business Around Open Source.pdf

16:00 – 17:00

Case studies of successful open source communities – Red Hat (Rui Vieira and Rebecca Simmonds) Open source use cases.pdf

17:00 – 18:00

Reception

18:00 – 19:00

Lecture: Open source challenges in the enterprise – Mark Little (Red Hat) Turing Institute 2019.pdf

Day 2

The first part of the day was organised into 4 clinic sessions where Turing researchers presented their projects and discussed their individual challenges with regards open source and community building. Red Hat participants commented on the individual cases. The projects were

The second part was a round-table where we discussed what Red Hat can offer and what might be useful to the Turing.

Summary of actions/suggestions/ideas to consider

  • If the Turing want to build communities around its software they need to increase its visibility in the IT community. Does it need an Evangelist to present its most promising projects at tech industry conferences (Red Hat target these)? It should also encourage project team members to do this so as to increase the visibility of its projects.

  • When Red Hat want to build a community around a product they have a planned campaign including blog posts, tutorials, examples and social media. This is driven by an expert on community building who solicits and publishes content from the engineers. Should the Turing recruit someone into this role?

  • While industry-based business models/roadmaps for open source are well understood, there are different options for research-produced software (for example, a series of research grants can sometimes be used to continue to develop software and build a community to the point where commercial revenue is an option)

  • The Turing is already doing a good job in educating users in the basics of creating open source projects. In particular, there is an existing Software Carpentry effort that the Turing support and we should avoid duplicating their work. Where Red Hat can help is at the level above: e.g. community building, scaling, and project productisation.

  • Red Hat can also provide access to experts in specific aspects of open source: e.g. legal issues and security

  • The overall Turing relationship with Red Hat will be managed by the Turing partnership team (Hushpreet Diwali as the main contact). The main Red Hat contact is Mark Little.

  • Only two academics signed up: it would be good to understand why that was, and to think about how to attract them to future events.

  • The Turing would like a figure for the in-kind contribution of Red Hat to the workshop.

Red Hat offered other ways in which they could help:

  • releasing those presentation slides that can be made public under a reusable license (e.g. creative commons) to see whether it is useful for incorporating into projects such as e.g. the Turing Way

  • repeating the event, or parts of it (one options is to compress it into a day)

  • mentoring staff, students and projects (for example, the Julia project are keen to have a Red Hatter to give advice and guidance)

  • internships for staff and students

  • collaboration on specific projects

  • working with The Turing to productise projects

  • “next stage" tutorials (intermediate, or advanced-level) building on the material/level used at The Turing. Facets mentioned particularly were:

    • issues with scaling up an existing community (growing pains)

    • practical project governance

    • automation (using CIs etc)

    • containers

    • understanding how a Turing project might be productized, including how Red Hat might select a project, and the process of productization