OSS Watch Advisory Committee Meeting Minutes, 03 December 2009

by Steve Lee on 11 December 2009

Introduction

Present: Lou Burnard (LB), Matthew Dovey (MD) chair, David Flanders (DF), Ross Gardler (RG), Gabriel Hanganu (GH), Steve Loughran (SL), John Norman (JN), Stormy Peters (SP), Paul Walk (PW), Scott Wilson (SW); Steve Lee (SAL) [minutes], Nigel Runnels-Moss (NRM) [guest]

Apologies: James Farnhill, James Governor, Phil Hands, John Mckenzie, Andrew Savory, Joseph Shields

ACTIONS

  • RG: produce draft based on discussion and circulate to AC for comments
  • GH: arrange conference call to discuss feedback from AC
  • RG: circulate licence chooser notes to JN and other attendees

Welcome

MD opened the meeting at 10.00am welcoming those present.

MD described the JISC view of OSS Watch and how its role has changed over time. Starting with the original mandate to explore open source policy in an HE setting, it moved on to encouraging open source to be fairly represented in procurement processes. The role then changed once more to encouraging development projects to embrace open source and open development, with an emphasis on open development from the start of a project. Now after 6 years there is less of a need for OSS Watch to fill the advocacy role as we see the long tail of change in favour of open source. PW expressed reserve about open source adoption, as only 55% of instituitions seem to have a policy in this sense, and most legal departments are not ‘clued in’. RG pointed out that Moodle skews the statistics as it has 87% uptake. MD ascertained that OSS Watch has raised awareness, even if that has not translated into practice. He suggested that the main barrier is closed minds in certain places, which can’t be easily changed. RG felt that OSS Watch could potentially work with suppliers to improve their understanding and help them promote value for money and sustainability to purchasers. MD indicated that effort should be spent building skills and knowledge in the next generation of IT directors, and suggested the OSS Watch legal and licensing materials could be transferred to JISC services (e.g. JISC Advance), leaving OSS Watch concentrate on providing services that support development projects (providing specific advice, encouraging community, meeting JISC open source requirements). Looking ahead, the Government’s review of public sector funding is likely to have an impact, and OSS Watch should be looking at additional funding streams.

Introductions

Everyone introduced themselves and outlined their activities.

OSS Watch Report

GH presentedf the main activities of OSS Watch since August 2008 (pdf). Key areas included:

  • Content creation and management
  • Liaison and collaboration
  • Studies
  • Conferences and events
  • Business and community engagement
  • Project registry
  • Project support

This was followed by a discussion on the likelihood of changes to HE funding and how they would impact OSS Watch. MD said the main concerns are: 1) public spending may be cut to cover deficit, which may result in expanded focus on overseas students 2) there is a debate on tuition fees cap with either no change or cap off - possibly going towards the US private model. He expects middle ground to be taken. SL suggested it would be stable until the middle of next year. RG asked how timing of changes will effect OSS Watch funding proposal. MD replied that JISC are trying to work that out; ideally there will be no changes to funding, however in worst case there will be no money, and OSS Watch materials will be moved to other advisory services. He advised OSS Watch to look at other funding sources and possibly focus on alternative intervention points that JISC could fund.

SW observed that OSS Watch provided a vital service in the past and should be maintained. RG said that funding applications were submitted to European (OpenSE - OSS Education in Computer Science training) and EPSRC calls (though extra staff will be hard to find). NRM is also looking at potential external collaborations. SW pointed out that effects of cutbacks could be expensive to implement. SP observed that in the current economic situation some extra funding is available from companies, as GNOME foundation finds out. MD noted that universities lag behind in recession times. They depend on HEFCE, who wait for the Treasury. Decision on OSS Watch funding will probably be made around May/June based on the best possible indications concerning the priorities of the new Government.

Open Development and Open Innovation

RG gave a presentation OSS Watch “reinvented”? [pdf document], which explored possible key activities for OSS Watch in 2010-2012. He also asked a series of questions about focus, implementation and the opportunity for a software foundation. Substantial discussion occurred during the presentation.

Legal Advice This has now largely become routine, though somewhat limited, as OSS Watch is not staffed by lawyers. It could be handed off to JISC Advance. RG asserted that the university legal departments focus on legal issues and are not interested in the community aspects of software development. The discussion then centred around how projects get legal advice. PW said that lawyers are there to give pessimistic advice on what one can and cannot do, and projects are left to decide if they want to proceed. RG noted that in Oxford if one ignores the provided legal advice no more is available afterwards. LB said this is not always the case, and some universities are more progressive than others. PW felt engineers need to know the right questions to ask of lawyers. It was felt that education in the IP issues is important, as PW observed that we may be seeing the long tail of universities producing FOSS without considering IP issues at all.

JN said that Creative Commons managed to boil complex licensing issues down to simple questions, and perhaps OSS Watch should consider a similar approach for software licences. RG explained the code situation is much more nuanced than content. OSS Watch has created a tool to provide licensing guidance to projects, and while this is not released yet, we can make it available for AC members to try. In summary, MD said that advice on legal matters could move to JISC Advance, with OSS Watch left with specific project consultancy and educating new managers. It was suggested that OSS Watch could potentially provide advice, consultations and formal training. RG pointed out that training is difficult to scale.

Procurement Support According to OSS Watch surveys, less that 15% of institutions consider FOSS fairly, although > 55% have an open source policy. RG thought that policy development and working with suppliers could be a route to a more level playing field, but MD said that could be hard to sell. RG noted that we don’t start from scratch in this respect, and MD thought shared services may provide an opportunity to cater for the common software needs of UK universities.

Materials Production The OSS Watch website has 8,000-10,000 visits, with 12,000-14,000 unique views per month. The blog has 800-1000 visits, with 1,100–1300 unique page views per month. However there is an 80.2% bounce rate, indicating that OSS Watch attracts visitors but doesn’t keep them. MD thought these could be people directly finding what they want and leaving satisfied.

Consultations RG noted that OSS Watch cannot currently say ‘we recommend that you use X’, which leads to frustration, and IP management in projects is often ineffective. MD agreed that the advice can’t be too prescriptive in suggesting which licence projects should use. JN suggested that at least a first level of distinction between licenses can be provided, and this needs to be simple but carefully worded. RG agreed that OSS Watch could explain e.g. how to choose between permissive and copyleft licenses.

On sustainability consultations, RG explained that the ‘strategic projects’ have been very successful. OSS Watch will focus on nurturing community leaders through community development in order to match the limited OSS Watch resources. The discussion then covered how projects often cut corners due to pressure to get on with development. While some felt OSS Watch should focus on developers, others thought that middle managers where a better target. RG stated that OSS Watch workshops are for both developers and project leaders. He also noted that basic process skills are out of scope and should be addressed at Computer Science course level. JN though it was important to look at the gap between ‘real life’ OSS projects and the HE grant-giving exercise, where funding is one of the most substantial distorters. It was felt that there is a need for sustainability use cases that show how to create community and enable contributors to join after creation.

Open innovation RG introduced the concept of ‘open innovation’ and explained how OSS Watch could help make it a focus for the sector. He then explored a number of potential opportunities and barriers, and emphasized the link between open inovation and sustainability. MD and PW thought that if one is hacking a prototype or working on a small project one doesn’t always need to care about sustainability. RG responded that even in such situations one can’t know who may want to use what was built. NRM felt that software should always build on previous work, like academic scholarship does. JN said that real problems occur when people hack a prototype thinking that this is what development work is about. MD noted that sometimes one just hacks to see if an idea is good. PW felt the need for a more detailed decision-making process about when to rewrite rather than reuse code. JN stated that usually funders look for patent-based models, which is a big barrier that goes beyond OSS Watch. RG agreed that most of the issues discussed concern wider issues than just OSS Watch.

Lunch Break

Break for lunch at 12:00. SP left to catch a plane.

Forward planning

The meeting reconvened at 13:30 for continued discussion in a single group around the table after a recap and direction from MD.

The discussion started with what should be the focus of sustainability: software, research or community? MD said that little software is carried forward from projects. JN felt that community is more important that specific code. When RG asked if OSS Watch should focus on something other than what is currently being funded for, SW noted that institutions can find it hard to work on something like FOSS projects, which are not directly funded. NRM wondered about software curatorship: with a project coming to an end and outputs getting shelved, how can one locate something that does precisely what one needs? LB suggested that archiving projects might follow data archiving practices in which institutions themselves are responsible, but JN felt that software aged too quickly for this to be useful. MD stated that curatorship of software is a research domain, which is not an immediate focus for OSS Watch. JN, MD and PW felt that encouraging good open practice is valuable and will spread through the sector. Simal could be used for gathering such statistics, and even if this is not its core function, someone could indeed use it this way.

At this point the discussion refocused on RG’s Open Questions slides. The first topic was ‘future focus’. It was felt that OSS Watch should avoid evangelism, and find instead strategic gaps which JISC should address, such as curriculum design. OSS Watch could also provide analysis, although JN felt that OSS Watch should not become akin to a Consumers Association ‘Which Magazine’. He also saw collaboration at institutional level as a means to provide better software at low cost, and felt that OSS Watch should expand beyond its UK focus. (At this point SL left) A discussion on encouraging reuse and its value for industry followed. RG thought that about 40% of the project bids OSS Watch have seen would benefit of an existing FOSS project, but PW guessed that there were fewer in industry, and wondered if there was enough demand. RG thought that we don’t always know the sector well enough to provide a definite answer. Next LB wondered what other benefits open development brings, appart from saving cash. NRM said that in Germany they focused on innovation and doing it faster. LB felt that in education that meant researching faster, leading the field, and acquiring prestige, and NRM added that this also makes better software.

The discussion then moved to the need to fill the gap in educating on OSS development skills. SW said that his institution (Bolton U) is looking for ways to improve their course provision to address this gap, and felt OSS Watch could provide a service if JISC was interested. MD said JISC were not particularly interested. One view was that OSS Watch could feed into a CS course, but that was not very practical; perhaps BCS would be better placed to do that. Teaching should NOT be provided by OSS Watch, as JISC won’t fund teaching, though materials from OSS Watch could feed into education. DF felt it is important to provide various ways to learn - not just short intensive courses.

Funding models were then looked at in some depth. JN likened OSS Watch to a startup, where initially one chases any available cash sources in order to get established, and later focuses on what reinforces its core objectives. He wondered if OSS Watch has now reached the level of maturity to decide what to focus on. He felt OSS Watch should find ways to become independent of grants. One possibility would be to look at subscription income, where the inertia people have in leaving can be a steadying factor, compared to the uncertainty of academic funding, extremely dependent on economic circumstances. MD agreed that funding required some chasing after the current funding period. LB wondered who would pay a subscription, individuals or institutions. MD suggested 2 possible models: 1) JISC provides core funding and the organisation finds other (e.g CETIS) 2) occasional funding from JISC calls that supplement other streams (e.g. MIMAS). GH asked if these 2 models of funding were exclusive or could be combined, and MD replied that they could be combined.

SW suggested that OSS Watch could charge for strategic project support, but RG pointed out that this would mean we could not be selective about suitable projects. Discussion then moved on to when and how would be best to interact with projects. DF said he has new projects that need mentoring, and that creating groups of mentors would help with scaling issues. RG said that OSS Watch is already attempting this with the Wookie and Shuffl link-up. He also noted that in the current funding round OSS Watch said they would pick a couple of projects and then develop subscription. The goal is that OSS Watch would no longer need to intervene at the beginning of projects, as they do now, however it will perhaps take years before we reach that point. PW said things could change fast - like the recent JISC RI call for which DF said 50% of his RI calls use some community tools

RG said that for open innovation the point with most impact in the life cycle of the project is discovery of potentially useful software, and this discovery should be taken outside the walled garden of HE/FE. SW added that a key performance indicator for JISC Innovation is the number of open source projects one engaged with.

Finally PW observed that JISC has not decided on any particular sustainability model, to which MD stated they supported any model, as they do not have a fixed policy. Rather, they look for sustainability, which may also be achieved through e.g. open standards or commercial models. FOSS is the default route, but most projects just dump code, and OSS Watch should promote open source as part of a complete sustainability plan. RG mentioned that in this funding round OSS Watch changed position to saying ‘if you do FOSS, then you should do open development’, although we have also done some work in looking at other areas, like ‘open innovation’ and ‘community source’. JN noted that, according to the Fluid project, strategy and accessibility are ‘precarious values’ which should be funded. He also mentioned the Islandor case study, which suggests including in the project proposal funds earmarked for contributions from external collaborators.

RG mentioned that he would produce a draft document based on the discussion in this meeting and circulate it for comment before making it available to the JISC

Round Robin

The discussion concluded with asking each participant what would be the most important thing OSS Watch should focus on.

  • JN: emphasize role of open development in helping UK HE R&D build and use better software at lower cost
  • PW: support grassroot developer collaboration, hands-on mentoring on software sustainability
  • SW: support open development as active practice
  • LB: explore how open development can help wider categories of audiences with their strategic goals
  • NRM: make engineering student training include community skills needed for software industry (team work, collective code ownership, version control etc.)
  • DF: advocate sustainability as means of keeping good people on projects
  • MD: build and sustain communities supported by JISC

AOB

No other business was raised

Date of next meeting

It was suggested the next meeting would be towards the end of February 2010, date to be fixed and circulated later.

The meeting was closed at 3.30pm.