Governmental research funding

[mjtamlyn]

An idea I've had knocking around for a while is governmental funding for OSS. Obviously this would be a hard one to manage, especially given the global nature of the projects, but there's actually a pretty strong political case here - investing money in open source projects helps power business in a profitable and growing sector, keeping the costs down for new (and large!) tech companies. Most companies put billions into funding science for the greater good, why not OSS as well?

Below is the content of an email I sent to my local MP (and minister for tech industry) yesterday.

As I'm sure you are aware, software engineering is a multimillion pound business which is growing rapidly in the UK. Many of the small and medium enterprises are built on open source software - tools such as Linux, Git, PostgreSQL, Python, Django, Ruby on Rails and jQuery. All of these projects are currently almost entirely volunteer run and volunteer driven. They are worldwide projects, but many of them have strong presences within the UK community. Personally, I am strongly involved with Django - a python web framework used to build sites such as Instagram and Pinterest, as well as many sites for NASA, Disney, Google etc. I have also recently seen an uptake in use of the project to build sites in the scientific community. For example, I was working with Oxford University on a STFC funded project giving visibility to atmospheric satellite data. This site was built using Django, and as such saved hundreds of hours of developer and scientist time on the project. Looking more at education, the huge success of the Raspberry Pi and the BBC Micro:bit would not have been possible without open source technology. I trust you can see the value in these tools.

Open source software is in a crisis. It seems every week I am hearing of volunteer developers who are stepping back from their projects, burning themselves out trying to fulfil the needs of the tech community in their spare time, normally around a full time job of their own. In many cases this can leave critical tools used by thousands of businesses completely unmaintained, vulnerable to security attack and eventually dying off as they don't stay up to date with the community. I would like to open a debate around governmental funding of this vital infrastructure. Research councils UK has an annual budget of £3billion - just 1%, just from the UK, would revolutionise the story for these projects and open source software in general. The stronger open source is, the faster tech companies worldwide and in the UK can innovate and develop.

The excellent gov.uk site is built on these open source projects, and the forward thinking which created the team who built it has given the UK government a strong international reputation for understanding how to utilise modern software engineering. This could be a fantastic opportunity to further enhance our international reputation as leaders in the tech community, attracting more businesses and talented people to move here.

This is obviously a bit of a long shot pipe dream, and would require political campaigning and effort to get it to happen. However, it has a lot of strong points. For example:

• There would need to be a governmental body who decides which projects get grants. These bodies would be able to easily set governance requirements for the projects in question - you must have a registered organisation, with an elected board, fair processes etc etc. This leans towards larger projects such as the DSF, PSF etc rather than smaller ones, but funding could be distributed to "community" projects like DRF by the DSF
• The amount of potential cash flow is ENORMOUS. As it says above, the combined body research councils UK has an annual grant budget of £3BILLION, and that's just one relatively small country. Across the EU, Australia and the US there must be an annual publicly funded science budget pushing £100billion. 0.01% of this would make a big difference to OSS.
• Political pressure could be put on companies to match funding from governments
• Can be politically spun as creating jobs, supporting vital infrastructure in the digital world, supporting research, investing in future businesses and probably several other ways I've not thought of

Biggest drawback? Involves dealing with politicians ;)

[freakboy3742]

This came up in a discussion with @jackiekazil at DjangoCon US - and based on that discussion, it might not be as much of a pipe dream as you think.

There are often existing grant programs available with existing funding pools. Yes, you need to get into the competitive bidding process for those funds, but you don't have to convince a politician that funding is needed at all, because the funds have already been allocated.

A key point here - the US Govt is a user of Django; it's easier to sell the "maintaining the infrastructure you're using" argument.

[mjtamlyn]

To be honest I actually would argue that the existing funding pools aren't specialised enough, at least from what I could find the UK. There is no "software engineering research council" - and the bodies which you could apply to are specialised towards scientific research. If there was say, $5m available in a year for just OSS development, there would already be an enormously competitive process for that money just between OSS projects - speculative and pre-existing. And I don't know much about non-web dev projects which would be after that cash.

I want to convince the politicians that this is work prioritising along side things like "materials research" - it should have it's own department/body/etc not be lumped in with general science.

[nanuxbe]

+1 on the main idea, a lot of governments are slowly switching to open-source and also donating money for various reasons to various people.

[tomchristie]

Similar, but more incremental on existing frameworks: Treat R&D tax credit applications in IT sector in more positive light if contributions are clearly open sourced.

[freakboy3742]

@tomchristie Now that is an interesting idea. It's obviously something that is going to need a lot of multinational lobbying to make it happen, but it's also something that should be easy to make the case - it's worth government giving an R&D Tax credit, because society as a whole will benefit from the investment made by a company.

[fijal]

EU is already doing that, e.g. pypy got quite a bit of R&D money from EU few years back. It's also problematic in the same sense as funding academic projects is problematic - a lot of paperwork, bureaucracy and connections to academia necessary

[mgaitan]

My humble contribution will be invite @fperez to share his knowledge on the subject, being a key player in the IPython/jupyter's financial success.

[tomchristie]

It's also problematic in the same sense as funding academic projects is problematic

@fijal - Indeed. For v small companies it may be too much of a burden, but certainly an incentive for mid-larger companies already using or considering R&D applications.

EU is already doing that, e.g. pypy got quite a bit of R&D money from EU few years back.

Even a slight change in the wording of the policy could make an impact. eg broadening the polices to treat open sourced software as a qualifying factor in itself. (Rather than the existing situation where OSS work is judged against the same criteria as closed software)

it's worth government giving an R&D Tax credit, because society as a whole will benefit from the investment made by a company.

Exactly, yes. Broadly the same rationale for the credits in the first place. Assessing qualification could be a bit fuzzy, (not obv that simply making a repo public should be deemed "qualifying" - possible require evidence of engagement) but that's also true if the existing credits system (some co's are good at milking it, others should be applying but don't)

It'd also be a huge incentive for new companies to take an OSS-first approach (Eg sentry) which would be a huge societal benefit.

[mjtamlyn]

Good to know the EU is supporting projects like Pypy - the problem is at present that it's bundled out of academic funding. Open source organisations aren't academic institutions, and they need different processes.

We also need to find out what the processes for applying for this kind of funding are, and make sure they'e well known in the community.

[olasitarska]

Speaking of tax credit: http://www.nysenate.gov/legislation/bills/2015/S161

[freakboy3742]

@olasitarska I'm torn. On the one hand, nice to see legislatures getting behind open source as an idea; but on the other - $200? Don't spend that all at once... :-)

[coderanger]

The Alaska UBI (oil tax credit) is at a similar scale, $1000-2000 per year. A nice starting point, but not actually useful for living.

[jayfk]

@hpk42 gave a talk at CCC this year with the title Hacking EU funding for a decentralizing FOSS project

FOSS and hacker culture meeting the EU buereaucracy. What can possibly come out of that? We'll discuss what is involved for FOSS projects and other interested parties to get $$$ funding by the European Union. Hackers deal with rule systems and their execution. And the European Union issues a lot of rules which are executed by the "commission" and its employees. Within the Horizon2020 framework programme 80.000.000.000 Euros will be distributed towards research projects across Europe between 2014-2020. Shouldn't some of that money go to purposes deemed useful by 32c3 attendants?

https://media.ccc.de/v/32c3-7300-hacking_eu_funding_for_a_decentralizing_foss_project#video