Posted by s.aragon on 6 June 2019 - 9:53am

By Laura James, Software Sustainability Institute Fellow

On Tuesday I was up at the University of Manchester for the first meeting of the 2019 Fellowship. The Institute gave us a great introduction and it was good how much they were keen to hear from us what we needed by way of support

Most of the day was taken up with short talks from all the Fellows.  Some particular highlights for me:

Lucy Whalley talked about getting research software published, as a way to get recognition for coding and raise the importance of software outputs. The Journal of Open Source Software is designed to support both good software practice and good research.

Ben Krikler has coined the term "hackshop" for a combined workshop and hackathon, taking place over a few days, with mixed ability attendees. He's been thinking about doing one on FPGA programming, and also about remote hackshops to reduce carbon emissions. I'd love to see more remote events - not just because of carbon-intensive travel, but also because in person events can be difficult for some people to attend and we tend to forget this.  International events may be getting tougher anyway:

Insightful article on Nature, indeed The choice of conference location is getting more tricky with US not even allowing certain nationalities to apply for a visa irrespective of their country of residence, and Canada/UK visas getting Harder.. https://t.co/Muq1uAhoQc

— Hamed Haddadi (@realhamed) May 24, 2019

Mozhgan Kabiri chimeh is facilitating the use of accelerated architectures - making it easier to do complex systems modelling on GPUs by abstracting away the complexity of the GPU architecture.

Daniel Hobley is focussing on code quality and testing practices - especially important for his work at LandLab, where they are trying to bring best practice from open source into the research software world.

Jess Ward wins the prize for effective use of emojis and diagram design in her presentation.

Sorrel Harriet made the excellent point that sustainable does not mean enduring. Software needs to be maintainable, able to respond to change. There's no checklist for this, it's more of a philosophy.

Many Fellows mentioned the Carpentries - I knew of Software Carpentry and Library Carpentry but I hadn't realised there was a whole set connected together. It's a super practical way to improve research practice, not just in universities but in NGOs and other non-coding-centric organisations.

We met in the Atlas room in the Department of Computer Science, surrounded by "relics". This nice processing unit was almost as tall as me.