Ancestors

Toot

Written by David Lawrence Miller on 2025-02-05 at 10:06

science software funding chat

how often do folks who review (or otherwise look at) grants (or projects) see "we will contribute to X package/library" where X is not written by the grant authors?

I see (and have been involved in) a lot of projects who say "and we'll write a package for this" but very few (almost none) that talk about external contributions to existing software.

The latter seems far more useful, but obviously less flashy. Are the incentives ("originality") in the wrong place for this to happen or am I suffering from sample bias?

=> More informations about this toot | More toots from millerdl@mathstodon.xyz

Descendants

Written by David Lawrence Miller on 2025-02-05 at 10:08

...

it seems to me that your actual impact would be much higher by contributing to an existing piece of software, with an existing user base.

Yes, adding to existing software is harder: you have to understand someone else's code before you start, but the users don't have to learn something new. So, overall, we probably save time?

=> More informations about this toot | More toots from millerdl@mathstodon.xyz

Written by Chip Butty on 2025-02-05 at 10:09

@millerdl funding for FLOSS in academia is so very, very broken

=> More informations about this toot | More toots from otfrom@functional.cafe

Written by David Lawrence Miller on 2025-02-05 at 10:12

@otfrom YES! I worked on a suite of R packages for 15 years in a previous job and tried numerous times to get money. I had zero success getting money from UKRI or other traditional grant agencies.

Improvement funding came from: Canadian government (env&climate change dept), the International Whaling Commission (bad name, actually now a conservation org) and the US Navy.

=> More informations about this toot | More toots from millerdl@mathstodon.xyz

Written by Chip Butty on 2025-02-05 at 10:24

@millerdl when the digital catapult was still called that, I told them they should be funding the fundamental floss underpinnings of the economy. I couldn't make them understand it though

=> More informations about this toot | More toots from otfrom@functional.cafe

Written by David Lawrence Miller on 2025-02-05 at 10:39

@otfrom yelling "infrastructure" into the void? I know that feeling! :(

=> More informations about this toot | More toots from millerdl@mathstodon.xyz

Written by Dan Stowell on 2025-02-05 at 19:20

@millerdl I can't remember ever seeing either of those in a proposal, TBH. But they do say more generically they'll do "open source" etc. It'd be good to see "we will contribute to X package/library" - as a reviewer, I'd call that "a clear route to impact".

=> More informations about this toot | More toots from danstowell@mastodon.social

Written by Phil Browne on 2025-02-05 at 19:57

@millerdl I can't help add to your actual question but came across a deliverable in a project I've been co-opted into. "We will release X as open source software". X is a single line computation. I have no idea what they actually expect from it! Maybe they want the LaTeX code for the equation

=> More informations about this toot | More toots from phil_browne@fediscience.org

Written by David Lawrence Miller on 2025-02-06 at 08:07

@phil_browne !!! well I guess that's an easy objective at least ;)

=> More informations about this toot | More toots from millerdl@mathstodon.xyz

Written by Martin Modrák on 2025-02-06 at 08:04

@millerdl Under my country's (Czechia) research assessment system, contributing to software maintained by others is basically impossible to claim as a "research output", whereas writing your own package (however small) is a recognized output.

=> More informations about this toot | More toots from modrak_m@bayes.club