diff options
Diffstat (limited to 'posts/2023')
| -rw-r--r-- | posts/2023/grants-for-operating-systems.md | 158 |
1 files changed, 158 insertions, 0 deletions
diff --git a/posts/2023/grants-for-operating-systems.md b/posts/2023/grants-for-operating-systems.md new file mode 100644 index 0000000..270fdac --- /dev/null +++ b/posts/2023/grants-for-operating-systems.md @@ -0,0 +1,158 @@ +<!-- +.. title: Grants for Operating Systems +.. slug: grants-for-operating-systems +.. date: 2023-11-14 10:00:00 UTC+01:00 +.. tags: arch linux, funding, ngi, nlnet, prototype fund, software development, sovereign tech fund +.. category: archlinux +.. link: +.. description: +.. type: text +--> + +Over the past years I have written (unsuccessful) funding applications for free software projects, associated with the [Arch Linux] Operating System. +This article is about my experiences with applying for numerous funds and my advice for people trying to get their work funded. + +TL;DR: Writing funding applications is extremely tedious and the selection process mostly intransparent and discouraging. Depending on what you apply for and who you apply with, you may never get funding due to other, additional factors. + +<!-- TEASER_END --> + +## Projects, projects, projects + +For some time now, I have been working as a freelance software developer. As such, my work time is more flexible, than that of an employee. +As *my hobby*, I have - aside from packaging - been working on several projects related to [Arch Linux], such as [archiso], [releng], [repod] and more recently [ALPM related Rust crates]. +As you can probably imagine, working on a software project can be very time consuming, especially when it comes to longterm maintenance efforts or improving legacy code. +With all of the above projects I have mainly been working on infrastructure related topics. Do note, that even this handful of projects is just a small subset of all the infrastructure related [software projects], that keep Arch Linux rolling. All of these are developed and maintained by volunteers in their free time. This is very different from Linux distributions with commercial backing (e.g. [Fedora] or [Ubuntu]), which often have a budget for internal projects. + +After spending significant time with packaging and package related tooling, I did quite a bit of work on [repod], in the hopes of improving the package repository management story for [Arch Linux] (see [Managing binary package repositories]). +This allowed me to estimate work packages and get a grip on how long implementation of certain features would take. +Most of the work on the project took long consecutive hours and was therefore not necessarily something just for a quiet afternoon or evening. +At this point I was contemplating to work full-time on the project for a bit to get more features off the ground. + +## Funds + +As a German citizen I am blessed with local, as well as European assocations and foundations, that provide funding programs for free software projects. +In 2022 I evaluated the programs of [ProtoType Fund] (funded by German tax money, driven by [Open Knowledge Foundation Deutschland e.V.]) and [NLnet] (funded by [European Commission] money). +Both have funded projects big and small over the past years and select their projects in several rounds per year based on received applications, which are reviewed by a public jury or an intransparent gremium of unknown people. +Funds are provided to freelance individuals or small groups of freelancing individuals. + +[ProtoType Fund] has - as the name implies - a focus on prototypes of public interest tech, which can be funded with up to 47.500€ for six months per project. The association selects projects with the help of a rotating, publicized [ProtoType Fund jury]. + +[NLnet] interfaces with several programs of varying points of focus, which it maintains with several other organizations and provides funds between 5.000 and 50.000€ per project. The foundation selects projects using a multi-stage process based on an unknown set of *"independent experts from the internet and open source field, academia and the public sector"*. +Large funding sums, that are handed out by [NLnet], are in fact provided by the [European Commission] *Next Generation Internet* ([NGI]) initiative and are made available to [NLnet] and a set of other organizations through [NGI Assure] and [NGI Zero], to be spent on projects. + +In 2023 I evaluated the [Sovereign Tech Fund], which is a relatively young organization, that offers funding for maintenance and improvement of existing ecosystems. Spending German tax money, the organization provides projects world-wide with funding starting at 150.000€. +The review process for applications is coupled with German law and no specific information on jury or reviewing experts can be found. + +## Writing applications + +As mentioned in the beginning, writing applications is quite tedious, especially if you are not used to writing these types of texts. +My takeaways from having written a few by now, and not being a professional in that field, is: + +* ensure that the content and framing of your application is a good match for the theme of the grant you apply for +* start early +* do several iterations +* invite people from various backgrounds to read the application and provide feedback +* use simple language +* do not get too technical (non-technical people will review your application, too) +* stay concise +* ensure you specifically target mentioned points of interest (aka. "hit the buzzwords") + +Although this process may seem overwhelming at first, it is useful to look at it from the angle of promoting your project to non-technical people and those unaware of your field of interest. +It is important to stay concise while doing so. All application forms have either a char limit (😬), or a word limit. + +To give a rough outline of an application process, it is worth noting, that although one has handed in an application, the review process only starts after the deadline of the application process has been reached. +Depending on organization, the review process may take between four weeks up to three months for an applicant to receive some form of acknowledgement or turn down. +If the funding organization has a multi-stage process, they may get back with further questions and clarifications. + +### ProtoType Fund + +In 2022 I wrote applications for [repod] in two funding rounds (12 and 13) of [ProtoType Fund]. +While the process was straight forward, the char limit of the submission form proved quite annoying. + +The evaluation process took around eleven weeks in the first application round and ten in the second. Both applications were turned down. +There seem to have been 171 applicants in round 12 (of which 21 - 12.28% - were selected) and 150 in round 13 (of which 23 - 13.45% - were selected). +The numbers outline, that competition is quite high for these grants. + +Although a project like [repod] meets probably some of the funding criteria, in hindsight it is likely too much of a backend and infrastructure project to be of real interest to [ProtoType Fund]. +Unfortunately, [ProtoType Fund] does not provide any specific feedback when turning down an applicant (the answer is kept quite formal and generic). This makes it very hard to understand whether a problem with the form or the topic of the application has been the grounds for rejection. + +### NLnet + +In 2022 I also started writing applications for [repod] for [NLnet] managed funds (i.e. [NGI Assure] and [User-Operated Internet Fund]). The application process was straight forward, this time with an enforced word limit. Due to technical issues, I did not receive a confirmation email for the applications, but after prompting the [NLnet] team about it, they replied swiftly and confirmed them manually. + +Both applications were turned down within seven weeks in the first review round. To the applicant it is unfortunately not transparent how many other projects applied and even finding a program-specific list of funded projects is a bit tedious on the organization's website. The email answer did not provide further details as to whether a problem with the form or the topic of the application has been the grounds for rejection. Receiving a generic rejection message makes iterating on potentially just misaligned applications not possible. + +In a second round in 2022 I applied again for [repod] for [NLnet]'s [User-Operated Internet Fund], [NGI Assure] and [NGI Zero Entrust] programs. All three were again turned down within seven weeks in the first review round (in fact I only ever got rejection mails for the [NGI Assure] and [User-Operated Internet Fund] programs, but after contacting [NLnet], it became clear that the rejection mail for [NGI Zero Entrust] just got lost). + +In a third round in 2023 I applied for a new set of [ALPM] related projects (more on those in a future article) with [NLnet]'s [NGI Zero Core] program. After a bit more than seven weeks I received a reply, that the application had been chosen for the second round of reviews and that I would receive further questions. Three days later I received a list of more in-depth questions, that I needed to answer (and did to the best of my abilities). +Six weeks later I received the rejection for this application as well, again with a generic message. + +### Sovereign Tech Fund + +In 2023 I applied for funding for the [ALPM] related projects as part of the *"Improving FOSS Developer Tooling"* topic of the [Sovereign Tech Fund challenges]. +The application process was fairly well laid out, albeit on a tight schedule. + +After a bit more than eight weeks I received a generic rejection email, that again does not help the applicant to improve on their application. +Another seven weeks later, the organization announced the [selected projects of the challenges program]: Nine out of 70 (12.86%) projects had been selected. + +A more generic application for [Sovereign Tech Fund] for the [ALPM] projects has also been sent, but given the above success stories, I am not holding my breath. +Furthermore, the organization appears to target more mainstream and industrially interesting Operating System projects such as the [Yocto Project] and [Gnome Foundation] (which each are backed by other companies and foundations as well). + +## Conclusion + +Although the previous sections read rather depressing (which admittedly they are to me), I think there are a few takeaways to be had from all this. + +Several things come to mind, when evaluating past experiences with funding organizations: + +* organizations providing funding for free software projects are very important for the ecosystem as a whole and there should be more of them +* the industry at large should be funding all software they rely on for their business (news flash: most do not) +* organizations providing funding opportunities should be more transparent about how their decision making processes work (especially if they are handling public money) + +For people trying to apply for funding for their free software project I would like to provide the following advice: + +* write applications with several people to bounce off ideas and to share the discouragement of receiving many rejections +* get ready for many rejections, unless you happen to hit a funding organization's sweet spot +* getting funding for infrastructure related topics, especially on Operating Systems without a huge interest in marketing, is *hard* +* write *many* applications +* document your process, as it is easy to lose track of ongoing applications due to the long waiting time and low probability of receiving a grant +* if you are working on a low-level backend related topic, [ProtoType Fund] is likely not what you should be applying for +* do not apply for funding for a topic, that you might be working on (slowly) in your free time, to not block yourself on waiting for a reply from a funding organization + +To sum up my past efforts: + +* I wrote seven applications for [repod] and three for [ALPM] projects (half of which were probably terrible) +* out of ten only one even got considered for a second round review +* in total I waited 56 weeks on replies since March 2022 (not counting still ongoing applications), averaging on 8.5 weeks per application + +I will likely try to apply for funding again at some point in the future (but most likely not with [NLnet], but more on that in the next article). +However, if you are interested in funding any of my work for [Arch Linux], consider contracting me for a project or sponsoring me on [GitHub]. + +Closing, I would like to thank everyone, that helped write and/ or improve any of the applications written in the past years. + +[Arch Linux]: https://archlinux.org +[archiso]: https://gitlab.archlinux.org/archlinux/archiso +[releng]: https://gitlab.archlinux.org/archlinux/releng +[repod]: https://gitlab.archlinux.org/archlinux/repod +[ALPM related Rust crates]: https://gitlab.archlinux.org/archlinux/alpm +[software projects]: https://wiki.archlinux.org/title/Getting_involved#Software_projects +[Fedora]: https://fedoraproject.org/ +[Ubuntu]: https://ubuntu.com/ +[Managing binary package repositories]: https://sleepmap.de/2022/managing-binary-package-repositories/ +[ProtoType Fund]: https://prototypefund.de/en/ +[Open Knowledge Foundation Deutschland e.V.]: https://okfn.de/en/ +[NLnet]: https://nlnet.nl/ +[NGI]: https://www.ngi.eu/ +[ALPM]: https://gitlab.archlinux.org/archlinux/alpm +[User-Operated Internet Fund]: https://nlnet.nl/useroperated/index.html +[NGI Assure]: https://www.ngi.eu/ngi-projects/ngi-assure +[NGI Zero]: https://www.ngi.eu/ngi-projects/ngi-zero/ +[NGI Zero Core]: https://nlnet.nl/core +[NGI Zero Entrust]: https://nlnet.nl/entrust +[ProtoType Fund jury]: https://prototypefund.de/en/apply/jury/ +[Sovereign Tech Fund]: https://sovereigntechfund.de/en/ +[Sovereign Tech Fund challenges]: https://sovereigntechfund.de/en/challenges/ +[selected projects of the challenges program]: https://mastodon.social/@sovtechfund/111261375834330869 +[European Commission]: https://en.wikipedia.org/wiki/European_Commission +[Yocto Project]: https://yoctoproject.org/ +[Gnome Foundation]: https://foundation.gnome.org/ +[GitHub]: https://github.com/dvzrv/ |