An interesting proposal from @Chris G came up in the discussion about grant funding. This community could pay for some cloud/IaaS resources for its members' projects, to help them experiment. In Christopher's words:

Make infrastructure available — Need a virtual machine? Some lambda functions? Some storage?

Keep it simple — have 1 account, with multiple access groups so a person/team can keep their work private if they like (or in the open, per Maikel's nice point above), and just pay the tab on that account

I'd love to know — if the community offered to pay for some cloud/iaas resources, specifically to help YOU work on your not-otherwise-meaningfully-funded FoC project, would you take advantage of that? What resources would you need? How might you use them? I love the idea of making resources like this communally available. It reminds me of what labour unions used to offer, in the good old days when the rail unions had well-stocked libraries and offered classes and such for members. Or, if you prefer, what the early-day perks at Google or the dot-coms were like. Extra goodies that are meaningfully valuable to an individual, but relatively inexpensive for the larger collective to offer.

One thought that occurred to me after reading it: perhaps we should spend something to host this community ourselves. (No, I'm not advocating paying Slack cash money.)

That idea was floated when we were discussing leaving Slack. It fell on the same sword as the other promising ideas — the constraint there isn't the financial cost, it's the human cost. Moving means losing more people than we'd lose by staying put. But that's another discussion entirely — one I'm deferring until next year at the soonest, for the love of all things.... (haha)

Speaking of org structure, I presume we're not yet at the state where we have a non-profit? Just might come into play when it's time to create the account - but will presume we'd setup a regular account. (I work at IBM Cloud, I won't advocate for it, I like AWS and Azure, too, they have more goodies - but if we do have a non-profit or any mission paperwork, I might be able to cajole us some free/cheap resources.)

I feel like a non-profit is in the future, but I'm kicking that can down the road until it hurts my toe to keep kicking, since it'll increase the amount of ongoing work for me pretty substantially (in relative terms). But if something like this grant happens, or if the sponsorship money ever goes from basically what it costs for hosting and transcription to more like actually enough that we would want a community bank account, then yeah.. non-profit.

Interesting discussion! I guess that those kind of FoC community shared cloud resources could be useful for some project. One thing to keep in mind: there are often free tiers for cloud services which could be enough for some projects (like mine, see below). So maybe we can focus on things not already covered by this kind of offers? Like the need of virtual machines. And maybe the added value will be on the collaborative/sharing features? To illustrate the kind of needs, here is what I have done for adacraft, which is not founded and for which I was looking for free/cheap cloud ressources to make it works. I choose to do as much as I can into serverless because 1/ I like it, and 2/ it will mean less work/stress for me for the infra stuff. So, it is mainly static web (for the web portal and the modified Scratch editor), a DB (for projects and users info) and file storage (image, sounds, raw project files). Here are the cloud resources/services I use. Netlify for the static web files, for some few lambdas and for user authentication. FaunaDB as a servereless database. S3 compatible object storage hosted by OVH for the file storage. All this are paid services with free tiers that are very comfortable for a small project like mine. I also use GitLab.com as a code repository (also free). The only thing I am paying now is analytics from Netlify. Which cost 9$ a month. I use it because at that time it was the easiest and quickest way to have analytics. Since then, I thought it would be cool and fun to reimplement a very basic analytics system based on FaunaDB, but I haven't took the time for that yet. So for example, one need for me for a common cloud ressource would be a web analytics solution (Matomo or something else). One thing I want to play with in the future, is some kind of server side execution of some users code. At that time maybe I will need some VM or some more advanced lambda (if the Netlify ones are not enough).