Maximising donation input from community

[seniorivn]

There are some discussions in the community about fundamental issues of nixpkgs, lack of maintenance in some parts of it, security concerns, issues with releases etc

There are a lot of disagreements on what is the best way to approach each of this, but at the end of the day someone is gonna have to do the job, and sometimes it requires a lot of resources.

In my opinion the best way for community to approach this is to delegate implementation of resource intensive solutions to a full time workers(members of the community ofc)

Which means nixos project needs to focus on it’s finances, more importantly income(donations)

There should be an estimate of the costs of planned projects and goals per project and monthly and all of that should be regularly seen by an average member of community, therefor advertised to them.

People prefer to donate to a specific goal so most pressing issues can be presented as a reason to donate to the project and therefor solve issues.

It’s ok to ask for donations on every occasion:

1. there are some educational videos on youtube, they could include an ad to donate to nixos project, so maybe active nix youtubers could be contacted and asked to do that in the future(possibly with optional affiliation links to incentivize participants to do that)
2. github issues could have a pinned issue with invitation to donate
3. discourse could have an ad banner with information about donations and donation goal

I post it in Marketing category because it’s mostly about advertisement

[blaggacao]

Nix Community | Nix & NixOS

Nix Community | Nix & NixOS

Donate | Nix & NixOS

Would all three require small amendment to reflect those goals (“paid effort on maintining nix repositories”).

[domenkozar]

As I’ve done three such campaigns (and learned a few things) I’d say what matters the most is finding companies/individuals that would really like something fixed and then setup that thing as a goal.

I plan to do more of such funding campaigns, but first I’m focusing on getting the macOS effort going.

If someone finds they’d like to contribute to a cause or set up such campaign, feel free to contact me.

[seniorivn]

wouldn’t you agree that there is a room for improvement on donation advertisement?

Shouldn’t it be addressed in an organized way?

As i understand “nix for macos” is already fully funded, correct? Or should we say has reached minimal operational funding?

It might be helpful if you described how you advertised it, and maybe suggest how we should make a more broad campaign to support all nix related efforts all the time

[blaggacao]

I think @davidak 's suggested https://snowflake.coop is very interesting.

I’d hope this platfrom would come up with ingenious angles to the problem.

I’d guess I’d be interested in allocating small amounts to a kubernetes-with-nix campaign (outside of NixOS/nixpgks !) and the divnix/devos project.

[blaggacao]

@seniorivn

Shouldn’t it be addressed in an organized way?

I think a blue print community process could go a long way (like blueprint laws issued by federal governments):

“Look, do this, register there, set it up such, publish here, here & here.”

It’s home? Don’t know. People will probably say “the manual”.

[domenkozar]

I agree there’s room for improvement and being very specific where we should advertise donations would go a long way.

We do have it on our website and GitHub, but it could be more explicit.

I think it would help however to go to each company in the community and ask them to contribute.

All of that requires someone to do the heroic work

It’s not easy to transfer my knowledge about this as it’s not a matter of writing a blog post. But I’m happy to help/guide any such effort!

[seniorivn]

@domenkozar

I think it would help however to go to each company in the community and ask them to contribute.

All of that requires someone to do the heroic work

having a Nix evangelist as a paid position in nixos foundation might be a good idea at some point, such position could include regular public talks, fundraising, just like politicians do when they are campaigning

But from perspective of a company it seems logical to prefer support of a specific issue(just like with the mac campaign) than a foundation with a broad mission, so it’s again a question to setup flexible system of donations with goals and projects

[domenkozar]

Exactly, at this stage it makes sense to fund specific, achievable goals.

Once we reach a certain size in the industry, I’m all for doing something similar than Haskell Foundation.

[ChrisJefferson]

I realise this might be difficult to organise, but I (from a business perspective) find it very hard to donate. However, it’s relatively easy to pay for something. This (from the outside) seems to be how sqlite works for example – you can pay for support, which gives you email replies from the sqlite developers.

I realise this raises many complex issues and might not be acceptable for NixOS.

[nixinator]

I like the term customer success engineering…, rather than support… Making customers and users successful with the technology… This stuff is complex and because nix changes a lot of things, it requires an open mind and the ability to learn new things. Some ‘titans’ in the tech industry really hate change.

Many companies just want to stick with the crowd, if another company says technology X is good, then company Y will adopt it… because ‘it works for company X’.

This is really dangerous group think… It’s important to get Proof of Value and Proof of Concept trials of Nix into companies… Mostly this is done with engineers introducing nix from a grass route perspective , or ‘slowly sneaking nix in to your companies toolset…’

The support model was the redhat model in the early days… you can learn a lot from history!

It’s a conundrum, how do you make a ground breaking product and get every one paid… while keeping your open source and FOSS integrity…

I’ve not seen a solution as yet!! Maybe it just can’t be done!!

Nix was off my radar for 10 years, so at least the marketing is working and more people telling the world about this stuff is only a good thing!

Interesting times!

[blaggacao]

@ChrisJefferson

I realise this might be difficult to organise, but I (from a business perspective) find it very hard to donate . However, it’s relatively easy to pay for something.

It is hard for the buying center. Probably impossible. However, I’d sustain it’s easy on a visionary or strategic level (still comes with tax breakes / brings you sparkels / access to talent / visibility / bla bla bla).

In any case, it has to be appealing. I suspect, in that context, effectiveness — that is output per effort/dollar-donated — is a good proxy for appeal.

Hence, we’d have to make sure we master the “science of effective organizations” — as per their language and understandinga.

For example I’m explicitly offering a (small) donation to the community. But should no reasonably distinguishable counterparty emerge / exist / or I’m being redirected to in the first place, any mutual aspirations about a meaningful and effective charitable transaction are somewhat in vain. (if that makes sense; time and money are really just substitutes)

[clefru]

Thanks for posting this, @seniorivn . I set up a small monthly donation. While evaluating whether I should do this, I looked at the links above that say that cache.nixos.org is the expensive part. It almost stopped me from donating, because seeing my donation go to CDN providers has zero emotional gratification. I rather donate to a project that implements Suggestion/Feature: use Bittorent or IPFS to download packages and turns down cache.nixos.org, so if you can come up with some project based donation plan, that would be great.

[asymmetric]

It would be really cool if we could get more transparency into the costs/budgets of the Foundation. Right now it’s hard to understand how badly my financial contributions are needed.

[asymmetric]

One more thing: I was trying to figure out how to contact the foundation to ask them if there are plans to make its finances more transparent, and I couldn’t find an email address, IRC handle, anything.

I created a PR addressing that (heh!), but IMO it would look more professional if the Foundation had its own email address.

[graham33]

This is a great topic. My company is using Nix (nixpkgs) heavily, and we’d like to contribute more and see the community continue to thrive, so I wanted to add some thoughts from that point of view.

I definitely agree that having some paid maintainers would be really helpful for long-term sustainability. I also agree that supporting specific areas, rather than just general ‘please donate to the NixOS foundation’ requests, is likely to attract more interest and probably more donations. In our case, for example, we’re big users of the nixpkgs Python ecosystem, and we’d be keen to see some kind of full-time maintainer within the community. This is something we’d likely be willing to contribute to.

I think there might be some interesting questions with such maintainer roles around governance and ‘independence’. Ideally, paid maintainers should act for the good of the whole community rather than concentrating on the things donors care most about. It probably makes sense to have some kind of charter for the role agreed up front. Transparency is also important. I suspect other open source projects have tackled this.

Aside from language ecosystem support, there are also cross-cutting areas that we’re interested in, for example support for new architectures and technologies, or specific features like the content-addressed store. There are also features that we really rely on that we want to see maintained, for example things like the hashed mirror functionality, or sometimes legacy ecosystems that we might (unfortunately!) want to keep around a little bit longer than the fast-moving open source project would otherwise.

It’s worth explicitly saying that some work items within a system like nixpkgs can have non-trivial value to companies, and it can be relatively easy to attach a dollar value to them, simply by estimating the amount of time it would take a company employee to do the work instead. So I wouldn’t underestimate willingness to contribute for specific actionable tasks.

One more indirect way that companies can contribute to nixpkgs is via third party consultant/contractor companies. For example, we’ve paid for such a company to add support for several new libraries and features to nixpkgs, and continue to do so. If there enough companies like us doing that, it encourages a strong ecosystem of such third party companies, who in turn will create paid positions for people to work on nixpkgs and create a demand for people with Nix skills. Clearly companies like mine can also make code contributions themselves too, and that’s something we do.

Finally, I would say there’s some scope for companies to help evangelise Nix within industry forums, with vendors and even competitors. It’s very much in our interest that it continues to be widely used.

I’d be more than happy to discuss any of this further if people are interested.