Notices by jon ⚝ (yala@degrowth.social)

@postmarketOS I was very surprised recently that Alpine's musl resolves DNS differently than glibc.

- https://biriukov.dev/docs/resolver-dual-stack-application/5-getaddrinfo-from-musl-libc/
- https://jvns.ca/blog/2022/02/23/getaddrinfo-is-kind-of-weird/
- https://wiki.musl-libc.org/functional-differences-from-glibc.html#Name-Resolver/DNS
- https://martinheinz.dev/blog/92
- https://purplecarrot.co.uk/post/2021-09-04-does_alpine-resolve_dns_properly/

In our case it breaks the search domain for Docker containers which would like to discover each other via DNS. This limitation of the search domain might also apply to mobile devices on postmarketOS in networks that set one.

@tomw I don't feel a generic rejection needs any justification. As a first, I just don't like the term.

My reasons for that can be debatable, but not the sentiment.

@FediverseSymbol Nice symbol. Do you also have a proposal to replace the term Fediverse? I know many people, me included, who don't want to or cannot use it as a term in political debates, myself included.

Social Web was once one proposal, Open Socials is another. Maybe you've heard of other alternative naming that is more accessible in its name does not put the effort into a niche?

@zak
GitLab CE just works (TM).
Email service desks with custom email matures up. Adding external participants to confidential issues is useful, too.
There is the work items migration, which unifies issues, incidents and tasks, offering new ways of interlinking.
No epics, which is a label for us.
What have I missed?
Ah ja, the Kanban board and separate incident timelines.
Also integrates well with external services and the rest they have to offer. The API and SSO integrate it in return.

@aral
Do as you please. We're not going anywhere.
@ecobytes

@aral As an @ecobytes member, we could always fire up a few containers for the small-web.org.

@aral I was thinking of running an instance for the computational common allmende.io, alongside the existing lab.allmende.io GitLab.

We don't always need all of its features and sometimes I just need a git and that's it

The application is simple and we still have enough compute left at @ecobytes

Also I wanted to offer you our self-managed PowerDNS at some point, which is now closer to being available for API use, if you wanted to test that.

Should we roll a git.allmende.io Forgejo instance?

@evan
Yes, what happened to C2S anyways?
@erincandescent @Seirdy @starshine

@ntnsndr @GuerillaOntologist @skyfaller @shauna @Matt_Noyes Thanks for mirroring the arguments. As I might feel tempted to answer to the second example later, I'm inclined to offer a slight variation of the prior:

Nobody accused anyone of lying above, if I'm not mistaken, which might leave us with the following, instead:

"The techno-social architecture of distributed ledgers is not sufficient to redistribute wealth equally to all, why we must choose to abandon this and persue other prospects."

@GuerillaOntologist @ntnsndr @skyfaller @shauna @Matt_Noyes Josh, can we name three specific critiques that are missing a response in this thread?

I'm with you about scrutiny in empirical argumentation of the subject.

From the above discussion, I'd pick:

1. Distributed ledgers are not decentralised in terms of protocol design and P2P bootstrap.
2. Philosophical arguments about novel governance and renumeration schemes ignore the fundamental critique of immutable, ungovernable protocol design.

@GuerillaOntologist @ntnsndr @skyfaller @shauna @Matt_Noyes

3. The identification and naming of quantification as institutional governance model, which is used to enscribe and reproduce monetary power hierarchies within a private property narrative that encloses common resources, limits the discourse within groups and the public plus assumes and presents social stratigraphy as an ontological fact.

@ryanc
This stands out for me:

"discrimination is justified by the difficulty of updates to the legal system and the cost of updating IT systems"

It means legal and informational systems are bound by material constraints, such as availability of labour, coordination costs, modular and standardised architectures, plus regulatory procedures.

Which all in all makes a strong argument for using FLOSS collaboration patterns to decrease auxiliary costs for all participants in the distributed system.

@smallcircles Maybe just put the list in Wikidata? Seems easier to maintain collectively in the long run.

You can always create any kind of representation from that raw data.

@evan Many fedi! (:

@youronlyone Okay, thanks for your details. I will see how this can integrate with the Docker setup. @alan

@youronlyone

Are there good instructions on how to do so?

@alan