<jpk68:matrix.org> vtnerd: I know you didn't ask me, but I agree with kayabaNerve that CBOR would be better than MessagePack

<jpk68:matrix.org> The former has deterministic encoding, and solves a lot of the latter's standardization issues; it also separates text from binary data (which is good IMO).

<vtnerd> I don't know that we need deterministic output, it's going to be annoying to implement. I don't know what you mean by separating text from data

<jpk68:matrix.org> msgpack/msgpack #121

<vtnerd> Like cbor deterministic output is nice for signing but hell if you're trying to avoid a dom when outputting (because c++ variadic templates have unspecified order of being invoked)

<vtnerd> I don't understand the criticism of msgpack in that issue - there's a distinct string type for utf8 and a separate binary type

<vtnerd> The issue was closed over 10 years ago, so maybe this has been resolved (still going through issue)

<jpk68:matrix.org> Not every version of the spec supports it, though. Consumers of the RPC might be using incompatible MessagePack libraries, when they think it's the same thing

<vtnerd> I mean this happened 10+ years ago? I haven't seen an implementation that has yet to be updated

<vtnerd> And if anything that deterministic output is pushing me against implementing cbor, it's literally shit for output from c++ data structures unless I can figure out how to guarantee an expansion order

<jpk68:matrix.org> Are you planning on implementing your own parser for it?

<jpk68:matrix.org> It could also be because C++ is shit ;)

<vtnerd> Or else buffer specially in cbor output I guess. Probably implementating my own output

<jpk68:matrix.org> Non-deterministic encoding was enough of a concern for WebAuthn, for example, which uses CBOR. I don't know, it just feels weird that spec-compliant parsers can arrive at different encodings with each valid

<vtnerd> Right but webauthn specifically needed signatures. Our use cases for cbor are "specifically allow multiple format types to make it easier for user"

<vtnerd> I could realistically add cbor and msgpack and let user figure it all out - deterministic order would only be present with cbor - assuming I can even implement it in a sane way.

<jpk68:matrix.org> Fair point. It's not a big deal or anything, just my preference

<dancer> I have never been roasted so hard. thank you. this is lovely. also, we will move on

<dancer> people assume you guys are a bunch of autistic nerds, and that might be true, but who cares and also.. who knows how awesome your sense of humor is?

<ofrnxmr:xmr.mx> off topic

<ofrnxmr:xmr.mx> #monero-offtopic:monero.social

<ofrnxmr:xmr.mx> Stop spamming please

<321bob321> Kick

<vtnerd> looked into cbor some more. cbor is not deterministic, unless the implementation implements the dbcor or CDE variants. looks like dcbor is newer/preferred (but not really sure this is annoying)

<jpk68:matrix.org> Deterministic encoding seems to be part of the original CBOR spec, it's just not entirely enforced for it to be valid

<jpk68:matrix.org> On the other hand, it's not part of MessagePack at all

<jpk68:matrix.org> I must admit, though, I wasn't aware of that until I just looked it up :)

<dancer> @321bob321: cute.

<dancer> absonned folly will not define us

<ofrnxmr:xmr.mx> @banhammer:matrix.org

<dancer> i am trying to be nice about it. a fork is coming forth

<dancer> @vtnerd: we should talk more about this

Vibe-coded repo, that has imports from github which are 404 now, and p2pool.cache in the repo as a cherry on top :D I never laughed so hard in the morning :D

<ofrnxmr:xmr.mx> Yep. Total spam

<ofrnxmr:xmr.mx> Bro making programs using duckai

<dancer> sech1: mrelay.p2pool.observer/m/monero.social/fJyGuewLucuQThaSMwKFEywg.png (image.png)

<dancer> I've told you that I love a good roast, and that I can appreciate God's sense of humor.

<dancer> look at what is happening all around you.

<dancer> you literally can't develop anything better than this. not because you're not talented. it's because the entire concept of "monero is money" lacks the ability to be what money is. it is hillarious. but we will remove gthe problem on July 1st.

"we" = me, the voice in my head, and my ai?

<ofrnxmr:xmr.mx> +drugs

<dancer> sech1: look at who tries to do good, and look at who is trying to stop us. you are not "us."

<ofrnxmr:xmr.mx> Who is "us"

<ofrnxmr:xmr.mx> Bro did everything except answer the question

.merge+ 10621 10624 10625 10628 10629

Added

.merges

10512 10531 10533 10543 10621 10624 10625 10628 10629

FYI I've just updated the DNS records with 0.18.5.0 hashes so auto update should trigger

.merge+ 10578

Added

.merge+ 10665 10623

Added

.merge+ 10642

Added