XSF Discussion - 2020-05-26

Neustradamus_ Do you mean a feature where a ban can be defined for a certain period? That might make for an interesting add-on - but I'm thinking that more than 90% of the use case of banning would be to remove people permanently.

But I do not immediately see a reason why something like a ban period cannot be added.

Guus: Yes! But in XEP-0045 not in a new

Neustradamus_ XEP-0045 is not final, so you can propose a change.

I'd suggest avoid being banned by not going on people's nerves in the first place :->

😂

Neustradamus_, why is it important to be in XEP-0045 and not a new XEP?

Link Mauve, AFAIK, @xml:lang is allowed everywhere and you don’t have to add support for it

if you really wanted to specifically point at @xml:lang, you could declare it in the schema under a different prefix -- nobody forces you to use xml as prefix

> https://github.com/zoom/zoom-e2e-whitepaper Impressive that they got Mathew Green as an author

> .pdf

*sigh*

yeah 😛

COMMIT ALL DEM BINARIES!!!

hey, at least it's not a Word document.

that zoom e2e thing makes no sense to me

"Ed25519 is not FIPS approved but it's better, so we just do Ed25519 *and* P-384 for everything. Ed25519 for the securit and P-384 to make regulators happy"

😀

This is what happens when you design crypto protocols with business in mind

It is better to have a full perfect XEP that several.

Why not put everything into a full perfect RFC 6120?

don't feed him :D

😀

I'd write the IETF!

fwiw, to me, this particular feature seems more suitable to be injected in the existing XEP more than as a stand-alone one.

Neustradamus_, it doesn't make a difference if you open issues everywhere "support new XEP for banning with period" or "support new version of XEP-0045"

be nice, guys.

Guus, I think XEP-0045 is already too big and already has too many things not properly defined whereas others are defined down to the detail

I mean, just a random example but but §10.3 and §10.6 are 1.5 pages virtually identitical text

I completely agree with that, Guus

larma not adding this won't change that, but at least it'll remain consistent if this is added. Also, I'm most unlikely to die on this particular hill. ✏

I mean, for this specifically I also won't mind

The only problem I see is that it moves finalizing further away

Sure. I think we're pretty much agreeing on things.

I think that xep-0045.xml should only receive bugfixes at this point, new features should go into inbox/

not sure if that’s your choice ;P

I'm also not going to die on the hill that is the discussion on what hill I'm going to die on.

also, I find this morbid fascination slightly disturbing.

you started it

Yes, talking about my own fascinations mostly 🙂

also, the wife and I are watching The Good Place atm, some slight increase in comical morbidity is to be expected. (note: The Good Place is full of plot twists, so don’t read a plot summary if you’re curious)

Trying to play out Netflix eh? Fair warning: the end boss is hard.

Sigh, jabber.org

Have you tried turning it off and on again?

the turning off part is quite battle-tested at this point

I'm not sure about that

(ignore this) ✏

interesting how frequent I now see message edits being used in XMPP 😀

very useful feature apparently

And WhatsApp doesn't have it!!!

pathetic!!!

But WhatsApp was based on XMPP, therefore it's XMPP that's pathetic!!!!1!!!!!!eleven

vanitasvitae: There is always the problem about XEP-0308: Last Message Correction: History, there are all changes...

Neustradamus_, ask your client dev to fix this

It's possible for most cases

Also possibly combined with 0421

pep.: The XEP has been updated?

Which

XEP-0308

It doesn't need to?

The fact that clients don't group corrections in MUC history is only due to them not being able to assert it's the same user. 0421 gives them the ability (for MUCs that have it deployed, obviously)

rion: ^

Jabber.org has problem since several days no?

yes. The admins are already aware of the problems.

Guus: Thanks to confirm me! I have not contacted several days ago, but today there are lost messages...

I've emailed Isode support. I've no idea what's wrong, and a restart didn't help

AFAICS it's working, just overloaded.

I'm watching authentications get logged. If I telnet to the port I get responses (slowly)

I sometimes got successful connections when I tried to debug why I ran into timeouts while checking it

so a fun DoS?

Possibly.

Actual DoS or post-restart flood?

Don't know.

needs more cloudflare

Seems to be completely responsive now, to me.

And CPU is no longer pegged.

Great

Dear Jabber.org team, it is possible to have a special mucroom for it?

Thanks for checking... if you don't see the cause then pretty sure I had no chance. I'll look at traffic next time perhaps

I couldn't see anything particular from any one host, it just seemed to be 'lots'.

As Zash says, the same sort you normally see swamping post-restart. ✏

'lots' being one more than many-many-three?

It's been to long since I opened a discworld book...

https://discworldquotes.tumblr.com/post/159017563008/everyone-knows-trolls-cant-even-count-up-to

RIP

also: $ curl -vvsSL -o/dev/null https://observe.jabber.network 2>&1 | grep X-Clacks-Overhead

it is surprisingly difficult to extract an HTTP header from curl

> Any namespaced elements can be used within a <command/>. Hmmmmmm.

Does anyone know if anything anywhere does ad-hoc commands with anything other than dataforms?

I think my draft for ToS tried to