XMPP Service Operators - 2026-02-09

> yeah its whatever, i have appropriate serverside limiting to prevent gajim or prosody from melting, i was just sharing with the class I found a bug in python3 that prevent's the random number generator from ever being seeded. ↺

among several other interpreter bugs

looking for alternatives to Gajim at this point. Security in python just seems far-fetched.

It doesn't seem like a good language to base a graphical client in.

I certainly have thoughts on this but not really on topic here is it?

> looking for alternatives to Gajim at this point. Security in python just seems far-fetched. in 2 weeks ill make a client 🚎 ↺

profanity

fr

> looking for alternatives to Gajim at this point. Security in python just seems far-fetched. so you think rewriting an entirely new client, using an entirely new library is then going to be secure? ↺

it is plagued by the same issue with "rewrite in rust", you want to ensure the ecosystem matures and there has been enough eyes on the codebase to find trivial security vulnerabilities. Anyways I will drop the topic as its OT

Anything is better than Python3/gtk4 at this point.

> Anything is better than Python3/gtk4 at this point. python isn't inherently insecure. gtk is used by most clients, and writing your own UI framework is going to cause even more vulns ↺

your aim to be "more secure" will end up leaving you with holes. Just saying

That and you shouldnt be thinking about hypothetical security problems and should be instead caring about demonstrable security vulnerabilities

"python has a bug that breaks the rng" okay, prove it affects gajim in a tangible way to break security

>> Anything is better than Python3/gtk4 at this point. > python isn't inherently insecure. gtk is used by most clients, and writing your own UI framework is going to cause even more vulns That’s debatable tho

> looking for alternatives to Gajim at this point. Security in python just seems far-fetched. On which platform?

Very.

> >> Anything is better than Python3/gtk4 at this point. > > python isn't inherently insecure. gtk is used by most clients, and writing your own UI framework is going to cause even more vulns > That’s debatable tho any more debatable than javascript/typescript, or Java? ↺

its not usually the programming language which is the problem but the project

https://news.ycombinator.com/item?id=46944245

Interesting discussions about Matrix, and XMPP, in the thread

Couldn't think of anything worse than protocols becoming more and more adopted by government

Just my personal feelings though :)

silo users (like hn commentors) exclusively have bad opinions though

> Couldn't think of anything worse than protocols becoming more and more adopted by government in Germany, Matrix has a very dominant usage among government and public institutions. openDesk, which lately has gained a lot of attention and traction, has Matrix/Element as its chat solution and public health insurances use it (https://element.io/de/matrix-in-germany/projects/ti-messenger), as well as others (https://bundesmessenger.de/) I wish XMPP would get as much as dedicated public funding... ↺

> in Germany, Matrix has a very dominant usage among government and public institutions. openDesk, which lately has gained a lot of attention and traction, has Matrix/Element as its chat solution and public health insurances use it (https://element.io/de/matrix-in-germany/projects/ti-messenger), as well as others (https://bundesmessenger.de/) > I wish XMPP would get as much as dedicated public funding... Yeah great idea, rely on the government, so when you feel their boot you have no choice but to comply ↺

What Matrix is doing is signing their death certificate, XMPP needs more funding outside of begging a government for funding. Once you rely entirely on public funds, you are unable to oppose the government.

most of them are siloed off, because they know that as soon as they would federate, they couldn't comply with GDPR because of the syncing to all participating servers

> most of them are siloed off, because they know that as soon as they would federate, they couldn't comply with GDPR because of the syncing to all participating servers So even worse then, I can't tell if you are supporting this or opposing this. ↺

``` Your XMPP server wants to connect to jabber.fr, do you accept to share your personal data with it [Accept all the XMPP cookies] [Reject] [Customize] ```

> ``` > Your XMPP server wants to connect to jabber.fr, do you accept to share your personal data with it > [Accept all the XMPP cookies] [Reject] [Customize] > ``` XMPP cookies? ↺

Free cookies? where can I get these? /j

https://xmpp.org/extensions/xep-0464.html

Maybe thats what we need to do at FOSDEM next year... "Free cookie on sign up!"

> So even worse then, I can't tell if you are supporting this or opposing this. turning off federation is bonkers in my eyes, but so is using Matrix for the purposes they outlined 🤷‍♂️ ↺

> I wish XMPP would get as much as dedicated public funding... This won't happen without us actively pushing for this. The Matrix community has been very successful in this, but there's no reason that we cannot do the same. In the latest XMPP summit, we briefly discussed that this was desirable, and have started thinking on pragmatic first steps. There recently was a call for involvement on the XSF's members list around this, too.

Who has been very successful at this?

Not the matrix community for sure

https://mail.jabber.org/hyperkitty/list/members@xmpp.org/thread/47CJ5TN5SXIMGFBJEGMBTZETCOKK77Q3/ <-- I'd love more feedback there.

VC funded element yes

I consider them part of the larger Matrix community. Let's not split hairs.

its important to split these hairs

processone is one of many XMPP companies that has been very successful too right?

The important distinction is that there are people actively pushing for things there, where we have far fewer of those efforts.

I don't see any evidence that matrix has more efforts or adoption than XMPP

Fine - I don't particularly care to win that argument. My main point is that I'd love for our community to come up with (and execute) more on the outreach and positioning of XMPP.

XMPP *only* runs on 100% of android and apple phones for push, and all nintendo switch consoles, and zoom uses it etc etc, its got far wider adoption than matrix ever dreamed of

I agree. I also agree that hardly anyone _knows_ about this.

Yes but there's a huge gap on sending custom proprietary XML in a TSL socket and call it XMPP and the XMPP I (we ?) want to push.

I'm having real-world conversations with policy makers that _know_ about Matrix, but have never heard of XMPP. They're even stating that there's no RFC-based IM protocol. _That_ is the problem that I think we need to solve: we need to be educating people about the brilliant stuff that we have to offer.

eXtensible is literally the first word in XMPP, you can't call it not XMPP because someone extends it

I'd have to agree that there's no such thing as "no strings attached" public funding

An non federated XMPP implementation with just 5 out of 500 XEPs used is it still XMPP or a solution found by the company to easily transmit messages between A and B by taking a server and libraries of the shelf ? ✏

afk dinner

> I agree. I also agree that hardly anyone _knows_ about this. I would say more confs, push it at more events, Matrix has been very successful in getting Linux groups to use it. ↺

> I'm having real-world conversations with policy makers that _know_ about Matrix, but have never heard of XMPP. They're even stating that there's no RFC-based IM protocol. _That_ is the problem that I think we need to solve: we need to be educating people about the brilliant stuff that we have to offer. Policy makers never have and never will care about protocols, they care about products. eg ejabberd, openfire, element ↺

I also think we need to provide an incentive to adopt XMPP

a bit like how homes build up around business, we need to provide the service for people to "move into". Currently we have no real communities on XMPP other than dev chats for all the different clients.

BSD has a small community, theres a few channels like the open hardware channel, but for the most part theres not that diversity

Matrix there is...

For example, Matrix has been adopted by a few Minecraft Servers for their main platform over Discord, outreaching to gamers.

XMPP Minecraft Server when

I might write a POC idea me and Fishbowler discussed at FOSDEM to provide an easy way to install an XMPP server for newcomers, which doesn't mean dumping it into a docker container.

> XMPP Minecraft Server when nothing is stopping you, apart from potentially the hardware ↺

MC servers are notoriously inefficient

and the ecosystem is plagued with spyware, and scummy code, even if some of it is open source

micycle.net, a group of people I met up at FOSDEM come from those circles, we could make it happen.

> For example, Matrix has been adopted by a few Minecraft Servers for their main platform over Discord, outreaching to gamers. Reminds me of this blog post https://slugcat.systems/post/24-12-12-the-rise-and-future-fall-of-discord/ > 1. Some players make a Discord community for the server. > 2. Us in IRC go “fine I’ll sign up but I doubt we’ll use it”. > 3. IRC was dead a month or two later. Complete ghost town. > Discord was just that good. ↺

surely you mean luanti not microsoft ™️ minecraft ™️

> Reminds me of this blog post > https://slugcat.systems/post/24-12-12-the-rise-and-future-fall-of-discord/ > > > 1. Some players make a Discord community for the server. > > 2. Us in IRC go “fine I’ll sign up but I doubt we’ll use it”. > > 3. IRC was dead a month or two later. Complete ghost town. > > Discord was just that good. well I wouldn't say Discord was "good", Discord provided ease of use for newcommers, something IRC still majorly lacks. ↺

IRC you have to worry about: - What network - Registering with nickserv - Configuring your client to auth - channel modes and chanserv ✏

XMPP is definitely easier than that :)

> surely you mean luanti not microsoft ™️ minecraft ™️ luanti is niche, you wont branch out that way ↺

try to gatekeep XMPP only for freedom respecting software and you will find nobody will use it, open it up to all even if they run Windows, and write proprietary code, then you build an ecosystem where people can actually discuss to a range of different people.

In the blog that quote recalls what happened in 2015. IRC and for that matter XMPP were very different than what we have today. And the blog goes on to talk about switching away from discord. It's a fun read if you're interested in the topic

> In the blog that quote recalls what happened in 2015. IRC and for that matter XMPP were very different than what we have today. And the blog goes on to talk about switching away from discord. It's a fun read if you're interested in the topic I dont have the time to read it but my experience with getting rid of Discord involved relapsing 7 times because XMPP is like shutting yourself in a lead box, unable to speak to anyone on the outside. ↺

Too few people even know what XMPP is, your friends and family get angry at you for not using Discord/Whatsapp/Instagram

you struggle to make friends in person, when they ask what your Discord is and you go "actually I only use XMPP" ✏

my experience with discord and matrix is I've still never used either, seems useless

> my experience with discord and matrix is I've still never used either, seems useless Alright gramps, we know you are old enough that they predate you, but when I was growing up Discord was the standard, and thus its why I first used. I then branched off into IRC and XMPP and have never looked back, but for most people, they see it as a downgrade, not an upgrade. We need to be able to show them that XMPP *IS BETTER* than Discord. ↺

> Reminds me of this blog post > https://slugcat.systems/post/24-12-12-the-rise-and-future-fall-of-discord/ > > > 1. Some players make a Discord community for the server. > > 2. Us in IRC go “fine I’ll sign up but I doubt we’ll use it”. > > 3. IRC was dead a month or two later. Complete ghost town. > > Discord was just that good. https://discord.com/press-releases/discord-launches-teen-by-default-settings-globally ↺

Lets see how this is handled

> Lets see how this is handled People will just jump to Matrix ↺

> Video selfies for facial age estimation LOL if you use such a thing you get exactly what you deserve

No offence to the people here, but if you look at the matrix crowds, and the XMPP crowds, the XMPP crowds are much older, and the Matrix crowds much younger.

Why? I dont know

Obviously because you get smarter as you get older

mopar you aren't helping lol

your rigid purist views will not help here.

No purism here, I use what works best

no but you want to push FOSS-only on XMPP...

you need to compromise to get the usercount up.

welcome stratself,

> no but you want to push FOSS-only on XMPP... ...why would we not? ↺

Even on a FOSS-only front IRC and Matrix see more usage in the real world for group chats in my experience. Both FOSS and Gaming are demographics XMPP has not reached as much as it's peers

welcome all. that was fast

> ...why would we not? luani as an example, most people play Minecraft, if we then try to push a niche on a niche platform, you get a niche community. If you instead conceed that Minecraft is proprietary and not freedom respecting, however make an XMPP community for it, you are opening up a niche platform for a mainstream userbase, which helps boost the userbase. ↺

> Even on a FOSS-only front IRC and Matrix see more usage in the real world for group chats in my experience. Both FOSS and Gaming are demographics XMPP has not reached as much as it's peers I think Matrix thrives on the fact it is a little bit of a silo, those who are techy enough to self host it, and have the money to afford a datacentre to do so, can still have federation, but for the majority, when you talk about matrix people know element well, and element is the standard on all devices, so it kinda replaces Discord with the all in one Element client on the Matrix main server. ↺

> you need to compromise to get the usercount up. first you'd have to convince me why I should care about this when everyone I need to contact uses XMPP already ↺

I think this is the same reason signal and telegram are so popular too, they are silos that make it easy

>> I'm having real-world conversations with policy makers that _know_ about Matrix, but have never heard of XMPP. They're even stating that there's no RFC-based IM protocol. _That_ is the problem that I think we need to solve: we need to be educating people about the brilliant stuff that we have to offer. > > Policy makers never have and never will care about protocols, they care about products. eg ejabberd, openfire, element That's simply not true, not generically. I'm talking with people that very clearly understand this distinction between product and protocol, can point to various RFCs for other technologies (like email or ldap), know about Matrix, but not of XMPP.

again discussed at FOSDEM, XMPP has the issue of: - What client should I use? - What provider should I use? - What channels should I join?

Because no one cares about the protocol, only the product

nor should they

Clients are a foreign concept to most people, "what I can pick any client and they all work with one another?" is a common question I get asked. Then providers can be overwhelming, and in most cases Germany is the only real pick you got, which is why I would love to become a public provider (legal issues like I said at FOSDEM).

Then when it comes to channels, the standard is SJN, which is renown for only including channels the operator agrees with, I shall not name who the operator is, but this is a problem!

> a bit like how homes build up around business, we need to provide the service for people to "move into". Currently we have no real communities on XMPP other than dev chats for all the different clients. I do have the same opinion, seeing no community on here whatever matrix did can suck, but at least they spent on rallying a community complaining about how sucky they are ✏ ↺

provider has only 1 sane choice: yourself

> provider has only 1 sane choice: yourself and this is why you are unhelpful ↺

you cant accept that you are likely more intelligent than the majority, and refuse to accept that service providers enables everyone to enjoy the same freedom you do, albeit being reliant on a provider.

> I do have the same opinion, seeing no community on here > > whatever matrix did can suck, but at least they spent on rallying a community complaining about how sucky they are exactly, Matrix has community circles for different things, on XMPP you look down the list and only see dev channels. ↺

And some channels which are branched out of that norm get removed from SJN as the operator disagrees with them.

if y'all want xmpp to be successful, I have a very wild suggestion. Perhaps write *very good* bridges to matrix, so when people eventually get tired of it they move to xmpp and not lose out on communities

i didn't know SJN also has censors, lol.

> if y'all want xmpp to be successful, I have a very wild suggestion. Perhaps write *very good* bridges to matrix, so when people eventually get tired of it they move to xmpp and not lose out on communities there is bridges already, nicoco has been working hard on slidge, but I think personally bridges are counter productive, they dont encourage people to move, it makes it easier for people to *not* move ↺

I go more on the basis of "if you want to talk to me, get XMPP"

> Then when it comes to channels, the standard is SJN, which is renown for only including channels the operator agrees with, I shall not name who the operator is, but this is a problem! then make another search engine ↺

> then make another search engine There has been talks about it :) ↺

> then make another search engine there are 15 standards ↺

\s

in fact i would probably help you, an alternative is badly needed

> I go more on the basis of "if you want to talk to me, get XMPP" i can see your point to an extent. But when all the communities are from matrix and you wanna capture them... yeah let's bridge some contents over here ↺

XMPP already has a method of channel discovery based on what channels your XMPP server is peered to (dont know the XEP for it)

the issue is, its inconvenient, so SJN was a more centralised approach which kicked off from it.

> XMPP already has a method of channel discovery based on what channels your XMPP server is peered to (dont know the XEP for it) if you mean the published directory of a single server, then i don't think it helps too much ↺

An alternative is to write a crawler, which crawls all public providers, or potentially hooking into Guus XMPP server graph to crawl anyone who publicly announces themself and their peers on the network, and then have a zero filtering policy for it.

> XMPP already has a method of channel discovery based on what channels your XMPP server is peered to (dont know the XEP for it) Disco ↺

This message has been retracted by the sender.

> An alternative is to write a crawler, which crawls all public providers, or potentially hooking into Guus XMPP server graph to crawl anyone who publicly announces themself and their peers on the network, and then have a zero filtering policy for it. this still partially centralises, but could be self hosted. ↺

> there are 15 standards Well, there's a difference between "standards" and "implementations" or "deployments". The core software running SJN is open-source ( https://codeberg.org/jssfr/muchopper ) and I think it's totally reasonable to have more than one search engine.

> > there are 15 standards > Well, there's a difference between "standards" and "implementations" or "deployments". The core software running SJN is open-source ( https://codeberg.org/jssfr/muchopper ) and I think it's totally reasonable to have more than one search engine. this is _some_ of the code ↺

as i said, an alternative is badly needed. a full alternative

the end product is proprietary..

Sure. If you're going to build a search engine, expect to put some work in.

and the operator moderates based on personal preference.

are chatwith.im and conversations.im admins in here?

as i said polarian i'm willing to work on this with you if need be

The moderation is not done by one person, so that's not quite true

> Sure. If you're going to build a search engine, expect to put some work in. But this still centralises the problem... but the decentralised solution right now is annoying, because you need to discover every peers channels manually, and if you aren't peered, you cant see their channels. ↺

isis members where using their servers

https://www.infostealers.com/article/killings-torturing-and-smuggling-how-an-infostealer-exposed-an-isis-cells-xmpp-network/

> as i said polarian i'm willing to work on this with you if need be I would look into the idea, but I dont think making more centralised solutions are the best idea, I think it would be an *improvement* but not the *solution* ✏ ↺

> as i said polarian i'm willing to work on this with you if need be I would look into the idea, but I dont think making more centralised solutions is the best idea, I think it would be an *improvement* but not the *solution* ✏ ↺

> https://www.infostealers.com/article/killings-torturing-and-smuggling-how-an-infostealer-exposed-an-isis-cells-xmpp-network/ well we love to see XMPP adoption right? ↺

> https://www.infostealers.com/article/killings-torturing-and-smuggling-how-an-infostealer-exposed-an-isis-cells-xmpp-network/ oh shit... this is going to cause some states to talk about banning XMPP ↺

and I am sure the UK will be one of them, as they seem to be the world leader in stupid legislation the past few years.

>> https://www.infostealers.com/article/killings-torturing-and-smuggling-how-an-infostealer-exposed-an-isis-cells-xmpp-network/ > oh shit... this is going to cause some states to talk about banning XMPP i mean terrorgram runs basically entirely on proprietary services ↺

it might be the whole banning matrix/element thing again

nah I think it will be more targeted at OMEMO

they can use warrants to yoink MAM

> Sure. If you're going to build a search engine, expect to put some work in. - and a modern federation tester: https://federationtester.mtrnord.blog/ - and an SFU tester: https://livekit.io/connection-test (Livekit's is matrix's SFU, so galene should have its own testing page too) - and a good bot ecosystem: https://matrix.org/ecosystem/integrations/ ✏ ↺

> - and a modern federation tester: https://federationtester.mtrnord.blog/ > - and an SFU tester: https://livekit.io/connection-test (Livekit's is matrix's SFU, so galene should have its own testing page too) > - and a good bot ecosystem: https://matrix.org/ecosystem/integrations/ We have some fragmented tools around ↺

you can do PGP on any platform

> and I am sure the UK will be one of them, as they seem to be the world leader in stupid legislation the past few years. the uk loves censorship. i dont think ive seen a place in recent history where elderly and disabled people are thrown on jail for holding a sign ↺

What's a SFU tester?

Martin, has done a lot of work on tooling for this stratself, would recommend you talk with him.

this isn't really 100% preventable

> We have some fragmented tools around i'm aware of the tools. And i don't wanna say this as discrediting anyone, but just as a point of comparison to say there's many things to be done ↺

> you can do PGP on any platform another encryption standard which has been discussed about being blocked in the past ↺

thats an age old discussion of "banning PGP"

never came to fruition thankfully

> Martin, has done a lot of work on tooling for this stratself, would recommend you talk with him. can you (or him) link to some of his works? would appreciate ↺

> thats an age old discussion of "banning PGP" that's impossible to do effectively ↺

> can you (or him) link to some of his works? would appreciate He hosts it on the Debian git thingy, but I can't remember what is is, sorry. Wait for him. ↺

i think you overestimate the digital literacy of MPs

> What's a SFU tester? a Selective Forwarding Unit allows more efficient group calls. And the linked webpage tests whether it's working well ↺

> that's impossible to do effectively Like 99% of the legislation passed by governments are enforcable, the recent findings by the EU commission on tiktok literally highlight the EU can fine any platform they feel like as all platforms violate the legislation. ↺

this

This message has been retracted by the sender.

> i think you overestimate the digital literacy of MPs this ↺

> i think you overestimate the digital literacy of MPs No I don't, I know they don't know what the fuck they are talking about. But it won't stop them suing you or throwing you in prison for bogus reasons. ↺

Good luck fighting a court case when the poorly written laws are impossible to abide by

true

i doubt this will be an issue anyway. normies will likely stick to proprietary services even if they enshittify further

Well this is why we need to try to convince more people to hop ship.

at least they hopped to Matrix :')

>> Martin, has done a lot of work on tooling for this stratself, would recommend you talk with him. > can you (or him) link to some of his works? would appreciate I didn't follow the conversation but I have a tool that checks srv records and connecting to the server if that's what you are talking about: https://xmpp-dns.mdosch.de

Martin, thanks

> at least they hopped to Matrix :') but that is counter productive to XMPP though. If people hop to Matrix and feel its "good enough", then why would anyone use XMPP? ↺

i do see you make some other xmpp bots too

> but that is counter productive to XMPP though. If people hop to Matrix and feel its "good enough", then why would anyone use XMPP? We need to maintain the stance that Matrix *IS NOT good enough*. However, arguing that to people who are not as techy as us is very difficult. Trying to explain that matrix is hard to self host, most people wont care, trying to explain that matrix is heavier or has a monolithic design, people dont care. As Travis said, people care about the end product, and if the end product is easier to use (which matrix unfortunately is), then its going to be adopted over XMPP. ↺

People jumping ship are more likely to jump once more. Outside of chatting apps you see this a lot with "distro hopping" for example

> People jumping ship are more likely to jump once more. Outside of chatting apps you see this a lot with "distro hopping" for example Techy people jump ship, average people don't. ↺

A lot of people I know will never delete Discord and are painful to even get to use XMPP just to talk to me, because everything they want is on Discord and even if Discord enshitifies, they will pay nitro, they will deal with the worse platform, just because everyone else uses it.

i actually started selfhosting matrix before running my own xmpp server. Running on both sides I can only wish both "platforms" prosper

i guess that might surprise you lol

No. Matrix is far easier to set up a server for lol. But not really relevant to this user adoptability conversation IMO

> i actually started selfhosting matrix before running my own xmpp server. Running on both sides I can only wish both "platforms" prosper same, running matrix was a lot more work and needed a lot more maintance and moderation. i stopped hosting it for the time being ↺

> No. Matrix is far easier to set up a server for lol. But not really relevant to this user adoptability conversation IMO Someone has a datacentre lying around obviously. ↺

and some really thick fibre to handle the metadata DDOS ✏

> We need to maintain the stance that Matrix *IS NOT good enough*. However, arguing that to people who are not as techy as us is very difficult. Trying to explain that matrix is hard to self host, most people wont care, trying to explain that matrix is heavier or has a monolithic design, people dont care. As Travis said, people care about the end product, and if the end product is easier to use (which matrix unfortunately is), then its going to be adopted over XMPP. but imo if the barrier of selfhosting xmpp down, you'd get more people running the network, and they can be for the families and friends group well snikket is already doing it ↺

>> No. Matrix is far easier to set up a server for lol. But not really relevant to this user adoptability conversation IMO > Someone has a datacentre lying around obviously. Set up and active maintenance are different. And while matrix succeeds at one, it fails at the other. Ups and downs ↺

i don't really think discoverability should be something to fight for on xmpp. I mean, not in a way that matrix does it with a centralized https://matrixrooms.info or matrix.org server where everything happens

Fight as in argue against, or fight as in promote?

>> but that is counter productive to XMPP though. If people hop to Matrix and feel its "good enough", then why would anyone use XMPP? > We need to maintain the stance that Matrix *IS NOT good enough*. However, arguing that to people who are not as techy as us is very difficult. Trying to explain that matrix is hard to self host, most people wont care, trying to explain that matrix is heavier or has a monolithic design, people dont care. As Travis said, people care about the end product, and if the end product is easier to use (which matrix unfortunately is), then its going to be adopted over XMPP. Again, this simply is false. Sure, there are people interested in products rather than protocols, but that is certainly the case for the other way around, too. _Especially_ people that are tasked with making decisions on how various groups of people communicate are very interested in learning about the protocol. Dismissing either group as non-existing is borderline silly.

> Fight as in argue against, or fight as in promote? "fight for", so promote ↺

Another issue, scummy OS.

https://dontkillmyapp.com/

only a few phones today will actually function. Otherwise its play-store only

Daniel is a life saver for this, and has days the app is free. I am sure if I asked at a conf I would advertise XMPP at he would make it free for a day or too, hell I would happily donate for it.

Monal IM is having a UX rewrite, beta should be about a month or so I was told???

Guus, remember how long it was at the dinner? The dev did say (sorry forgot his name already)

> A lot of people I know will never delete Discord and are painful to even get to use XMPP just to talk to me, because everything they want is on Discord and even if Discord enshitifies, they will pay nitro, they will deal with the worse platform, just because everyone else uses it. i mean, sure if you need 90% of the world on your protocol, then the normies need shipped products with cute little stickers and custom emojis and animated avatars. But sincerely we're not there yet, and I think you'd have a better chance getting there by polishing the protocol for interested "power users" and "decision makers" ↺

> Guus, remember how long it was at the dinner? The dev did say (sorry forgot his name already) I don't remember that conversation.

> >> but that is counter productive to XMPP though. If people hop to Matrix and feel its "good enough", then why would anyone use XMPP? > > We need to maintain the stance that Matrix *IS NOT good enough*. However, arguing that to people who are not as techy as us is very difficult. Trying to explain that matrix is hard to self host, most people wont care, trying to explain that matrix is heavier or has a monolithic design, people dont care. As Travis said, people care about the end product, and if the end product is easier to use (which matrix unfortunately is), then its going to be adopted over XMPP. > Again, this simply is false. Sure, there are people interested in products rather than protocols, but that is certainly the case for the other way around, too. _Especially_ people that are tasked with making decisions on how various groups of people communicate are very interested in learning about the protocol. Dismissing either group as non-existing is borderline silly. Guus, I have to disagree because people you talk to are those interested in tech, how do we sell XMPP to some random person who is tech illiterate? Those interested in protocol are a small minority of the potential userbase. ↺

> I'm having real-world conversations with policy makers that _know_ about Matrix, but have never heard of XMPP. They're even stating that there's no RFC-based IM protocol. _That_ is the problem that I think we need to solve: we need to be educating people about the brilliant stuff that we have to offer. Guus, do you have any hints why this obscurity is the case? ↺

> i actually started selfhosting matrix before running my own xmpp server. Running on both sides I can only wish both "platforms" prosper Same here ↺

I got told about XMPP on matrix

stratself: I believe that this is caused by Matrix doing a _much_ better job at evangelism / outreach / positioning than the XMPP-world is doing. This is why I asked this to be discussed during the last Summit, which is where https://mail.jabber.org/hyperkitty/list/members@xmpp.org/thread/47CJ5TN5SXIMGFBJEGMBTZETCOKK77Q3/ comes from.

icebound.dev: if you talk about _selling_ XMPP, then targeting the smaller group of more influential people (technologist, standardization people, architects) makes more sense than targeting individual end-users. I'm certainly not saying that this is an either/or thing, but dismissing one group is really, really unwise in my opinion.

> stratself: I believe that this is caused by Matrix doing a _much_ better job at evangelism / outreach / positioning than the XMPP-world is doing. This is why I asked this to be discussed during the last Summit, which is where https://mail.jabber.org/hyperkitty/list/members@xmpp.org/thread/47CJ5TN5SXIMGFBJEGMBTZETCOKK77Q3/ comes from. starting to think I should have went to the summit ↺

I thought it was just dev stuff, but it sounds like operator/outreach would be able to fit in too :p

> icebound.dev: if you talk about _selling_ XMPP, then targeting the smaller group of more influential people (technologist, standardization people, architects) makes more sense than targeting individual end-users. I'm certainly not saying that this is an either/or thing, but dismissing one group is really, really unwise in my opinion. I dont want to dismiss one group, I feel we preach too much to one group, and we need to outreach more. ↺

I believe I kicked off this discussion by stating that we're currently not servicing one group at all - those technologists / standardization people / architects (et.al.) hardly know we _exist_, in my experience.

> I believe I kicked off this discussion by stating that we're currently not servicing one group at all - those technologists / standardization people / architects (et.al.) hardly know we _exist_, in my experience. Hmm, I assumed they did... ↺

my assumption would have been wrong then.

I've _literally_ seen official national standardization documentation that claims no RFCs exist for chat and IM. - that scared me _a lot_. One could argue that the people involved in writing that are at fault, but you could equally argue that we should've done a better job promoting ourselves.

As the XSF, we _can_ do the latter (we can't do the former).

Hence the call to action.

welp im a small fish in a very *VERY* large ocean :p

the XSF is a bigger fish :p

> welp im a small fish in a very *VERY* large ocean :p will still try to push XMPP this year at the confs I attend. ↺

maybe do an XMPP talk :p

That's the kind of thing that I believe we should do more. Make people interested again. We've long ago lost the hype-status (which Matrix is currently riding out). I feel that we've grown complacent, assuming that everyone knows what XMPP is. Instead, we're running the risk of being forgotten.

> The mission of the XMPP Standards Foundation (XSF) is to build an open, secure, feature-rich, decentralized infrastructure for real-time communication and collaboration over the Internet. > The XSF’s product is protocols; the XSF’s market is developers > But we do not market ready-made solutions to organizations or individuals: that is the job of the developers we serve. From the XSF mission statement page

> That's the kind of thing that I believe we should do more. Make people interested again. We've long ago lost the hype-status (which Matrix is currently riding out). I feel that we've grown complacent, assuming that everyone knows what XMPP is. Instead, we're running the risk of being forgotten. well from the sounds of it XMPP *is* forgotten :/ ↺

maybe thats a good thing, starting from scratch, rebuilding XMPP in a modern image?

if i'm being honest, the xsf is mainly a technical institute. So if you really need popularity either change its mission statement or have sibling institutes doing the grunt marketing

I think a proper clone of Discord functionality (as far as the concept of voice channels and whatnot) would be a huge stride

> maybe thats a good thing, starting from scratch, rebuilding XMPP in a modern image? not this old protocol from the 1980s/90s but this modern and modular protocol ↺

I suggest Mumble to people and they just say "isn't that only for voice?"

> I think a proper clone of Discord functionality (as far as the concept of voice channels and whatnot) would be a huge stride This is being worked on :) ↺

Theres a new "spaces" XEP

which will then enable Discord-server like functionality

> if i'm being honest, the xsf is mainly a technical institute. So if you really need popularity either change its mission statement or have sibling institutes doing the grunt marketing you can say JoinJabber is doing its job but... you need more coherence in who you're marketing to because resources are scarce ↺

but writing a lightweight GUI client is really long and hard

the server backend for this stuff, dare I say it, is the easy part

> which will then enable Discord-server like functionality will this be supported for all clients tho?

actually designing the UX and making the client efficient, secure and maintainable is hard...

A xep isn't needed to enable any functionality and is like 0.1% of any work

> > which will then enable Discord-server like functionality > will this be supported for all clients tho? well its a XEP, so clients decide if they want to support it ↺

personally I wouldnt use it

XMPP is technically _fine_. Any one new feature is not going to make any difference. It's not about the status of specific parts of functionality. Instead, we have a protocol that has proven itself over and over again that it can adapt to changing demands. _That_ is what we should push.

ye if i wanted discord functionality i would just use discord...

Sure, but when I mention XMPP I get a reaction a lot like when I say I use IRC, "isnt that old and... legacy?"

i dont think people really know what xmpp is

> XMPP is technically _fine_. Any one new feature is not going to make any difference. It's not about the status of specific parts of functionality. Instead, we have a protocol that has proven itself over and over again that it can adapt to changing demands. _That_ is what we should push. then you need to push the notion that this protocol actually changes and has a robust community/foundation process doing so, and is not prone to standards hell ↺

Who is "we" ? Companies with marketing budgets are the only ones who "push" anything

at least the people i know

> Who is "we" ? Companies with marketing budgets are the only ones who "push" anything true

> > which will then enable Discord-server like functionality > will this be supported for all clients tho? Voice channels could be joinable as calls from other clients ↺

some clients dont support calls tho

profanity...

the issue with calls it that theres two ways of doing it

i think at least

> the issue with calls it that theres two ways of doing it really?

there's infinite ways of doing anything

what i take from "voice channels" is that the future XEP for Muji + SFU should only serve as primitives for building a calling experience

I dont know the protocol specifics, but there is jitsi's way which is bounce it off the server (iirc) and then there is p2p calls which dino implements, but when you scale this... you get tons of tcp connections between clients and the system grinds to a halt. ✏

"we", as in the community. Anyone interested in furthering the adoption of XMPP. Those can be individuals, organisations, companies. All of those are made up of people, that need to decide individually if this is desirable. If we're just pointing at others and go 'you should do this', nothing will happen.

would rather use mumble

> I dont know the protocol specifics, but there is jitsi's way which is bounce it off the server (iirc) and then there is p2p calls which dino implements, but when you scale this... you get tons of tcp connections between clients and the system grinds to a halt. p2p vs sfu: https://getstream.io/blog/what-is-a-selective-forwarding-unit-in-webrtc/ xmpp doesnt have sfu yet, but someone's working on implementing it with galene.org ↺

but lets not talk about specifics for a moment lol

> Then when it comes to channels, the standard is SJN, which is renown for only including channels the operator agrees with, I shall not name who the operator is, but this is a problem! This is completely false. Their bot can be freely invited and lists all channels it can discover. Only in rare cases and after repeated abuse reports are channels manually delisted. ↺

> This is completely false. Their bot can be freely invited and lists all channels it can discover. Only in rare cases and after repeated abuse reports are channels manually delisted. yeah yeah of course you would deny this, just like you saying JJ is inclusive. ↺

If JJ is so inclusive then why am I banned from it for simply joining the channel?

yet another JJ argument started by polarian in an unrelated room

Let's not go down this route. Let's keep to conversation open and inviting please.

> Let's not go down this route. Let's keep to conversation open and inviting please. Sure, but I would rather my concerns not be called "false" by someone who causes my concern. ↺

> yet another JJ argument started by polarian in an unrelated room Apologies ↺

> "we", as in the community. Anyone interested in furthering the adoption of XMPP. Those can be individuals, organisations, companies. All of those are made up of people, that need to decide individually if this is desirable. If we're just pointing at others and go 'you should do this', nothing will happen. what community? the rest of what you said is the important bit, there is no one community with a set of goals, if you have a goal in mind then find or create your own community to further it ↺

> This is completely false. Their bot can be freely invited and lists all channels it can discover. Only in rare cases and after repeated abuse reports are channels manually delisted. I noticed an economics channel was delisted ↺

So obviously economics to SJN is a banned subject

cmon guys, lets not go there.

moparisthebest: I'm active in various communities, and I'm trying to do my best to utilize each one: XSF, IgniteRealtime, GoodBytes, guus-the-lonely-dev.

Well my point was we need better ways of discovering channels, instead of a single centralised webapp for it. I didn't want to "go there", but if someone wants to discredit my concerns, and suggestions, then I have to defend my concerns.

I can do different things in different capacities.

(and I try to)

right, but everyone here keeps saying "we" like that's a single thing we all agree on

> Well my point was we need better ways of discovering channels, instead of a single centralised webapp for it. I didn't want to "go there", but if someone wants to discredit my concerns, and suggestions, then I have to defend my concerns. word of mouth is a strong thing ↺

sunglocto (sunglocto.net), Economics and Liberalism, which discusses mainly the Austrian school of economics was delisted, for reasons beyond me

> Well my point was we need better ways of discovering channels, instead of a single centralised webapp for it. I didn't want to "go there", but if someone wants to discredit my concerns, and suggestions, then I have to defend my concerns. I just do not see this as relevant to adoption of XMPP. I think relatively few XMPP users are actually even joining public channels, and even fewer are just looking for channels rather than like joining a link from a website or other community. ↺

No, I'm not pointing to one 'we'. I'm trying to use that to point out that everyone reading this and somehow is interested in XMPP is someone that I'm interested in taking part in this.

If we all use SJN, and it is so easy to delist a channel by the authority of one or a few people, then how do we convince people XMPP is a decentralised protocol resistant to censorship?

so yes I think my concerns are valid.

and yes sunglocto (sunglocto.net), I would consider working with you on an improvement

icebound.dev: create a second SJN - afaik the protocols are published. I'm not sure if the implementation is open source, but I wouldn't be surprised.

i mean there are other reasons to not use SJN rather than affiliations that jonas may or may not have. i'm mainly concerned with centralisation

> I just do not see this as relevant to adoption of XMPP. I think relatively few XMPP users are actually even joining public channels, and even fewer are just looking for channels rather than like joining a link from a website or other community. Not saying it's not a concern at all, just not to relevant to this issue ↺

> i mean there are other reasons to not use SJN rather than affiliations that jonas may or may not have. i'm mainly concerned with centralisation I share that concern. ↺

there should be support for custom "search services" in xmpp clients - conversations for example is hard-coded to either use the local muc server to search rooms, or SJN

As of now there are not any alternatives to SJN so that probably wouldn't be worth the effort

> I just do not see this as relevant to adoption of XMPP. I think relatively few XMPP users are actually even joining public channels, and even fewer are just looking for channels rather than like joining a link from a website or other community. Hmm this is true. But Discord for example has a discovery mechanism for channels. Right now if I want to discover a channel its SJN or word of mouth. ↺

I think if we saw a legitimate group of alternatives it would be worth

> there should be support for custom "search services" in xmpp clients - conversations for example is hard-coded to either use the local muc server to search rooms, or SJN ok cool, then start an alternative then bring it up to client devs? ↺

> But Discord for example has a discovery mechanism for channels I was a heavy Discord user for ~5 years and never once used this

I'm 99.999% sure that SJN wasn't offered in any client up until SJN started to exist.

Just saying "there oughta be" is useless

exactly.

case in point. if we make an alternative then it would probably begin to get used

iirc the MUC XEP specifies advertising the channels a XMPP server provides. So, its feasible that an XMPP server could keep track of all the MUCs its peers advertise, and then share this with clients (so clients dont need to recursively resolve them). Furthermore, it could be that all provides from the provider list by default are indexed by the server.

there is definitely an issue with people whining and then doing nothing to fix the issue

> Just saying "there oughta be" is useless I have suggested a few ideas... ↺

> there is definitely an issue with people whining and then doing nothing to fix the issue _I usually agree to fix the issue, and then never do it :p_ ↺

🧌

micycle.net, I find it interesting you are not concerned that one of your channels were delisted from SJN without so much of a reason for why?

well guys, I think you all started with discussing how to better xmpp for normies. Now it becomes a different goalpost by saying you need a better non-dictatorial room directory or whatever

i mean i doubt the economics muc was delisted intentionally. if it was in fact delisted intentionally then all of us know why

cronies will be cronies. that's unrelated to this discussion in my opinion however and there are other reasons to try and diversify our services

> i mean i doubt the economics muc was delisted intentionally. if it was in fact delisted intentionally then all of us know why You pointing fingers 🤣 ↺

sunglocto (sunglocto.net), thoughts on my suggestion?

it could be feasible to have a prosody plugin which indexes providers, and its peers public MUCs and then advertises this to the clients connected to it.

it would probably be more feasible to have this be a client-side feature

sunglocto (sunglocto.net), the issue with this is, if you got a big provider list, you would slam the client

just stop spouting ideas and build it or do nothing ;)

> icebound.dev: create a second SJN - afaik the protocols are published. I'm not sure if the implementation is open source, but I wouldn't be surprised. It is open source and the link was shared here earlier. ↺

imagine a phone on a public wifi network, connecting to potentially 1000+ providers, and then indexing all of its MUCs

so allow the user to choose a provider

true, but then what if they want to search _all_ providers

I think having the server handle this and allow you to query the server would be more efficient

icebound.dev, not a bad idea for a "plugin". It has to fetch data from somewhere (SJN for example) and scrape the room directories of everything it touches like midas but please ship it tho.

i think that's a tad bit superfluous. there shouldn't be significant differences between indexes unless we have such a fucking insane motley crew hellbent on controlling information

matrix does this by just having a couple of public servers that index rooms and you can pick which index to choose

> micycle.net, I find it interesting you are not concerned that one of your channels were delisted from SJN without so much of a reason for why? Didn't say I'm not concerned, I just said it's not relevant to the overall adoption of XMPP ↺

> icebound.dev, not a bad idea for a "plugin". It has to fetch data from somewhere (SJN for example) and scrape the room directories of everything it touches like midas > > but please ship it tho. It would come from the providers, which is a public list, but for the most part would be apolitical, the indexed results would be decentralised, just it would rely on the providers list, which could be mirrored or distributed. But my other suggestion was for it to be also the peers of the server, so my server has maybe 15 peers? It could recurse through them and index its MUCs. SO someone like conversations.im could provide a massive number of channels to search through. ↺

and it would be completely decentralised in the case of indexing peers.

btw Guus, if you're still there, wdyt of turning the XSF into an advocacy group? from what I understand it's not initially designed to be one, so idk if there are any ideas or frameworks guiding any of your advocacies of this protocol

i think that's a very heavy solution

> Didn't say I'm not concerned, I just said it's not relevant to the overall adoption of XMPP how so? Would you adopt a protocol which is decentralised up until the point where one/a few people control the main way to find new channels? ↺

(wdyt = what do you think)

> i think that's a very heavy solution Not so much, a daily or so index would be fine no? ↺

also i don't think SJN controls the way that people find channels. i think that the only person who has tried to do so made one and people use it

> how so? Would you adopt a protocol which is decentralised up until the point where one/a few people control the main way to find new channels? Again a very small minority of users are actually just looking for channels like that ↺

> Again a very small minority of users are actually just looking for channels like that true ↺

there are issues with SJN other than its creator however which you for whatever reason seem to be caught up on

I do see your point that people join channels from other places, say a website promoting a channel, or in the case of Discord, a lot of its channels were advertised through youtubers...

so its entirely feasible to "outsource" the finding of channels

i mean almost all of the rooms im in i joined via word of mouth

> btw Guus, if you're still there, wdyt of turning the XSF into an advocacy group? from what I understand it's not initially designed to be one, so idk if there are any ideas or frameworks guiding any of your advocacies of this protocol stratself: ask in xsf@ ↺

Right any and all Discord Guilds I ever joined were links from a community's website or just links from friends

anyone who's seen me say "JID?" knows exactly what i mean

> Right any and all Discord Guilds I ever joined were links from a community's website or just links from friends Sure. ↺

But SJN exists, why not improve on the idea and make it decentralised

> i mean almost all of the rooms im in i joined via word of mouth What if we crawl MAM history of MUCs so it firstly lists the most talked ones ↺

> But SJN exists, why not improve on the idea and make it decentralised at worst it doesn't help, at best it does? ↺

There's already nothing really centralized about it

> What if we crawl MAM history of MUCs so it firstly lists the most talked ones my idea was already branded "heavy" by sunglocto, this would make XMPP as slow as Matrix 🤣 ↺

> There's already nothing really centralized about it a crawler controled by one/a few people, who can manually delist channels they disagree with? ↺

If thats not centralised, idk what is :p

yeah that is extremely heavy. i think at the moment member count is fine, i think the lighthouse and joinjabber are the few examples where they have huge membercounts but very low traffic

> a crawler controled by one/a few people, who can manually delist channels they disagree with? An XMPP user can easily go their entire lives without ever interacting with SJN ↺

> An XMPP user can easily go their entire lives without ever interacting with SJN Sure, but your argument doesn't argue that SJN is redundant, but just that its not manadatory, I would still like an improvement from SJN for personal preference. ↺

I like discovering channels, I found the BSD channels through SJN

As several people mentioned, run your own instance then

i mean you could start making an alternative. right now

I dont like being fed a list of MUCs which have been fed through someone elses biases.

> As several people mentioned, run your own instance then this doesn't solve the issue ↺

> i mean you could start making an alternative. right now . ↺

this just adds a second player

Thus solving the issue..

then someone might not like my biases, and thus another one pops up

and so on and so on

> I dont like being fed a list of MUCs which have been fed through someone elses biases. then stop whining and fix it? ↺

or just the first

until you have as many indexes as you do XMPP servers, and you would just discover the channels through the protocol

> then someone might not like my biases, and thus another one pops up that's the beautiful thing about platforms like xmpp ↺

> then someone might not like my biases, and thus another one pops up Sounds like that would work as intended ↺

> then stop whining and fix it? I suggested how to fix it, sunglocto (sunglocto.net) is concerned its too heavy. ↺

if you don't like how someone's doing it then you DIY

You're literally complaining about centralization and then saying decentralization doesn't solve the issue

> if you don't like how someone's doing it then you DIY mate, I am not denying this ↺

icebound.dev: you cannot curate an index without having bias

stop suggesting, just fix

i literally said I would be happy to look into solutiosn

> stop suggesting, just fix the best ideas are those which are discussed and scrutinised ↺

i mean i don't really care about the centralised manner of it. we are using XMPP MUC which is by definition centralised

and if you don't curate it, you end up with spam or worse things

*by function

it gets better by hosting alternate instances

Exactly an index that refuses to discriminate becomes useless

sunglocto (sunglocto.net), well this requires writing it, SJN has some of the code, not all.

Funny to call MUC "centralized", but I guess I get what you mean

then write it

> Funny to call MUC "centralized", but I guess I get what you mean wasnt there a XEP to federate MUCs ↺

write the code lil bro

fMUCs I think they are called

Yes

This is like a Google vs Duckduckgo vs SearxNG vs YaCy discussion rn

> Yes crazy idea, sounds unable though :p ↺

unstable*

The only implementation I'm aware of is in an XMPP server almost exclusively used in military setups, so... maybe

i mean i won't be able to start making an indexer now because i'm working on my own project, but realistically this shouldn't be difficult to do

> The only implementation I'm aware of is in an XMPP server almost exclusively used in military setups, so... maybe Openfire has an implementation iirc ↺

as i said before, an alternative is desperately needed

kinda remains me of IRC networks tbh

> write the code lil bro maybe if I find time to :p ↺

currently got 101 operator duties to do, still discussing a UK public provider

my query in the providers MUC went ignored :/ ✏

discuss less, code more

fucking this

sunglocto (sunglocto.net), I did offer to include you in on the aim to get a UK public provider

never actually responded to the offer

for reasons that are way too offtopic for this chat. you know what they are

stratself: I'd not want to _change_ the XSF to an advocacy group, but I do believe that there's plenty that the XSF can already do well within their current mission statement. Like I already said: we discussed this in the last Summit. There's a discussion ongoing on one of its mailing lists, too.

> for reasons that are way too offtopic for this chat. you know what they are genuinely don't will ask in PMs ↺

if you want something done, do it yourself

stratself, there is SCAM

> if you want something done, do it yourself omfg I know, I am... ↺

Its also useful to discuss ideas...

> Its also useful to discuss ideas... go to jdev ✏ ↺

> go to jdev, ffs true... ↺

not forever with no action in the wrong place :P

i could literally get a prototype together tonight

> i could literally get a prototype together tonight glad you have time ↺

you need a crawler component and a frontend

I still have work to finish and deliver to a customer, I have a server which is half completed, and I haven't coded in a while.

> i could literally get a prototype together tonight also I have read your code, it very much is "write now, optimise later" ↺

but I guess thats whaat a prototype is

I am a perfectionist so my projects always end up being bogged down with that.

>> i could literally get a prototype together tonight > also I have read your code, it very much is "write now, optimise later" hence prototype ↺

> stratself: I'd not want to _change_ the XSF to an advocacy group, but I do believe that there's plenty that the XSF can already do well within their current mission statement. Like I already said: we discussed this in the last Summit. There's a discussion ongoing on one of its mailing lists, too. thanks for your response. I'll just follow the maillist and pounce any ideas over to @xsf as need be ↺

sunglocto (sunglocto.net), please answer my question in PMs

i'm good

> i'm good I kinda need the answer to it lol ↺

thanks stratself

Were there any operators at the XSF summit who were *not* also developers?

Boy, operator chat is spicy today.

Sorry :(

icebound.dev: the list of attendees is on the wiki :)

> icebound.dev: the list of attendees is on the wiki :) yeah but I don't know them all :( ↺

MattJ: what are you talking about that is only implemented in military setups?

https://www.army-technology.com/products/m-link-for-constrained-networks/

https://xmpp.org/extensions/xep-0289.html

Does anybody have a working implementation of XMPP over HF?

Isode is the only one I'm aware of with M-Link but they refuse to sell it to me because i'm not a military

Or even XMPP over AX.25

FX25

I2LP

There is a xmpp to meshtastic bridge on codeberg

Seems STANAG5066 might actually just be vaporware

Meshtastic is a toy protocol that fails under real world usage

I'm talking about running the XMPP protocol itself over a non IP or non-TCP protocol. Just something that provides ARQ or ideally that 'performance enhancing proxy' that allows for stanzas to be preempted out of order and whatnot.

IE, internet goes down so I federate with important servers over ARDOP2000 on a narrowband HF link.

Or I2LP packet

I've been doing some testing with running XMPP+TLS+TCP+IPv4+D* at a 900bps symbol rate and it's really way too much overhead

https://eindhoven.space/wp-content/uploads/2023/05/il2p-specification-draft-v0-5.pdf

> Interesting discussions about Matrix, and XMPP, in the thread I've seen a lot of 'Wait, Matrix has been around for years and it is still all slow and has issues' recently and a lot of praise for XMPP and people saying they want to get back into it. That and some people looking for alternatives due to the Digital Independence Day. Make of that what you will. I, for one, see that as a first sign of a renaissance of XMPP.

so no more having the same people on every muc?

> > Interesting discussions about Matrix, and XMPP, in the thread > I've seen a lot of 'Wait, Matrix has been around for years and it is still all slow and has issues' recently and a lot of praise for XMPP and people saying they want to get back into it. > > That and some people looking for alternatives due to the Digital Independence Day. > > Make of that what you will. > > I, for one, see that as a first sign of a renaissance of XMPP. a lot of the big cornerstones of matrix are falling, such as envs.net recently shutting down their matrix and now having an xmpp ↺

> silo users (like hn commentors) exclusively have bad opinions though Most, but I occasionally see some really intelligent points there, which always makes me wonder why they do post it there.

>> I've seen a lot of 'Wait, Matrix has been around for years and it is still all slow and has issues' recently and a lot of praise for XMPP and people saying they want to get back into it. >> >> That and some people looking for alternatives due to the Digital Independence Day. >> >> Make of that what you will. >> >> I, for one, see that as a first sign of a renaissance of XMPP. > a lot of the big cornerstones of matrix are falling, such as envs.net recently shutting down their matrix and now having an xmpp Didn't even notice. Did they post aboht their reasoning? Would love to read it.

Ill see if they posted much of anything, but as far as i know matrix was decided to be shut down due to ongoing cost (the servers are REALLY heavy to run) and then the admin just kinda discovered xmpp at some point

https://downloadable.pain.agency/file_share/019c4460-de4e-73c5-ab99-4d36632010f1/Screenshot_20260209_114825.png

https://bsky.app/profile/arathorn.net/post/3ly7fxuiw5s2a https://bsky.app/profile/arathorn.net/post/3ly7gaza45s2a https://bsky.app/profile/arathorn.net/post/3ly7g5s2yg22a

> Ill see if they posted much of anything, but as far as i know matrix was decided to be shut down due to ongoing cost (the servers are REALLY heavy to run) and then the admin just kinda discovered xmpp at some point Yeah, Matrix is the blockchain of messaging while XMPP is just coins and notes. Simple to carry, easy to maintain and not a burden on the environment.

ah yeah > important change at envs.net > > As many of you know, the Matrix service consumes a lot of resources. This has become increasingly problematic in recent months, as it’s making operation more and more expensive. > > For this reason, I’ve decided to discontinue the Matrix service. > [...] https://pleroma.envs.net/notice/B2DCyCSaONBuXeJFIW

> Yeah, Matrix is the blockchain of messaging what does that mean?

> > Yeah, Matrix is the blockchain of messaging > what does that mean? probably relating to the way room state DAG is calculated, it feels very web3 even though its kinda unrelated ↺

i never used matrix so ig i dont know

Regarding the screenshot: The simple fact that I at first didn't know whether I was looking at Matrix, XMPP or something else (until I looked at the UI elements) says a lot about how much worse one or the other is in turns of UI (although not UX).

>> what does that mean? > probably relating to the way room state DAG is calculated, it feels very web3 even though its kinda unrelated This

> i never used matrix so ig i dont know It wastes a lot of resources on accomplishing something that can be done easier. Just like Butcoin destroys the planet just so users can send some money... instead of just using a bank to transfer.

> Regarding the screenshot: The simple fact that I at first didn't know whether I was looking at Matrix, XMPP or something else (until I looked at the UI elements) says a lot about how much worse one or the other is in turns of UI (although not UX). gtk4 is quite clean feeling which makes gajim look very pretty, but a lot of the widgets seem used incorrectly... also the whole slow as shit single threaded python is kinda a bummer, gajim and random electron slop are the only programs that feel sluggish on my otherwise high end machine ↺

>> Regarding the screenshot: The simple fact that I at first didn't know whether I was looking at Matrix, XMPP or something else (until I looked at the UI elements) says a lot about how much worse one or the other is in turns of UI (although not UX). > gtk4 is quite clean feeling which makes gajim look very pretty, but a lot of the widgets seem used incorrectly... also the whole slow as shit single threaded python is kinda a bummer, gajim and random electron slop are the only programs that feel sluggish on my otherwise high end machine Yeah, growing pains. It's a new UI, it will get better. Just report your pain, perhaps the devs have not yet considered some of them as they might be deep im the code, but don't know how users perceive the UI, as often is the case. I've started to report every single bug a while ago and I can say that improved Linux telephones a lot and even some XMPP client things. Awareness of issues is the first step in solving them.

> I've started to report every single bug a while ago and I can say that improved Linux telephones a lot and even some XMPP client things. me and lovetox just end up arguing every time i try to report something, i honestly dont care at this point. plan to try my own hand at some point

>> what does that mean? > probably relating to the way room state DAG is calculated, it feels very web3 even though its kinda unrelated It feels kinda bad. I know the Matrix people are really nice people, but it isn't great seeing them trying for years and still struggling to have notifications that go away once you open the chat and messages that cannot be decrypted. Meanwhile, XMPP has solved many issues and they just don't look at prior art. :/

> Meanwhile, XMPP has solved many issues and they just don't look at prior art. :/ in fact they actively propagandize about the state of xmpp lol

> https://bsky.app/profile/arathorn.net/post/3ly7fxuiw5s2a > https://bsky.app/profile/arathorn.net/post/3ly7gaza45s2a > https://bsky.app/profile/arathorn.net/post/3ly7g5s2yg22a these are worth a read and laugh ↺

>> Couldn't think of anything worse than protocols becoming more and more adopted by government > in Germany, Matrix has a very dominant usage among government and public institutions. openDesk, which lately has gained a lot of attention and traction, has Matrix/Element as its chat solution and public health insurances use it (https://element.io/de/matrix-in-germany/projects/ti-messenger), as well as others (https://bundesmessenger.de/) > I wish XMPP would get as much as dedicated public funding... I wonder whether they use E2EE, I guess they need to keep logs for audits and legal reasons. Perhaps they don't even get "Unable to decrypt messages."

> ``` > Your XMPP server wants to connect to jabber.fr, do you accept to share your personal data with it > [Accept all the XMPP cookies] [Reject] [Customize] > ``` Please, I don't need nightmares! :p

>> So even worse then, I can't tell if you are supporting this or opposing this. > turning off federation is bonkers in my eyes, but so is using Matrix for the purposes they outlined 🤷‍♂️ Makes sense to turn it off, I don't want the German army to accidentally federate with matrix.putin.su :p

> I'm having real-world conversations with policy makers that _know_ about Matrix, but have never heard of XMPP. They're even stating that there's no RFC-based IM protocol. _That_ is the problem that I think we need to solve: we need to be educating people about the brilliant stuff that we have to offer. Some people even believe Matrix to be the internet standard. Many people have forgotten XMPP as well. But I'm hopeful the tide has started to turn, now we have to put the XMPP ship to use the fidal flow to bring standard messaging back into the harbour of communication. (I'm sorry, I know the metaphor has.. sailed too far.)

>> I agree. I also agree that hardly anyone _knows_ about this. > I would say more confs, push it at more events, Matrix has been very successful in getting Linux groups to use it. Start with Linux phone groups, XMPP works incredibly well on slow phones that are not yet at full performance (ask me how I know...)

>> I'm having real-world conversations with policy makers that _know_ about Matrix, but have never heard of XMPP. They're even stating that there's no RFC-based IM protocol. _That_ is the problem that I think we need to solve: we need to be educating people about the brilliant stuff that we have to offer. > > Policy makers never have and never will care about protocols, they care about products. eg ejabberd, openfire, element And they care about policy. Throw into the mix: interoperability, DMA, sovereignty, independence

> BSD has a small community, theres a few channels like the open hardware channel, but for the most part theres not that diversity I could get a lot more people interested if I just knew more interesting channels to recommend 🤔

>> In the blog that quote recalls what happened in 2015. IRC and for that matter XMPP were very different than what we have today. And the blog goes on to talk about switching away from discord. It's a fun read if you're interested in the topic > I dont have the time to read it but my experience with getting rid of Discord involved relapsing 7 times because XMPP is like shutting yourself in a lead box, unable to speak to anyone on the outside. XMPP isn't the lead box, gatekeepers like Discord are. They refuse to let you talk to the outside world. Like the phantom of the opera they drag you into their basement and tell you that from now on you can never leave. The only real difference to the phantom is, that by the end the phantom becomes someone you can sympathise with.

> my experience with discord and matrix is I've still never used either, seems useless My two hackerspaces use Matrix, but once Hackint and Biboumi support IRCv3 features, this might change. It's just annoying not to see pictures and display names, makes it hard to follow a chat where everyone else assumes all can see it.

> Why? I dont know One has been there early, then largely got forgotten, then new users discovered the other. \*shrug\*

the funny thing is XMPP is really commonly used in (well, was) business telephony - Cisco IM&P/Jabber is XMPP as the name suggests, a couple avaya products for both aura and IP office, etc I think the people pushing for matrix in business have probably been using XMPP this whole time without realising it (cisco and avaya are popular in governments and military because they support MLPP)

avaya workplace and cisco jabber can even federate lol

talking as a lowly user/single user server operator though, XMPP was so, so much easier than setting up synapse and uses like no resources compared to it - was a breath of fresh air and like many i started out with matrix as its what i knew of as it is very heavily pushed where as XMPP doesnt have all the fan fare around it

> again discussed at FOSDEM, XMPP has the issue of: > > - What client should I use? > - What provider should I use? > - What channels should I join? - What phone should I buy? - What provider should I use? - What plan should I choose? - What people should I call? I wonder what is similarly complicated that most people use :p

> provider has only 1 sane choice: yourself I imagine telling that my parents or grandmother... Yeah, they would just dismiss it with 'Yeah, whatever, you hungry?'.

> > provider has only 1 sane choice: yourself > > I imagine telling that my parents or grandmother... > > Yeah, they would just dismiss it with 'Yeah, whatever, you hungry?'. haha my mum has XMPP but doesnt know it (she is on my CUCM instance, i really need to federate IM&P to this instance but certs are a pain) ↺

> if y'all want xmpp to be successful, I have a very wild suggestion. Perhaps write *very good* bridges to matrix, so when people eventually get tired of it they move to xmpp and not lose out on communities THIS. This is what I have been repeating continiously for the last year. It would help sooo much.

>> if y'all want xmpp to be successful, I have a very wild suggestion. Perhaps write *very good* bridges to matrix, so when people eventually get tired of it they move to xmpp and not lose out on communities > there is bridges already, nicoco has been working hard on slidge, but I think personally bridges are counter productive, they dont encourage people to move, it makes it easier for people to *not* move We need server level, XMPP transports. What ejabberd does.

I'm glad I can keep in touch with my two hackerspaces until they get annoyed with Matrix and switch to XMPP in a couple years.

>> provider has only 1 sane choice: yourself > > I imagine telling that my parents or grandmother... > > Yeah, they would just dismiss it with 'Yeah, whatever, you hungry?'. erebion: why not? don't they install apps all the time? ↺

> No. Matrix is far easier to set up a server for lol. But not really relevant to this user adoptability conversation IMO I disagree, prosody and ejabberd and Snikket have all been easier than Synapse, or even Tuwunel [which is a lot easier than Synapse]. More over, Matrix servers are a lot more annoying to maintain.

and use more resources

Yup, my experience also having run both.

Yeah not really sure what that's based on because it took me way longer to get synapse set up than prosody.

> Theres a new "spaces" XEP Which one?

I've officially given up on reading the whole backlog after ¾ of it took an hour. lol

> talking as a lowly user/single user server operator though, XMPP was so, so much easier than setting up synapse and uses like no resources compared to it - was a breath of fresh air and like many i started out with matrix as its what i knew of as it is very heavily pushed where as XMPP doesnt have all the fan fare around it I already knew XMPP, but gave up on it a couple years ago when my last contact stopped using it and I literally had an empty roster and no one to talk to. Then a while later I suddenly had people to talk to again, so I had a server again. I guess there are many people out there that falsely believe XMPP to be dead.

>> >> I imagine telling that my parents or grandmother... >> >> Yeah, they would just dismiss it with 'Yeah, whatever, you hungry?'. > haha my mum has XMPP but doesnt know it (she is on my CUCM instance, i really need to federate IM&P to this instance but certs are a pain) Same, I set up a Snikket instance for friends and family, most have no idea what a protocol is, but it works well.

Biggest factor for me was notifications in Android apps work, unlike matrix which is till hit and miss even with battery controls etc turned off. This matters probably the most for my family. Never underestimate the wife factor for adoption of a thing.

Oh, just realised this probably isn't really the channel for this discussion heh.