Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Dino: An open-source application for XMPP messaging

dino.im

Post reply on HN
Print view View original

131–140 of 153 posts

Re: Dino: An open-source application for XMPP messaging

#131

Post by mekster »

kitd wrote: post #108
mekster wrote: post #81

It's sad that we don't have a very good self hosted chat solution. Last time I checked around, XMPP didn't have a client that I could recommend to business clients that looked easy/simple to use, has no gotchas and looks native. Currently I use Mattermost, which does basic things fine and things are stable but the formatting is just getting in the way making texts unnecessary big and bold when people never intend to…

Someone recently pointed me at DeltaChat [1]. 'Chat' over SMTP/IMAP. Apart from voice/video, it ticks all of the above (plus other features such as threading and e2e security), with the fallback of being able to use a regular email client in an emergency [1] - https://delta.chat/en/

I wonder if it can handle large attachments efficiently, assuming postfix is set to allow 1GB attachment or so.

Re: Dino: An open-source application for XMPP messaging

#132

Post by mekster »

lucideer wrote: post #102
mekster wrote: post #81

It's sad that we don't have a very good self hosted chat solution. Last time I checked around, XMPP didn't have a client that I could recommend to business clients that looked easy/simple to use, has no gotchas and looks native. Currently I use Mattermost, which does basic things fine and things are stable but the formatting is just getting in the way making texts unnecessary big and bold when people never intend to…

> WYSIWYG formatting instead of automatically parsing symbols into bold texts etc Assuming you know the minimum features non-technical users want/need in such a tool is going to exclude a lot of options that might otherwise be viable. WYSIWYG is probably the canonical example of this: non-technical users DON'T need WYSIWYG. Slack has been the (non-self-hosted) solution du jour for what you describe. My company use Sl…

> complaints were rife when they recently introduced WYSIWYG and we continue to get requests for help in disabling it.

Can you tell me how they don't like it?

Do they rather want auto formatting like markdown or no formatting at all?

Re: Dino: An open-source application for XMPP messaging

#133

Post by ryukafalz »

Andrew_nenakhov wrote: post #103
mekster wrote: post #81

It's sad that we don't have a very good self hosted chat solution. Last time I checked around, XMPP didn't have a client that I could recommend to business clients that looked easy/simple to use, has no gotchas and looks native. Currently I use Mattermost, which does basic things fine and things are stable but the formatting is just getting in the way making texts unnecessary big and bold when people never intend to…

(self-promoting here) We're building Xabber in exactly this direction. Currently, of your list wee lack multiple people voice/video calls, and also a native client for desktop. Web version is rather mature already, with android / iOS clients coming to shape fast and will soon be released. We'll likely make an Electron-based desktop version, though of course it far from ideal However, doing all this great stuff we had…

Is this MUC replacement MIX or a custom thing?

In any case, I may have to give this a try - I used Xabber a bit on Android years ago.

Re: Dino: An open-source application for XMPP messaging

#134

Post by neiljohnson »

SamWhited wrote: post #116
ajconway wrote: post #96

Earlier quoted context omitted.

Matrix seems to be the future now. The present is, however, Slack.

Matrix will likely never take off, sadly. I was hoping for an alternative for a while too, but not having a proper standards body to advocate for and push the standard probably means it's not displacing XMPP any time soon. Also the protocol just isn't great for chat; if you need a giant distributed graph database, great, but for chat it's just very heavy.

The evolution of the Matrix protocol is governed by the Matrix.org Foundation (https://matrix.org/foundation/).

XSF contributed the XMPP spec to the IETF after ~10 years, Matrix is a younger protocol (2014) and perhaps it will do something similar in the future.

Given the adoption rates of Matrix, I think it is fair to conclude it is a good fit for chat. Though it is fair to call out that it works quite differently under the hood from many other chat protocols. The decentralised architecture adds a resilience not possible in a purely federated system.

Re: Dino: An open-source application for XMPP messaging

#135

Post by neiljohnson »

SamWhited wrote: post #119
paule89 wrote: post #98

Glad they do it right from the beginning. "Encryption, encryption, encryption Insert Balmer developer meme here " My question is what does matrix do better than xmpp. What does xmpp do better than matrix?

Matrix works well as a big distributed graph protocol, not so much for chat. XMPP works well for chat or anything where you need efficient near-realtime streams (eg. stock trading and tickers sometimes use XMPP, some accessibility caption systems, etc.). Also XMPP is better standardized (by virtue of having a long-term non-profit standards body, Matrix is sort of trying to do that, but also trying to figure out how t…

There are multiple companies building on top of Matrix, some quite publicly like New Vector (https://matrix.org/blog/2018/04/26/matrix-and-riot-confirmed...), some choose to do so more privately.

However, the protocol itself is governed by the Matrix.org Foundation acting as a neutral guardian (https://matrix.org/foundation/), so it unclear how the spec itself would become part of the product.

To suggest that Matrix itself is not well suited to chat is to ignore the rapid adoption of the protocol now used by millions of people.

Re: Dino: An open-source application for XMPP messaging

#136

Post by rakoo »

ppjet6 wrote: post #92
rakoo wrote: post #64

While I've been a believer of XMPP and still believes it has a place, I have to admit that it doesn't have the appeal it once has. Now the IM protocol du jour is Matrix, which might or might not know the same fate; who knows. And another one might come later, and there are already others today that exist and are widely used. I think it's high time we had a concerted effort to have one great UI (probably per platform)…

I want to say "it's complicated". Probably any single protocol could be used to build a client that Slack or WhatsApp or [insert your target here] users would appreciate with good designers, (and/or with really little to no education.) Multi-protocol clients are generally not optimal. The set of features that can be used will be the lowest common denominator, and if it's not the case there will be discrepancies in th…

Of course I'm not saying that people should stop working on what they want, they are free to spend their time however they want. I'm looking at it the other way: what if I want to create a protocol but can't be bothered to work on the UI, because I'm no designer ?

I'm not even talking about multi protocol clients because, as you said, it tends to level downwards rather than upwards. But if I could "fork" the UI and slap my protocol on it, instead of having to build it from scratch, I'd save a lot of time.

Re: Dino: An open-source application for XMPP messaging

#137

Post by dependenttypes »

BlueTemplar wrote: post #82
dependenttypes wrote: post #80

Earlier quoted context omitted.

It is actually a good idea to use https rather than your own protocol because then your network provider/government will not be able to (easily) find out that you aren't just making https requests using your browser. Useful for people living in oppressive governments as well as employees/students under restrictive firewalls. In addition to that this will get more efficient once we have http over quic.

Quic seems to be the UDP equivalent to TCPcrypt? Both of which should be used by default on the Internet? But why should in your examples "the authorities" care or not whether you are using a browser over HTTP(S) or any other kind of software/protocol? If any communication channel is available, it's possible to send any kind of VPN traffic through it. Bits are bits. Especially when encrypted, so the censor has no ide…

> But why should in your examples "the authorities" care or not whether you are using a browser over HTTP(S) or any other kind of software/protocol?

You are going to be more suspicious if you are using $protocol-specific-for-encrypted-chat than if you are just browing the web.

> Bits are bits

I agree, but this does not stop my ISP from throttling my SSH speed.

Re: Dino: An open-source application for XMPP messaging

#138

Post by Andrew_nenakhov »

ryukafalz wrote: post #133
Andrew_nenakhov wrote: post #103

Earlier quoted context omitted.

(self-promoting here) We're building Xabber in exactly this direction. Currently, of your list wee lack multiple people voice/video calls, and also a native client for desktop. Web version is rather mature already, with android / iOS clients coming to shape fast and will soon be released. We'll likely make an Electron-based desktop version, though of course it far from ideal However, doing all this great stuff we had…

Is this MUC replacement MIX or a custom thing? In any case, I may have to give this a try - I used Xabber a bit on Android years ago.

No, it's not MIX. I believe that MIX will never fly, because it requires support by a participant's server. So unless all server developers will implement it, XMPP users will not be able to use it. Also, it puts too big emphasis on compatibility with old MUC, which is beyond saving.

Our approach does not require anything from a participant's server, and it does not even require a specialized support by a participant client: a reasonable fallback is provided for legacy clients, even if a participant is connected using 2 clients, one of which does support the new standard, and another one does not. It was achieved by clever dancing around how message carbons work and allows to sync clients nicely.

Re: Dino: An open-source application for XMPP messaging

#139

Post by Zash »

neiljohnson wrote: post #134
SamWhited wrote: post #116

Earlier quoted context omitted.

Matrix will likely never take off, sadly. I was hoping for an alternative for a while too, but not having a proper standards body to advocate for and push the standard probably means it's not displacing XMPP any time soon. Also the protocol just isn't great for chat; if you need a giant distributed graph database, great, but for chat it's just very heavy.

The evolution of the Matrix protocol is governed by the Matrix.org Foundation ( https://matrix.org/foundation/ ). XSF contributed the XMPP spec to the IETF after ~10 years, Matrix is a younger protocol (2014) and perhaps it will do something similar in the future. Given the adoption rates of Matrix, I think it is fair to conclude it is a good fit for chat. Though it is fair to call out that it works quite differently…

A correction:

The Jabber project started in 1999 and https://tools.ietf.org/html/rfc3921 was published in 2004.

https://mail.jabber.org/pipermail/xmppwg/2002-May/000000.htm... suggests that the IETF process was already started in 2002.

Re: Dino: An open-source application for XMPP messaging

#140

Post by neiljohnson »

Zash wrote: post #139
neiljohnson wrote: post #134

Earlier quoted context omitted.

The evolution of the Matrix protocol is governed by the Matrix.org Foundation ( https://matrix.org/foundation/ ). XSF contributed the XMPP spec to the IETF after ~10 years, Matrix is a younger protocol (2014) and perhaps it will do something similar in the future. Given the adoption rates of Matrix, I think it is fair to conclude it is a good fit for chat. Though it is fair to call out that it works quite differently…

A correction: The Jabber project started in 1999 and https://tools.ietf.org/html/rfc3921 was published in 2004. https://mail.jabber.org/pipermail/xmppwg/2002-May/000000.htm... suggests that the IETF process was already started in 2002.

Oh interesting, thank you for the correction.

In any case Matrix would not rule out working with the IETF in the future, but for now the Matrix.org Foundation manages the governance of the protocol.

Post reply on HN