Gajim - 2023-09-19

umu: please participate in a healthy way.

Salvor_Hardin: otr is not able to handle multi device situations and comes with.other drawbacks. It was removed from gajim, and it won't come back. I'd advise your contacts to switch to other clients, since pidgin's xmpp support is quite outdated, and it won't work reliably

https://pidgin.im/help/protocols/xmpp/supportedxep/

wymmm

pidgin supports plenty of xeps

Supporting xeps doesnt mean anything. Which xeps is the questioe

Aed the answer is old xeps that havent been updated in over a decade

With old encryption based on otr. As was said

pidgin has an omemo plugin

and supports almost all of the xeps you'd expect for a chat client

No it doesnt. Go check the xeps. And i will leave it at that because its ot here :)

wym

how does this make the client unable to chat

its not missing any xeps that you need

Hi, today I came across the concept of (proxy obedience)[https://2019.www.torproject.org/projects/torbrowser/design/#proxy-obedience]. I'd like to check if Gajim does any network activity outside the proxy configured by the user. How can I do this reliably? Looking for suggestions. :) ✏

umu: pidgin does not work reliably, as several reports showed. It.is not very helpful to promote it (here)

im not promoting it

Good

> Hi guys. All very well? I wanted to suggest something to Gajim. I'm a user of this client on another PC and if there was another encryption option, like OTR, it would be great. I have several friends who use Pidgin and if there was at least encryption by default between both clients it would be helpful. I hope I didn't bother. Hugs!

this is what i was responding 2

Why does Gajim take so much CPU when running idle in the background?

https://upload.gequatsche.odyss3us.net:5281/file_share/5IkZf0zpuBSR030bKtdPRUNf/Bildschirmfoto%20vom%202023-09-19%2009-49-45.png

3.0

ham5urg: make sure you're running the latest version

Also you can have a look at Gajim's xml console, check if there is much scrolling by

cal0pteryx, I use 1.8.1. The xml-console shows an entry ~around every 2nd second.

That should be ok then. If you want to dive deeper into this, you would have to profile what gajim actually does

This isn't typical, just so you know

> cal0pteryx a écrit : > umu: pidgin does not work reliably, as several reports showed. eh? I found it mostly reliable but dated (no edits, no archive replay in the main window, GTK2 GUI, etc) in my short time using it (which I don't currently because it was frustrating). Also the omemo support plugin is questionable. if all you need is Jabber, and the workspaces paradigm is not obnoxious to you, Gajim is a superior option.

Thanks to everyone for answering. I spoke to Pidgin people doing the same proposal and there is a work ongoing, as a priority: https://messaginglayerscurity.rocks/

Things will improve! ;-)

About the pidgin plugins, I know you can add. I suggested to the team to put as standard in order to avoid problems. For example, it is not possible to install plugins if using Tails. This is because we have to change permissions, which is not easy.

Anyway, I hope the community think not only at Pidgin and Gajim, but Dino and others. It seems interesting to make a client "polyglot". Including mobile customers.

Do any clients support MUC history limits so I can test them in practice?

Lightning Bjornsson (they, he, xe/hir): Pidgin and most of plugins for it are written in C, which destroys its reliability. Some plugins crash often, some eventually, but in general it's a big PITA. I did experience crashes even with the stock IRC plugin.

> bodqhrohro a écrit : > Lightning Bjornsson (they, he, xe/hir): Pidgin and most of plugins for it are written in C, which destroys its reliability. this does not follow. > Some plugins crash often, some eventually, but in general it's a big PITA. I did experience crashes even with the stock IRC plugin. right... i'd be keenly interested to study this

I had reported many of them and did even fix some of them, but this is an eternal whack-a-mole.

C is bad what else is new :P

Lightning Bjornsson (they, he, xe/hir): there's a clear relation. In scripting languages, programmer's error means an exception. In C, programmer's error means a crash, or a memory corruption and a crash long after that with unpredictable consequences. In an app actively communicating with the network it's highly dangerous.

MSavoritias (fae,ve): Metaprog? ×DDDDDD

irrational C hate

very boring, i'd rather get drunk than hear more

Pretty rational. Feel free to cause a memory corruption in a hipsta Node.JS app.

(you still can in the engine, because it's written in… surprise… C++)

sounds pretty easy, Node.js is inherently corrupted because it's interpreting JavaScript ✏

Modern CPUs differ so much from what we had in the start of UNIX era that they need some programming approach made completely from scratch. By someone with an extremely out-of-box thinking, ideally not knowing how does modern software actually work so they're not blinkered with existing solutions.

The way how bloated LLVM IR and modern C/C++/Rust compilers are clearly shows evolution has went the wrong way.

looks like irrational javascript hate to me :)

The whole text programming paradigm is corrupted.

i hear scratching

Nice pun.

modern CPUs pretend to execute linearly but in reality issue instructions 6 a go

several milliard times a second.

But Scratch is merely text programming enclosed into puzzle elements. Metaprog too. HiASM too. Virtually anything graphical I know too. Maybe some functional languages are not, but they're even more abstracted from the real CPU execution than the dated imperative programming paradigm.

this, itself, is inherently corrupt.

Basically, provide an autistic kid with a CPU, don't allow them to use computers, especially YouTube/Tiktok tarnation, and watch what useful would they come up with.

Maybe that would be the point when we reach technological singularity.

> bodqhrohro a écrit : > Basically, provide an autistic kid with a CPU, don't allow them to use computers, especially YouTube/Tiktok tarnation, and watch what useful would they come up with. exploiting the disabled and different now are we do you have another MUC you're in so we can free this one for topical noise?

I recently watched into the kernel code. likely/unlikely macros everywhere. Weird ASM hacks to "control" branch prediction. Why the heck did we even run into this? Throw microcode down. ✏

Lightning Bjornsson (they, he, xe/hir): it's not exploiting, it's providing with interesting tasks :P I wouldn't flood here if I would in one obviously (not).

The CPU expects to be spoken to as a linear executor, but with hints about the likelihood of various futures. But it executes in parallel.

The specter of SPECTRE. The madness of MELTDOWN. All caused by this disconnect between the CPU's reality and its programmer's reality.

> Do any clients support MUC history limits so I can test them in practice? Bump.

(like discussing other clients is less of offtopic here, hehe)

Also, is ofrom useful for anything? Gajim just not seem to use it, yet nbxmpp stores it.

Please stay on topic in this room. bodqhrohro you've been abusing this a lot

I know :C

> > Do any clients support MUC history limits so I can test them in practice? > Bump. NVM, BombusMod supports it.

(only by message count, but that's enough, I suppose, and it's specifiable by user)

good morning

when is omemo 2 going 2 b supported

It is supported.

Actually, not yet (in Gajim)

how does the macports one work?

is it any good

looking to do a large deployment of gajim