XMPP Service Operators - 2019-03-29

Is there any way to increase file transfer reliability?

file transfers with my friend keep failing

Could you be more specific? Error code or the failure message displayed from the client?

tom: server? Server software? Client?

Gajim 0.16.9 sending to Gajim 0.16.6 server is.... hold on

whatever 404.city is on and....

tom: oh....that Gajim is how many years old?

Things have changed...

Prosody

Licaon_Kter, oh?

I don't think there's a release of Gajim past 0.16.x that still supports gtk2

Yeah, eg that upload namespace thing

tom: eh....

If you can tell me what's wrong I could probably figure out how to patch Gajim 0.16 myself

My friend is also stuck on Gajim 0.16.x because of the forced gtk+3 and he's a programmer as well.

Licaon_Kter, what's up?

tom: error logs?

uhm.. let me check

strange

Does Gajim have error logs somewhere? or are you talking about server logs

are you still with me Licaon_Kter ?

tom: I dunno the answer...hence....

ok

I'll keep looking into it, so far it just show a X icon next to the failed trasfers. Not very descriptive

tom, you should update, the 0.16 branch is abandonned, and so is gtk2.

And there are many features which have been added in 1.0 and later, such as more failure-proof file transfer.

Heh, nice, I don’t even have gtk2 installed on this laptop, no software I use depend on it anymore, I didn’t know. ^^ ✏

gtk2 isn't abandoned, it still receives patches, if if that were to change I will patch it should a need arise

and gtk3 is not a generic toolkit, it's the gnome graphical toolkit

it's also optimized for tablets, not computers.

and it isn't stable

I'd much rather fork Gajim 0.16 if nobody else is going to

backport useful features

gtk3 is so unusable to me that the effort required to backport and maintain 0.16.x is worth the effort for me

even spending money to pay for development help

tom, gtk3 will not see any newer minor version, only patch releases, that is the Debian definition of “stable”.

It also is definitely not “optimised” more for tablets than for anything else.

And yet also, not targetting GNOME in the slightest, there are just people involved in both projects.

I've used the 1.x release

of gajim

and it's probably the worst release i've ever used ui wise

now if there is another jabber client out there that you'd recommend that doesn't force gtk+3 and doesn't fall for whatever the latest UI fads are and that isn't based on nodejs I would like to know

Please open issues for things which could be improved.

tom, hmm, maybe look for the Qt ones?

Swift for instance, or Kaidan, or JabberCat.

Link Mauve, I don't think Gajim wants to improve. I took some of my concerns to their muc and they just insulted me

Or even Converse, it uses node and python2 and C++ to build but then it’s fully client-side JS at runtime. ✏

I also don't see a valid reason for switching off of gtk2

tom, I’m pretty sure it does.

tom, oh, there is a very valid one: it isn’t supported by anyone.

Well I will take your word for and and open a ticket

Link Mauve, What do you mean supported?

If your suggestion is just “switch from gtk3 to gtk2”, of course nobody will do that.

But if you have valid UI or UX concerns, that’s relevant.

tom, no one is pouring time to fix or improve gtk2 anymore.

what is there to fix?

Really?

Have a look at all of the reasons for gtk3’s existence. :)

The main one for me was Wayland support.

gtk3’s is damn good.

gtk2’s, well, inexistant.

the last patch to gtk2 was 14 days ago

commit c505d3f what do you mean not supported?

also, what reasons for gtk3's existance?

wayland isn't even stable yet, and doesn't work on the BSDs

Look at the release notes if you aren’t aware of it.

and X11 forwarding over SSH doesn't work on wayland

tom, erm, can you please stop with the false accusations please? Wayland has been stable since 2012 and has worked on any system with UNIX sockets and shared memory.

tom, it sure does, as long as your compositor provides an Xwayland integration (and all do, to my knowledge).

This escalated fast....

really? I didn't know

Does wayland work on OpeBSD?

tom, sure.

I mean, it certainly has unix sockets, and shared memory support, as it exposes a POSIX API.

So, as a person who has used X for a decade and has not had a problem with it, why would I want to use Wayland instead? and Why do most operating systems not ship with Wayland, but X instead?

Some compositors may have additional requirements, such as OpenGL support for rendering, or udev for input, but at least when I had to run Weston on FreeBSD for a client that was working.

tom, personally, it was because I was tired of tearing happening all the time, and because of the complete isolation between processes, as well as the very simple and easy to understand and to extend protocol.

I'm certainly not going to go ahead and install Wayland just to run a graphical toolkit that has not yet even justified it's existance over gtk2 that I don't even like

I've never experience tearing. Are you sure that wasn't just a graphics driver causing tearing?

For instance I wanted to add quad-buffering support to a 3DS emulator, because that sounded like fun to be able to run its stereoscopic capabilities on a big 3DTV or VR device, and that was pretty much impossible on X11.

tom, it’s a combinaison of many things, Xorg being single-buffered being one.

I am personally running 2 monitors, one of them being a 144Hz and 1080p while the other being 1900x800 at 60hz.

I don't experience any tearing besides what on the 60hz one if I play a videogame on it without vsync, but that's the monitor's fault. it's a cheapo unit and does that across any os

I’m pretty sure it isn’t due to the monitor, try it with Wayland for instance.

I also had a lot of misconceptions before getting deep into the graphics stack. ^^

I can see process screen isolation being a useful feature if it doesn't break things like window managers, oneko, and such

but other than that, I don't really see how that only justifies a huge compatibility breaking change like that

Of course it prevents your window manager from running, and oneko would have to be a priviledged client or part of the compositor.

that doens't sound very good

But it’s becoming very simple to build a full compositor by writing only code that you would write for window management before, with libraries like wlroots.

So it’s only a matter that people were targetting Xlib only, and now have to target either Xlib or wlroots or both.

so other than wayland compatibility, (which I don't see has justified or being worth the hassle) why else would someone want to switch to gtk3?

Depending on whether they want full control over the display mechanisms, or if they want to delegate that to Xorg.

In the latter case, they can’t support Wayland.

tom, have you just made the effort of opening the release notes or design plan of GTK+ 3.0?

If not, then there is no point to this discussion.

https://lwn.net/Articles/300303/

qrong link

I can only find the release notes for gnome3

would you provide a link?

tom, second result on ddg: https://mail.gnome.org/archives/gnome-announce-list/2011-February/msg00022.html

basing the theming api on CSS is a horrible idea

let's build on the shoulders of giants?

Note that that’s only the 3.0 release, not everything was done by then so the following releases added a lot of missing pieces, while always keeping it ABI compatible.

dbus integration is also a horrible idea. dbus sucks

tom, I don’t see what’s wrong with it, most designers these days are used to CSS for apparence.

just because something is popular doesn't make it good

also, what was wrong with the privious theming api?

*previous

Most likely, that it wasn’t popular enough, so random passerbys had to learn an entirely new language to theme their application or DE.

But that’s a thing better asked to GTK+ people from 2008.

I can only speculate on their intentions.

there were and are plenty of great gtk2 themes already out there

Although this change seems to have panned well.

also, learning how to use your tools is just a fact of life

regarding the last relase note, added an application chooser, I have a perfectly fine gtk2 based application chooser that didn't require a whole major release number to implement

tom, so what good does it give to have to learn two different languages, to make web applications and GTK+ applications, while it could be a single one?

https://0x0.st/zPsZ.png

because there is a difference between web applications and programs

that line should not be blurred

Eww, Flash.

That’s a thing I haven’t used in a decade.

tom, using CSS for theming is hardly “blurring” anything.

And why should it not, btw?

Some other toolkits like Qt went much further with the integration of web technologies, integrating a JavaScript engine, CSS, QtQuick, and even Chromium for their web needs.

when you attempt to blur those lines to bloat the web and make turn a web browser, not into a markup renderer, but a virtual machine

and Javascript is a terrible terrible language. It always was even from the beginning

You can polish poop but at the end up the day it's still a peice of shit

the compatibility breaking change of using css for theming does not justify itself in a cost to gain ratio. the gtk2 theming api is perfectly expressive enough for what it does and simple enough too

I've made and modified my own gtk2 themes, i'm not talking out of my ass here

>tom‎: when you attempt to blur those lines to bloat the web and make turn a web browser, not into a markup renderer, but a virtual machine sorry, something happened here. I meant to say: when you attempt to blue the line of web browser and programs, you end of turning a markup renderer into a bloated, virtual machine

*blur

does that answer your question?

just to give you an example of the bloat of the fallacies that have brought us "web 2.0" firefox 52 (even before you had to compile rust to compile firefox) took and still takes longer than GCC to compile

GCC used to be the most bloated and longest to compile thing of the entire operating system

>* More flexible geometry management, with support for height-for-width, for both widgets and cell renderers. Why did a compatibility breaking change be needed to implement this?

* Modern input device handling. The input device handling in GDK has long been a sadly neglected area. This has changed; with 3.0, GTK+ steps into the modern world of XI2 with full support for multiple pointers, keyboards and other gizmos. Sounds like a GDK problem not a GTK problem. Just patch gdk

Sorry I was doing something else.

no problem

But a compatibility break eleven years ago, with a lot of other things required to change, has long since given benefits.

The gtk2 theming API is just an artifact from the past nowadays.

WHAT BENEFITS?

tom, but I wasn’t arguing for bloating the web, it already does that on its own perfectly well, I was talking about using the useful parts of it, familiar to many, to make your own software more familiar.

tom, see, I’m a random who only knows HTML and some CSS, I want to change some visual things in my IM client.

Now, with gtk3 I can just apply my preexisting knowledge.

yes, and I'm saying maybe so (i doubt it though) but that doesn't justify breaking everything to change it

With gtk2, I’d simply be demoralised away by having to learn yet another language.

well that's your own fault

tom, the break happened for many other reasons, your gtk2 themes wouldn’t have been compatible with gtk3 even if they kept the language.

tom, sure, you can blame the user.

You can also accept criticism and aim for a more user-friendly experience.

being demoralized from having to learn new stuff is

I’ve met a lot of people identifying as web developers, who outright reject the idea to learn anything else.

And they can live perfectly well.

Same for Java developers, same for C developers, same for Python developers.

that's why they are web developers

and you can stay in in a bubble

Yes, that’s in their self-description. What does it bring to cut yourself from these people in order to undo an eleven years old language change?

With the hingsight, it’s only brought nice things.

what? I don't understand your metaphors

tom, there is no metaphor here.

no one is self mutilating

and years of constant breakage and instability is not a nice thing

also completely abandoning decades of UI research for hamburger buttons is not a nice thing

Heh, still arguing about using gtk2? :p

Mhm... Hamburgers!

a button for buttons

tom, “cut yourself out” means making a clear separation between you and a given set of people in English.

tom, that’s an application developer decision, whether they put menus or other things.

Nothing the toolkit can do for you.

Ge0rG, oh, good idea.

I should have breakfast, and start packing, I leave in less than one hour.

And get out of bed too.

because if you want to enter a new field, you should learn the tools of the trade. NOt attempt to adapt tools of a different trade before learning the traditional way of doing things.

at least learn from the more experienced before attempting to forge your own way

I don't go get a job as a car mechanic and refuse to learn a wrench and attempt to use a keyboard to change my oil

tom, but as a random user, do you want to enter a new field, or simply customise your chat client?

Using a chat client should never have the same requirements as getting a job as a car mechanic.

I want a XMPP client for my computers that doesn't have a phone client. Also, If I'm going to be upgrading I want the upgrade to at least be equivalent or slightly better than what I had before

*phone interface

Same for me, except I don’t care where the inspiration comes for the interface.

I want a computer interface on my computer. not a tablet interface on my computer

I mean, I use a client inspired by terminal IRC clients, of course I don’t. :D

I also refuse to lose customizability I had before

for example, I don't want material design, animations on checkboxes, and rounded bezels

Gajim's 0.16.9 interface is probably the best IM interface I've ever used. Like ever

then 1.x just completely throws it all out the window

why?

because windows are "old"

not because what they are using now is better, but because popups and windows are "old"

just because something is new doesn't make it better. and just because something is old doesn't make it worse

focus-stealing popups always have been bad

If nobody else is going to fix file transfer in Gajim 0.16.x I'll do it myself

I would however appreciate some help in figuring out why exactly the issues are occuring in the fist place

“10:54:10 tom> for example, I don't want material design, animations on checkboxes, and rounded bezels”, great, all of those are customisable using *wait for it* CSS!

tom, you can read the history between your latest commit and master.

But be aware that it’s very long.

And I wouldn’t advise doing that.

Why should I when gtk2 works absolutely fine?

also the file picker in gtk2

Nothing, not even the qt one comes close to it

Can you show me a gtk3 theme that looks like gtk2?

Raleigh?

hmm. ok. I could use that as a middle ground in the meantime to fix the glaring ui design failures, but still, as with all my previous posts, the change to gtk3 is not justified in a cost to gain ratio

With most software having moved, maintaining them all in an old version will only accumulate work.

So, good luck in your impossible quest I guess.

At some point maybe you’ll realise the cost to gain ratio is really not to your benefit.

In the meantime, I’m going on weekend, see y’all next week! \o_

Link Mauve, fuck you

:D

I'm sick and tired of your passive agressiveness

I totally agree with Link Mauve (/me throws oil on the fire)

why?

Maybe you should start a gtk2 support group for people traumatized with gtk3

maybe you should stop being ass asshole

If you manage to gather enough people to maintain gtk2 it might become worth your time

gtk2 is maintained, what are you talking about?

the last patch was only 14 days ago

are you actually asking or just trying to piss me off

What really needs to happen is an anti-redhat group

Every single thing to come out of redhat minus qemu patches and md has been a downgrade in the oss ecosystem

dbus

GNOME3

consolekit

PulseAudio

FreeDesktop

systemd

lol

ansible

that too

hahahaha because manually configuring every single server is so much better

Redhat's a special fucking snowflake and can't using a staging tarball like every other unix os

noting wrong with gnome2

I actually used to use it

and i'm comming form 10+ yeasrs of using only tiling managers

tom i mean gnome3

ok

is there anyone here that agrees with me?

i moved to gnome because they droppped what they were doing and decided to do something new (aka gtk/gnome3)

anyone at all?

seriously, what the hell

what the hell happened to the industry in the last 7 yeays]

tom, i guess it moved on

I dont see what your problem here is tbh

why?

because as everything things move on

Why is bloat and spyware the norm now?

because people are borred of doing the same thing over and over again

i dont know

that's too general

spyware?

your just fucking with me

where is the spyware in anything you mentioned?

I'm trying to have a serious conversation about the stuff coming out of redhat and web not being good and all I get is assholes trying to derail

like yourself

I don't know why I spend the effort explaing this shit

tom, imo all you do is shout some random shit, fueled by your frustration

maybe I think your capable of reason

obviously not the case

itsa very simple. you dont like gtk3, as you mentioned gtk2 is still supported. whats your problem?

someone higher said he likes web stuff. you dont. whats the issue. develop/use non-web stuff. period

I dont like propretary software nor apple/google/windows etc. I dont use it

how's that for a reason

the problem is I asked a simple question, what changed that made file transfers more reliable, what spec do i need to implement, and I'm just met wet "just upgrade to gajim 1.x, oh you don't like gtk3? you must be some of of fuckhead then i'm right your wrong and stupid

good luck with your impossible quest

I don't need you to fucking tell me to switch to fuckin wayland

tom, thats why i said i dont know what the initial question was.

tom, noone tells you to do anything afais

oh, were you not here for that conversation?

I'm sorry then, that wasn't for you

tom: the specific issue with gajim is that the older versions don't support a ton of feature that are rather necessary nowadays. You might not like gtk3, but then you're cutting yourself from these other features and lose your rights to complain about them not existing (just like using pidgin nowadays is shooting yourself in the foot). You could make a gtk2 version of the newest branch of gajim maybe, or port these missing features in the older branch

> a ton of feature WHAT FEAUTURES?

> I asked a simple question, what changed that made file transfers more reliable, what spec do i need to implement https://xmpp.org/extensions/xep-0363.html

tom: https://dev.gajim.org/gajim/gajim-plugins/wikis/httpuploadplugin

is that it?

That's my answer to your question about which spec offers more reliable file transfer, yes.

so there's no way to fix peer to peer transfer on XMPP, it's just been replaced with http uploads?

I *think* that plugin should even work with 0.16 so maybe there's actually nothing to do besides installing it?

tom: Well p2p is hard(er) to make reliable.

can it be?

let me rephrase

Most clients basically replace it with upload yes. Not just because reliability but also to make it work with offline peers, multi-device setups, and group chat.

if implementing perfectly by the spec, can p2p transfer be made to be reliable?

I agree it's not exactly elegant but the fact that it just works is kinda nice.

sure

tom: The problem is that there's multiple specs and incompatible revisions of those in the wild. Plus the limitations I mentioned above.

tom: Apart from that I think it can be made reliable, yes.

hmm.https://xmpp.org/extensions/xep-0363.html says it's still only draft

Yeah if you take the XEP status seriously you can basically forget about XMPP.

For p2p you'd do XEP-0234 these days which is 'Deferred'.

hmm. Looks like i'm going to have to switch XMPP providers to get that to work

not a huge deal, I've been looking to host my own for a while now

Haha, I'm not the only one grumpy about the Gnome CADT model and all the new shit that's working differently from the old shit that I know to operate.

BTW, hosting your own XMPP server is the first thing to be actually on-topic in this room

I don't quite get why everyone is mourning for the Gnome2 UI. It's just Windows 95 and there's still plenty (most?) DEs doing just that.

I think the problem is Gnome 3 with its opaque buttons where you have no way to find out what they do short of clicking and looking

Hmmm. Dunno I chose Gnome 3 as the first DE for my little son and he seems to cope just fine.

Not sure what buttons you mean exactly. There aren't many in the first place :-)

(And on-topic is overrated.)

Whatever, I don't care at all what others use. I'm just happy the GNOME people tried something else because I think there's still plenty old-school DEs if you prefer those.

Ah I used that myself one or two decades ago :-)

Holger, agree. I like gnome because its different. I used pretty much everything out there from fluxbox, xfce, kde, through awesomeWM and i3. I just deployed Gnome3 desktops at work and i was amazed by the reaction from normal non-tech folk that find it super comfortable to use. some even say its better then MacOS and i start recieving windows laptops form them (and now some of their friends that saw it) asking if I would "load" that linux thing on their personal computers.

so i guess gnome is doing something right. its not for everyone but so isnt Mate or tilling window manager

thank you for your help Holger

muppeth, I think that's the problem. GNOME3 and gtk3 alienate 'tech folk' in favor of computer illiterates when their previous userbase is power users

tom, totally disagree. I consider mysefl power user. I used tilling window manager exclusively for over 10 years. I moved to gnome now, being still power user

i do undertsand people want to use tilling managers still

heh

everything is one huge tile in gnome3! :D

i dont uderstand how using a DE make me less of a power user?! I use terminal for work. tbh most of the time i dont even need XServer to do my work

I find myself doing to opposite, slowly moving more and more into tiling managers because i'm being alienated by gnome

not because I like tiling

because I can't find another option

I"ve moved from gnome2 to matte, then they went gtk3

then to xfce

then they went gtk3

maybe because gtk3 is better

then to i3wm because I'm not sure whereless to go

tom: I understand how the GNOME transition was painful and alienated parts of the existing user base, but I totally appreciate *some* DE project targetting "computer illiterates" so I have something good to offer them.

there was already a project that targeted computer illiterates, and I don't see why gnome had to alienate it's own users to switch target userbases

But yes if it's about avoiding GTK+ 3 I see your problem. Qt not an option?

https://www.magicdesktop.com/en-US

People are almost never happy with change anyway, what's good for one is going to annoy the other. I'm fine if I'm not GNOME's target user anymore (or never was), I just don't use it

but what is your problem really. gnome poeple wanted to do something and they did. according to your logic, software cannot change because there is this tom guy and he wont like it.

tom: Well I think it's hard to build a single piece of software that makes both power users and illiterates happy.

qt is not as bad as gtk3, but it's still a downgrade. It also has no support for C

also the file picker

Holger, why? gnome3 or kde does just that. it makes non-tech and tech people happy.

doesn't worker nearly as well as the one in gtk2 does

Whoever is using C as a UI language, deserves all the resulting pain

muppeth, no, kde is hard to justify in that case.

just some poeple dont like it for whatever reason, but I'm pretty sure there is a DE that fits them too and if it doesmt, well time to write your own tom

KDE pissed people off soo bad and much that they went and created the Trinity Desktop Envrioment

in the kde3 -> kde4 transition

tom which people?! all 5 of them?>

how many greybeard linux nerds do you need to write a Desktop Environment?

muppeth, cut the smartass unhelpful remarks

I'm trying to have a conversation

tom, my remarks are no different then your whining. you blame decissions made by gnome team from your perspective.

whats the point of this 'conversation'?

whine that everyone

>‎[05:10:18 AM] ‎muppeth‎: tom which people?! all 5 of them?> >‎[05:06:01 AM] ‎muppeth‎: maybe because gtk3 is better

whine that everyones ideas suxx because its not what you want?

>[05:11:35 AM] ‎muppeth‎: whine that everyones ideas suxx because its not what you want?

tom, as you pointed out everyone moves to gtk3. it means it must be good right?

your attempting to derail to conversation

tom: you are derailing the conversation for hours now.

This is the XMPP Operators Room. It is about XMPP Operators.

muppeth: In the context of window managers I'd call "power users" those who want a high amount of configurability/flexibility/extensibility and/or powerful features that aren't necessarily intuitive (i.e. tiling WMs or so).

that doesn't justify muppeth's passive aggressive snipes

muppeth: In general I think offering more powerful UI often conflicts with intuitiveness because it requires a steeper learning curve.

tom: you lost the remains of your credibility when you told Link Mauve to f*** themselves.

muppeth, your rude

Holger, yes. exactly. check dconf (very unintuitive, anti-userfirendly) tool to change every single bit of your gnome experience 😛 very power user frioendly 🙂

muppeth: Hence I think it's often hard to make both target audiences happy in one go.

tom, I'm rude? I got active in here because i saw your insults towards others here.

muppeth: Nah dconf doesn't allow me to turn GNOME 3 into i3.

Holger, i think its a good balance. non-power users have very little options to get overwhelmed or screw up with, while other can change every single piece. not to mention you can hook it all up to ansible and manage/configure every aspect on every desktop you administer.

Holger, but also i3 doesnt allow you to become gnome. there are restrictions, although the tiling functionalities in gnome are getting interesting lately, so soon you will actually be able tyo make gnome into i3 while not the other way around 😛

anyway as i said there are no bad DE/WM just different expectations of poeple, and its nice that you acctually have choice.

Can we please get Poettering into here to discuss the merits of gnome3?

If you ask me all this shit is just the Extend in Embrace Extend Exterminate

and about non-power users have very little options

non-power users have Windows and Mac OSX

and they have https://www.magicdesktop.com/en-US

Why is the so much of a concern to write software for computer illiterates anyways?

tom: because most users are computer illiterates

Why are a lot of developers writing software for computer illiterates instead of themselves out of being paid to do it?

tom: the right question would be: why are so many developers writing software for computer illiterates, despite lacking the required skills?

no, I'm asking why they would want to be going through all the effort of writing software for someone else with different interests and abilities while not even being paid to do it

tom: because devs won't have work to do....?

if that's the case why don't they spend their free time writing software for themselves or their own interests?

> if that's the case why don't they spend their free time writing software for themselves or their own interests? My impression is that's still the case for most spare-time devs.

I'm a spare-time dev writing software for other people.

You're different!

Damn.

Another impression is that there fraction of free software produced in spare time is much smaller than back in the days. ✏

So it's more driven by profitability these days.

hm

Not all of my software is profitable.

You're different!

Ge0rG is?

why?

tom: I am why what?

> ‎Holger‎: You're different!

tom: apparently because I'm writing software for normal people, and because not all of my software is profitable

but that's what Holger was saying your different from yes?

Holger, what you care to clarify?

Ge0rG is working on an XMPP client that targets normal end-users in his spare time. And my impression is that this is a somewhat uncommon case.

Or maybe somewhat specific to the XMPP/chat field. Where you have an interested to chat with others, possibly with illiterates. Which requires software that's usable by them.

chat with illiterates! \o/

:-)

by normal you mean computer illiterate right?

Yes.

k

If I might make a reccomendation

if you could make a stripped down xmpp client that only works in google chrome

as a 'web app'

it's called converse.js :P

tom, I dont understand your argument. I'm not computer illiterate and i do like for example gnome. I have a feeling for you software for non-illetirates must look like its still 90s or something.

most computer illiterates don't know how to use software and think that installing programs gives them viruses

they do everything from a google chrome. just look at email

tom, not true. I run a platform that provides services to people and you would be suprised how many 'illiterate' people actually use it while you would think its only for the elitist power users.

world isnt 0 or 1

i know few 'illiterate' people what use old xfce. not because they are power users that dont like change, but because they are used to their old wondows95 UI

if your looking for adoption, that might be the way to go

your preference for software does not make you better then others.

or software better then the other

In my opinion computer illiterates should simply stay away from computers.

I fully agree with vanitasvitae on this one, but unfortunately that's not going to happen

You should only be able to use spftware, if you fully understand how it works.

well maybe, if you can isolate computer illiterates to smartphones that might work

Nonono

Smartphones are computers as well

Even lightswitches are, depending on the scientific model you use.

> You should only be able to use spftware, if you fully understand how it works. Nobody does that. Please turn off your computers immediately and become a potato farmer.

vanitasvitae: turing complete lightswitches?

Ge0rG: potatos can electricity ✏

about the chrome-only webapp thing, that's actually a fairly popular thing. Thats the enture premis of ChromeOS

You don't have a computer, you just have a netbook that only runs a web browser

tom: https://inverse.chat

Ge0rG: lets collapse society!

yeah, that doesn't even do anything but render a solid blue page in my browser ,even with javascript and cookies enabled

vanitasvitae: it was your premise. I'm perfectly fine using tech I don't understand

vanitasvitae, still it doesnt mean my workspace cannot look nice and needs to look like shit crafted in mid 90s. i still dont see how is gnome / gtk3 made for illiterates and hated universally by everyone that knows a bit of how stuff works.

"Mom, you're not a full stack developer? No Candy Crush for you!" Simple as that!

hahahaha

I hope by now its obvious that my statement was not meant serious :D

its more of a response to tom then to you vanitasvitae i think

though candy crash should be made illegal

I'm with you on that point ;)

Is there any reason the Prosody website recommends installing Prosody from Ports instead of Packages for OpenBSD?

Probably because someone familiar with OpenBSD told us to put that there

I'll take that advice then

Is there any reason to put the muc on a seperate subdomain?

or can you put the muc on the same domain as the main server, for example xmpp.domain.tld instead of xmpp.domain.tld and muc.domain.tld

or conference.domain.tld

tom: some clients will expect that, or confuse things when MUCs are not separate

tom: btw you shouldn't have "xmpp" or "jabber" in your xmpp domain, use SRV records for your main domain instead

sorry?

thanks

tom: when a client receives a carbon copy of a MUC-PM message it needs to check whether it's a MUC or not, and a set of clients will revert to checking that on the domain

I'm sure now that I've said "carbon copy of a MUC-PM message" you understand the problem.

so it's better to put in on it's own subdomain for buggy clients?

it's better to put them on their own domain because of an oversight in the protocol.

Ge0rG: really? This is how I've been using, xmpp.example.tld with A record and an SRV record to xmpp.example.tld. What kind of problems can arise?

lag: it's not a problem, it's just a remnant of the old times when you didn't have SRV records to allow using your normal email address as a JID as well

I wish it was possible to use nextcloud FIDs be the same as emails too

(federated id)

you can have that with xmpp for 20 years now.

nextcloud is fairly new

yes, and they went to reinveng messaging.

because there isn't enough messaging already with xmpp and matrix.

they did?

as far as I understood, yes

and then they realized people want federation, and were like: ugh what? how? what now?

Double check that, Nextcloud is more of a file/calender collaboration suite

IIRC it's called "nextcloud talk"

They have a video chat, but I think that's only for people who you need to talk to but like only have skype or something

I remember there being a plugin somewhere for xmpp integration, but really nextcloud is more of a file and calender sync service

I can only keep up with so many messaging silos.

ask me about XMPP.

ok

your have any experience with mod_onions and mod_darknet ?

no. I think that if you are into zero-metadata comms, you should use a protocol designed for that, like briar

no, it's more about stopping illegal dragnet mass survailance

transparent darknet integration is nice

Germany is forbidding the darknet now.

really?

I thought germany was a democracy?

I wonder how they plan to enforce that

I don't know yet.

tom: swat teams as usual? As CCC

seems super sketch, banning network anonymity

https://darknetlive.com/post/germany-s-proposed-law-to-ban-tor-nodes-in-english/

oh, so it's not law just a proposal

it was accepted yesterday

or at least passed one of the parliamentary chambers.

But the politicians are sufficiently clueless to let it through, especially with the right wing being rather strong

Ge0rG: I saw that not so long ago :(

communist china bans tor, but there's still tens of thousands of users

I think it's time to make more tor nodes

you beat me to it

There's no better time to setup tor than when governments start trying to ban it

tom The reason for transfer errors in large file sizes, no one thought a few years ago that large files would be sent via http upload

thank you 404.city

Link Mauve, https://www.youtube.com/watch?v=ON0A1dsQOV0

maybe it's time I take another look into qt

Does Psi support XMPP well?

tom: depends on your definition, it does in my opinion

sure, I mean does it implement most of the important XEPs properly

the comparison would be pidgin, which doesn't do that