XSF Discussion - 2021-09-19

> I wrote: > Isn't private group conversations on Xmpp spoken without verification? > I think some things need to be clicked on. Is there such a issue?

bung: I'm sorry, but I don't understand your question. What do you mean by "verification"? Verification of what?

> Sam wrote: > bung: I'm sorry, but I don't understand your question. What do you mean by "verification"? Verification of what? XMPP requires you to verify the key of the user you are talking to.

bung: that's a general requirement, not xmpp-specific. Otherwise you don't know who you talk to

> mjk wrote: > bung: that's a general requirement, not xmpp-specific. Otherwise you don't know who you talk to The question here is whether the cryptic conversation is ready.

Do you need to click on something or not?

Are you talking about a private MUC using OMEMO?

> Sam wrote: > Are you talking about a private MUC using OMEMO? Yes

You mean can encrypted communication be established without verification? Sure, it helps against passive attacks

I'm still not entirely sure what the question was, but I think the answer is that most (all?) OMEMO implementations use a method called trust-on-first-use. They trust the keys a contact sends, and its up to you to mark them as verified or not. So yes, you'd have to click something if you want lots of little green lock icons, but if you don't everything should still work fine.

> Sam wrote: > I'm still not entirely sure what the question was, but I think the answer is that most (all?) OMEMO implementations use a method called trust-on-first-use. They trust the keys a contact sends, and its up to you to mark them as verified or not. So yes, you'd have to click something if you want lots of little green lock icons, but if you don't everything should still work fine. Can I use it without marking?

But encrypted

Yes

There we go https://github.com/xsf/xmpp.org/pull/964#issuecomment-922464835

Looks neat Zash. How does it look in a narrow window though? ;D

In other news, my one-byte patch is finally here! https://github.com/xsf/xeps/pull/1106

mjk, https://github.com/xsf/xmpp.org/pull/964#issuecomment-922470850

Ah, so it does reflow the table... if push comes to shove

s/table/whatever-this-is/

The actual XEP table doesn't deal with it of course...

It's pretty viewable on mobile anyway

https://cerdale.zash.se/s/670GESAng33Jbmpe/UZjAbHoFTuWpyaiFfiJPSA.jpg

Perfection

Oh, it didn't re-render the shortcodes? Hrhrhrrr

Seems `hugo serve` doesn't deal well with accessing from a different device 🤷️ ✏

it caches some layouts and doesn't rerender, I forget what

It's very annoying.

Ah well, I feel like it's Good Enough for now. It can always be incrementally improved later.

> Perfection I say yes (not counting the listlessness)

No-CSS is best CSS!

I tried using definition lists

Ah, that. I almost didn't notice. I meant the literal lack of list-like rendering in the second and third lists ✏

And that's their default style?

Html :shrug:

yes, except in bootstrap undoes everything nice about them unless you add some class I can't find anymore

I like how https://prosody.im/doc/developers#core is rendered

But that's Bootstrap 3 and xmpp.org is Bootstrap 5 apparently, and now I don't know anything anymore

uggh, I didn't notice we were using bootstrap. It's gotten so bad. I volunteer as tribune, err, I mean, to purge bootstrap and create the tine stylesheet with a few dozen styles that's all we need.

(but seriously, if people are game I will do that later in the week when I have a day off)

Zash: for access with other devices you can run hugo with address bind to 0.0.0.0

Ah yes, the "I'll just create a tiny stylesheet with a few rules" trap. And before you know it you've developed another 10kLOC CSS framework.

I'll have a look at the column response in the evening, but that should be easy with the grid system breakpoints

> (but seriously, if people are game I will do that later in the week when I have a day off) Please don't

I have literally never had that happen. I have seen it happen when people don't know how css works. I'm no expert, but I've never had that happen.

Anyways, this can be simplified and I'm willing to do it and maintain it if someone is willing to merge it.

🤷️

I would rather use Bootstrap (or any other actual project) than begin maintaining our own CSS framework

I do not want us to maintain a framework either. That would be bad.

I just think that for a small site like ours there is no point to using a framework at all.

The benefit of frameworks is that they are tested across browsers and documented. Fidding with CSS rules for everything you want to do is far more annoying than just looking up e.g. the grid documentation for your framework.

I'm not particularly a fan of Bootstrap, but I think that all but the simplest of sites benefit from using an established project

I don't think ours is a trivial site, it has tables and menus, we want columns, and we want it to be responsive

I disagree, it is trivial to make things work across browsers these days. Everything supports grid and flex layouts. The browser issues that used to necessitate frameworks are more or less gone for simple sites.

> it is trivial to make things work across browsers these days

These sound like famous last words

Well, probably not, as long as the set of browsers consists of "latest firefox and latest chromium" :)

ITYM "~latest firefox and~ latest chrome ~ium~"

Nah, things aren't that bad with ff these days

> I don't think ours is a trivial site, it has tables and menus, we want columns, and we want it to be responsive Exactly. There is more than you might see at a first glance

I don't think that's true, bit maybe I'll do it and be proven wrong.

Sam: that's the spirit

Why though? It works fine doesn't it?

We now have something where people can come with an idea, look up solutions in the docs, put together existing building blocks, and have something up and running in minutes.