XSF Discussion - 2019-06-11

jonas’ any particular reason why you didn't hit the big green button on https://github.com/xsf/xmpp.org/pull/573, instead of 'only' approving it?

four eyes?

meh. It's easily reverted in case of something going awfully wrong. 🙂

:)

what would be the best way to share/link a pubsub item in a chat message ?

https://xmpp.org/extensions/xep-0297.html could do it

edhelas, XEP-0372 perhaps?

One requires the target users to have permissions enough, the other doesn't

indeed, looks fine !

dwd danke :3

could someone remove the sprint banner on the home page please? Sprints have happened.

will submit a PR

💝

there it is :)

jonas’, Hello, Editor!

Hello, dwd!

jonas’, Can I formally request a Last Call for XEP-0280, please? I'm willing to collect and resolve any Last Call comments (if Ge0rG doesn't beat me to it).

\o/

jonas’, I think XEP-0300, XEP-0313, and probably XEP-0357 are also all ready. ANd if not, I don't know what would make them ready. But I can only shepherd one at a time, I think.

jonas’, Oh, so '357 is pending Kev incorporating Kev's feedback. Ho hum.

jonas’, You want me to add '280's last call to the Council agenda I'm sorting? :-)

as someone who spent the weekend implementing 357 - what's Kev's feedback?

dwd, can we talk about this at 16:15Z?

I’m a bit busy right now

(I had trouble figuring out when a server should trigger a notification, which could be a valuable addition to 357 - hoping that that's part of Kev's feedback to be incorporated by Kev)

jonas’, No worries.

Guus, I don't know. It's on the list somewhere.

dwd: you are very evil, aren't you? > Can I formally request a Last Call for XEP-0280, please?

I'm still not sure it's ready, mainly because of the error bounce rules.

Also I'm sure that nobody outside of Council had a chance to appreciate the new namespace split.

JUST DO IT

Also I'm on holiday and can't provide useful feedback.

0313 is even worse.

And I'm going to -1 0357 as it is now.

Until Anu and Thilo and Daniel have written down all their special rules into it. And Holger. I'm pretty sure that Holger actually volunteered to do that.

So many XEPs with special rules that may or may not be implementation details :|

Zash: there are many implementation details which, if written into an XEP, will actually improve developers' lifes

Ge0rG, I'm hoping that a Last Call on '280 might actually generate a finite amount of feedback which can be incorporated, this time.

.

Also the rule I've added about "chat related" should be fully specified.

here we go, Movim now handle Pubsub item URIs as references in Chat

https://movim.eu:5280/upload/9d94237298995552fa13436420195fbca436dce7/SYveuOhUOR6S1od3YunMfhH3thJRF8YFR41lJgu2/Screenshot_20190611_180716.png

I'll try to handle Pubsub nodes and simple contacts and rooms uris as well

Very nice.

meanwhile I've added another chapter to https://wiki.xmpp.org/web/XEP-Remarks/XEP-0260:_Jingle_SOCKS5_Bytestreams_Transport_Method

edhelas: what happens if I don't have perms for these pubsub nodes?

Otherwise it does look cool

The card only show a preview of the post. But indeed it can leak a few info.

The issue is that I can't get the Pubsub node config as metadata from now

Well not since the last 0060 change

Hey, is it know that https://wiki.xmpp.org runs a Version that was released around 2015?

Forwarded to the iteam room, thanks.

Zash, ok thank you

jonas’, you are not in @jdev ✏

do you know can i satisfy the test vectors from 0392 with your current impl in jclib?

it seems im getting total different values

> string.unpack("<I2", sha1("juliet@capulet.lit"))/65536*360 209.41040039062

yeah i get the hue almost right

but if i pass it to the hsluv_to_rgb function it does not return the same values as in the test vectors

hm

What are you passing as the other values (s and l?)

I did `hsluv(hue, 100, 50)` whatever that means

yes i do this too

and this returns the correct numbers for you?

Not sure

`color_from_jid("juliet@capulet.lit")` → #008392

I guess that seems about right

ah i have it

he afterwards multiplys every color with 0.8

thats why it does not match for me

in jclib?

yes

r *= 0.8 g *= 0.8 b *= 0.8

not sure why, i dont see this in the XEP

Tweaking the brightness I guess?

That is most likely wrong, if you want to darken a colour you should convert to linear RGB, do the operation, then go back to sRGB. ✏

ah no i think he adjusts for white background

rc = 0.2*rb_inv + 0.8*ri; gc = 0.2*gb_inv + 0.8*gi; bc = 0.2*bb_inv + 0.8*bi;

as white is 0,0,0

he left it out

which leaves just *0.8

hm but the xep says first to invert the background color

so he misses this