XMPP Council - 2013-08-14

Same as last week's unmeeting, the only thing I see in the agenda for today is http://xmpp.org/extensions/inbox/dynamic-forms.html

ok

Will m&m be around for that IETF summary?

I will try

:D

I started the day quadruple-booked for meetings … now I'm down to double-booked (-:

heh

if I'm not here … I am −1 on dynamic forms, unless the author can explain why ad-hoc commands is not suitable

Meeting in 9 minutes.

m&m: Do you believe that dynamic-forms can be done using -50 then, without all the extra business logic that's the core bit of the proposal?

yes

I don't see it.

I mean, I can see how you could build the same stuff on top of 50 that this builds on top of 4, but it still seems to be building stuff on top of XX.

Present

ralphm: You've got 6mins yet :)

m&m: I don't see how -50 gives you ways of the client telling the server "I've got a value in this field now, but the form's not complete yet" or of the server replying with "Oh, OK, add these fields to the form, then".

-50 supports multiple stages, which could fulfill the same role

Yes and no.

Because if you want the -50 stages to support that, you then lose the ability for -50 stages to support multiple stages.

hello

that makes no sense

Peter Waher: Howdy.

m&m: It makes lots of sense :)

why can't the next stage send the form up to the point it knows of?

I've done this in the past

and now I'm in my next meeting

m&m: If you use multi-stage to be used for incomplete data submission, like the proposal needs, you can't also use multi-stage for multi-stage forms in their traditional sense.

don't know if I'll be paying enough attention here … sorry

stpeter: booh

Because you lose the ability to say "This is the next stage and I really mean it this time, please give me the next stage not the updated version of this stage".

And ding, it's time.

So 1) Roll call.

here

not really heere

m&m is absent with apologies in a strange manner.

double-booked meetings

I can't help it today

nor next week

here

maybe we should find a different time then, at the end of this meeting?

MattJ: Poke.

I'm amenable to other times.

Or other days, at least. The time slot is probably as good as we'll get.

MattJ only just joined, he can't be not here can he?

ralphm: it seems that the UPnP folks might be interested in using XMPP to cloud-enable their technology, so I'm helping them through some of the issues they're facing

I guess he's not.

2) http://xmpp.org/extensions/inbox/dynamic-forms.html Accept as Experimental.

stpeter: nice

m&m is -1, to post to list.

So I guess that makes everyone else irrelevant :)

not necessarily

since it might be relatively easy to respond to m&m's objections

Peter Waher: A single -1 is blocking, so it doesn't matter what the rest of us think.

it would be good to hear what all think

Present

i'm +0, post-back is some strange term (don't know a better one off hand) and if xep-0050 is not suitable it could at least mentioned in the introduction why not, since data-forms and ad-hoc are deployed and used

Sorry, clock I was watching seems to be a few mins out

for what it is worth, I am not really convinced that the dynamic stuff needs additional protocol wrappers

I don't see how this stuff can be done with what we have without a spec of some sort or another (be it significant business logic on top of existing protocol, or new protocol). So I'm not objecting.

I'm +0, since I haven't read the document yet, I'll read it and post my concerns to the thread if I have any

ralphm: So you're blocking it too, or not?

That said, I am not objecting to publishing.

the ad-hoc command was considered in the beginning, but was too different for the purpose. If m&m objects on list I can respond to why

considered writing something about it in the xep

but didn't

can easilly add why it was not used

OK.

So, 3) Date of next meeting.

Given m&m's the person with the conflicts, and he's not here, I don't see much point changing the date. Let's go for next week at 15:10UTC as usual, and we can change during the week if we want.

Peter Waher: design choices are always great to have inside specs like this

+1

Kev, fine with me

yes, I see that now. Thought about it, but for some reason I didn't add it. I'll correct that in my response to m&m on list

OK. 4) AOB?

none here

None here

Good enough.

Thanks all.

Thanks Kev

thanks

Thanks!

ralphm, you planned some changes to XEP's HTML rendering?

oh, right

in what direction?

basically some additional CSS

there are problems with rendering of images

Peter Waher, got an example?

img tags are rendered with width="" height=""

since width and height attributes are forbidden according to the schema

this works in FireFox

but makes images invisible in IE

(it's treated as width="0" height="0")

Peter Waher, ah..right

wonder if we can do something about that :)

if you want, I can mail you a mail I sent to Peter Saint-André earlier

you probably need to look at extensions/xep.xsl, where the attribute is rendered

yes, I've proposed a correction

great

to whom do I mail it?

can mail it to me

if you cc me then I can look into fixing it

tm@ayena.de

bear: What's your mail address?

bear@bear.im

mail forwarded

thx

got it

(y)

delivery receipts over xmpp! :D

tobias - I pushed the static site to the XSF github repo

I am using the github repo just as a working area, the final repo will be our normal one

i'll be adding more to the README to document what each area is

bear, we want to mix standards and our website repo?

mix?

well..do both stuff in a single repo

that probably needs to be talked about - a lot of it has just been a few of us just doing things

my reasoning for using github right now is just pure convienence, i'm in it every day

heh

and this test site needs some serious css love

MattJ, got a link to that repo?

Tobias: I see the gen.py depends on mercurial for versioning information?

Which?

ralphm, yes

so you can't even build the XEPs from the git checkout

you can using XSLT

right

gen.py renders the last version of a XEP that doesn't have an <interim/> tag in it

or all

part of why I spoke up to help - this needs to be "solved" for the static site generation

for some reason the XEP generated from the repo have a smaller font size than on the web site

ralphm, basically i went with hg in python because of the easy py api, i think back then there wasn't a good one for git

i can try changing it to dulwich, but that's at the end of my todo :)

I have a personal preference to hg, too

but our generation process should depend on it

should?

not. I pull and push from github using mercurial with ease

Let's not go changing our repo format :)

Kev, hahaha

Kev, ever used dulwich?

I don't have a problem with the 'official' generation having dependencies, either. It's easy to generate the XEPs yourself when you're editing them.

Tobias: No.

2) I will post my objections to the list by tomorrow (MDT)

3) apologies for next week; conflicts with JSON WG interim meeting

4) I can send a summary of the POSH BoF to some list, if people care

m&m - please do send POSH summary

+1

ok, I'll send that at least to council@, fwd standards@ … sometime this week

I need to do minutes. Probably several sets.

I think at this point you probably need to sleep :P

sleep is for the weak

It's only 21:30.

(-:

I'll try and get minutes done tomorrow.

Kev: thanks

The reason I've missed them is almost comical.

It's been hot in the UK (although not any more).

So I had my VM server shut off to save heating my study more than I needed. My TODO lives on one of the VMs. So I had nothing telling me to do minutes :)

hehehe

sounds perfectly logical to me (-:

and yes, m&m and I experienced the tail end of the heat when we were in Berlin

ugh

we Americans are spoiled by the ubiquitous and excessive Air Conditioning™

in our defense, it approached 37C outside …

just one day :)

it wasn't that much colder the rest of the week! *

* not including the deluges

not much but a bit, one day was quite rainy though

it was interesting how walking through the InterContinental, there were definite climate zones

heh

saw a tornado or so?

a few steps one way, and it went from fairly warm to chilly

fascinating and totally off-topic: http://twistedsifter.com/2013/08/maps-that-will-help-you-make-sense-of-the-world/

i find 9 particular surprising

I liked the literal translations of Chinese names for European countries :-)

moral-land, hahaha

Braveland!

That's us!

cf. map #3 :P

heh, http://www.slate.com/blogs/the_vault/2013/08/12/the_1931_histomap_the_entire_history_of_the_world_distilled_into_a_single.html is cool too

although it seems to end in 1931

I'm not convinced that claiming the UK uses the metric system is entirely right.

the UK is in transition, I suppose

"transition" :)

It's not, though, in any meaningful sense, transitioning.

It's halfway and stuck.

ah

As a result, I'm hopeless at kilometres, but useless at yards

many years ago one used to seek some road signs in kilometers in America, but not often any longer -- and that won't change until Satan buys a snow shovel

Driving distances/speeds are always imperial. Weight of a person is usual imperial. Cooking is a mix. Most things you buy are metric, but often in imperial round numbers. You buy a pint of beer (if your tastebuds are broken). Short distances could be either. Etc.

s/seek/see/

I'm looking forward to visiting England again for the IETF meeting next spring (although at the same time I'm working hard to finish up all my IETF work!)

I'm looking forward to that too, I might finally make it

I'll definitely plan an XMPP hackfest of some kind

Excellent :)

maybe I'll stay busy with IETF stuff through the Hawaii meeting ;-)

stpeter, hehe

stpeter, who around the IETF do we have to nag about updating jabber.ietf.org?

updating the website or the xmpp server?

xmpp

they are running some ancient ejabberd

I'll ping my friends at AMS about that

thanks :)

I offered to help them

they don't have to switch to prosody...but they could at least update to the current ejabberd release

stpeter: you have people living at the airport?

switching to Prosody is the plan

Kev: for what it is worth, even though the Netherlands is mostly metric, there are various things measured in non-round values

stpeter, maybe poke when 0.9 is out

MattJ: good plan -- I know they have some hardware migrations to finish first

note to self: don't set passwords longer than 40 characters, they might be easy to remember but they're hard to type

stpeter: About XSF servers, have you news about updates?

hmm

what kind of updates?

Debian 7, M-Link...

M-Link isn't an XSF matter

but that's being worked on

Jonathan Siegle has been working on Debian upgrades

it is on it like ejabberd before... ;)

ok !

stpeter: I recently learned that three simple dictionary words, separated by spaces, should have enough bits of entropy to be reasonably secure

Wonder if that's true

I think the usual pattern is four.

But I'll believe that three is sufficient.

Four certainly has more entropy than an 8character alphanumberspecial using the usual rules.

ralphm: well, this password is for my PGP key, so I was being paranoid :-)

right

But I think the maths is fairly easy.

Assume 100 valid chars in the usual passwords, so that's 100^8 combinations.

sigsegv:devel/swift/swift/> wc -l /usr/share/dict/words 8:59pm 235886 /usr/share/dict/words

Assume four words, so that's 235886^4 combinations.

>>> 100**8 10000000000000000 >>> 235886**4 3096054976816887360016L

http://xkcd.com/936/

Yes.

for those visual learners

heh, someone pinged just now about an XMPP implementation that doesn't really handle presence correctly for multiple resources -- but it seems that we're not clear about such things in RFC 6121, perhaps because we just assumed that clients would do the right thing about displaying presence correctly (or perhaps we need a XEP on this topic)

things like "don't show the user as offline just because one of the user's resource sent unavailable"

hah

I've heard many people refer to our priority stuff to bat shit crazy, too

stpeter, hehe

i think i stumbled upon that strange sentence while reading up probes stuff