XSF Editor Team - 2015-10-15

howdy

yet another RFC with your name on it

and you :-)

I still have 3 I-Ds to submit before the cutoff on Monday, sigh

do you want to skip on editor stuff then>

?

it's OK, let's make some progress

alright

first, something we're going to need to address

Kev found what he feels are substantive changes in XEP-0030

around how xml:lang factors into uniqueness

right, I saw a message about that somewhere

did we track down the check-in that modified this?

yes, sometime back in 2008

this paragraph, eh? "Each <identity/> element MUST possess the 'category' and 'type' attributes specifying the category and type for the entity, and MAY possess a 'name' attribute specifying a natural-language name for the entity; the <identity/> element MAY also possess a standard 'xml:lang' attribute, which enables the entity to return localized results if desired (i.e., the <query/> element MAY include multiple <identity/> elements with the same category+type but with different 'xml:lang' values, however the <query/> element MUST NOT include multiple <identity/> elements with the same category+type+xml:lang but with different 'name' values)."

yep

https://github.com/xsf/xeps/commit/bf110edfccddf0c22089df2861e0231bdb05c3a2#diff-86c7b1b1dbf6dd553b5a7b53a1cb87ce

I think I suggested list discussion, mostly (disclaimer, I'm ill today, not firing on all cylinders).

Just to check we weren't about to include a substantive change without discussion.

It looks like the change in question was ages old, but never included in a new revision.

And at some point it accidentally got regenerated into the current version, like the 313 stuff.

I'd need to review the discussion list history from June of 2008

and earlier

perhaps this was discussed as part of the run-up to 2.4 but was left out somehow.

I didn't do that. But yes, that would be sensible.

I can't recall.

I think I did a search of my mail archives for likely-looking queries and didn't find anything, but maybe I'm mistaking it for something else.

Anyway - my block was the temporary check-we-don't-accidentally-break-something type, not the this-must-not-happen type.

I don't have a quick answer.

I'm not sure that -1s are the appropriate mechanism here, but I don't think we have anything else available to say "Oh, wait, a minute, before we do that...", because otherwise the votes expire and it runs through on majority.

I might have been thinking this was a sensible clarification, I might have received some implementation feedback asking for clarification, it could have resulted from discussion about XEP-0115 somewhere, etc.

-1 on something that seems broadly right feels wrong, but there we go.

I'm now thinking of XMP Council's -1 as equivalent to the IESG's *DISCUSS*

cf. Section 5.1 of XEP-0115

e.g., "Sort the service discovery identities [15] by category and then by type and then by xml:lang (if it exists), formatted as CATEGORY '/' [TYPE] '/' [LANG] '/' [NAME]. [16] Note that each slash is included even if the LANG or NAME is not included (in accordance with XEP-0030, the category and type MUST be included)."

So they have to be unique.

m&m: right, "let's have a chat about this"

shall we take it to the list?

or is it already there? ;-)

the list discussion hasn't started yet

see also https://github.com/xsf/xeps/commit/bf110edfccddf0c22089df2861e0231bdb05c3a2#diff-5313e01f95bfb0963280ca4dfb497800L275

but this was in XEP-0115 earlier

this looks like the kind of thing that someone asked me about and we clarified it in two places

however, the constraint was already in XEP-0115 by then

now I'd need to go through the history of XEP-0115 too

probably not the best use of our time here :-)

probably not (-:

shall we move along to things we can settle now?

let's

If this is already covered in the general lore by 115 and it was just missing from 30, let me know and I'll remove my -1. I was concerned that it might be adding in multi-lang without discussion.

I do recall this being discussed for 115

BTW this xml:lang rule was not in version 1.4 of XEP-0115 and was added for version 1.5, published in February 2008

so we can look at the discussion leading up to 1.5

anyway … let's move on

of which there was much, I think - or at least lots of release candidate versions of 1.5 leading up to publication and approval in Feb 2008

nod

yep

next conundrum? ;-)

we've got action we can take for 115, which is to get back to Kev on details, then decide what's next

s/115/30/

https://github.com/xsf/xeps/pull/40

I believe this just needs to be merged into the next release

nope, I need to reach out to Kev and MattJ about reviewing the latest changes

I'll send that note out shortly

https://github.com/xsf/xeps/pull/41

this is ready to take to council

in the past, this would have been an 1.yrcX interim release

well, tmp publish

also BTW https://github.com/xsf/xeps/commit/6057a1fb42c18d1dddf07cdc0daa72696c327734#diff-5313e01f95bfb0963280ca4dfb497800 - that's where we made the change to 115

stpeter: you're having a tough time moving on today (-:

m&m: agreed on #41

I am - now I'm fascinating by what happened here :-)

I don't blame you … gone down a bunch of rabbit holes like this in the past myself

I've just been through (quickly) the 100ish post thread about 115. I'll remove my -1.

noted and thanks, Kev

Kev: yeah lots of discussion in January 2008! http://mail.jabber.org/pipermail/standards/2008-January/thread.html

for #41, I suggest we ask the council if they're good going to draft from the changelog

Mail sent.

and render upon request

trying to avoid publish snafus

on https://github.com/xsf/xeps/pull/63 I asked Lance to make some fixes - I can ping him about that

thanks

#82 is on hold, waiting on Mr. Wild

same with #83, actually (-:

#83 I need to sit down and rework; I just keep forgetting to do it.

Based on list feedback and discussion from last week.

it's still on my list to go through XEP-0198 proposals but I haven't carved out the time yet

so #84 on hold pending author

see above on needing to update a few Internet-Drafts :-)

right

also I'm way behind on a certain MUC/MIX discussion :-)

That one just needs you to say "Kev, make the changes" and watch while I fail to find time :)

I still haven't had time to read all of the changes to #91

failing to find time is a running theme

always :-)

Yes, sorry.

we're all guilty

Lance says he'll update XEP-215 per #63

cool

Unless we're desperately short of people I'll drop Council next year. Won't give me much extra time, but will cut back the amount of stuff I'm committing to and struggling with.

https://github.com/xsf/xeps/pull/94 LGTM as I replied there

Kev: yeah I might need to do some recruiting it seems

I'm the only person who's put a hat into the ring for anything ;-)

stpeter: agreed, just needs some massaging to get its editorial version

m&m ok good

#95 is pending a response from stpeter I think

yeah I _just_ replied

looks like #96 is also ready for merging

+1 on #96

I just took it over to track my failures (-:

#98 is just fixing examples, and stripping extraneous whitespace.

I think it's good to merge as an editorial revision

agreed

#100 needs some council discussion still, I think

need to make sure that makes it onto their next agenda

Can we go ahead and merge #99 so I can use it? It works fine (at least on Linux; might be able to install things with homebrew on macs). People can improve it later if they want to support their system more generically.

SamWhited: let me try it out … sorry, it kept slipping

I think it would be good to have a discussion on the list about deprecating stream compression entirely given the attacks that exist - or at least have a discussion about whether those attacks apply to XMPP, because if they do then deprecating stream compression would seem to be the right thing to do.

m&m: Many thanks :)

stpeter: agreed

or at least, "might" be the right thing to do

I'm not entirely sure these attacks are applicable to XMPP, given its stateful nature

but the discussion is good to have

I know we did a big security analysis when we implemented it; pretty sure the answer ended up being "as long as there are flushes on stanza boundaries on both ends it will be okay", but I'll see if I can't get the official report or discussion. Might have good info, I dunno.

s/flsuhes/full flushes/

that would be great, Sam

:)

thanks, Sam!

so, uh, #106 seems contested to me (-:

106 or 104?

gah, yeah 104

k

I merged 0016 and 104 in an odd way

<insert-evil-laugh-here>

I haven't looked at #106 yet

(FWIW)

Sam suggested that I need to post in the privacy lists discussion

Please and thank you :)

so, #106 is still pending author?

I think I suggested that Sam suggested that ... :)

indeed

I haven't been closely tracking that thread, but given that I've been the one slowly pushing forward with alternatives (invisibility and blocking commands), I think it's safe to say that I am not a huge fan of privacy lists - and we developed privacy lists originally to address a requirement that we misunderstood from RFC 2779 way back in the early days of the XMPP WG!

so I will post in that thread

(And so who really composed Beethoven's 5th?)

stpeter: I think that would be useful, thanks.

k

ok

so, #104 is pending more list discussion

#106 is pending author action

#108 just needs to be mergec

merged even

(with an editorial revision)

I need to depart for another meeting

I think we're done

thanks!!

and congrats on yet-another-RFC (-:

Thanks to our diligent Editors ;)

:)