Dear all, Per Don¹s email (below), the proposed agenda for the next WG call on Tuesday 9 December is: 1. Roll Call/Updates to SOI 2. Finalize WG recommendations on Category E ­ Relay (see attached document and Don¹s notes, below) 3. Review WG recommendations on Category G ­ Termination (see attached document ­ potential recommendations/questions for WG highlighted in blue) 4. Next steps Thanks and cheers Mary Mary Wong Senior Policy Director Internet Corporation for Assigned Names & Numbers (ICANN) Telephone: +1 603 574 4892 Email: mary.wong@icann.org From: Don Blumenthal <dblumenthal@pir.org> Date: Sunday, December 7, 2014 at 10:15 PM To: "gnso-ppsai-pdp-wg@icann.org" <gnso-ppsai-pdp-wg@icann.org> Subject: [Gnso-ppsai-pdp-wg] Issues left from the last couple of weeks

Some issues that remain from D and E. I hope that we can resolve them on the list and then during drafting rather than take call time.

D. Transactional/nontransactional definition Steve formulated a definition on the call two weeks ago, subject to comment from WG members who support the p/p eligibility categories. I haven¹t seen any response so here it is again from the transcript:

³[D]omains used for online financial transactions for commercial purpose should be ineligible for privacy and proxy registrations.²

E. We focused on Relay last week but didn¹t get as far as we had hoped because of audio issues. I copied language from the template and put additions based on where I think we wound up in bold. Please take a look and send comments on the issues as well as my language.

A P/P service provider [should] [must] notify the requestor upon becoming aware through a timely affirmative notification of a persistent failure that delivery of the original form of electronic communication has failed. Notification may be by means of a technical mechanism or other means that relays notice of the delivery failure to the requestor and/or provides the requestor with alternate ways to escalate the issue.

As part of an escalation process, and provided there has been a certain minimum number of affirmative persistent delivery failures of an electronic communication within a certain specified time frame [TBD], the provider [should] [must] upon request forward a further form of notice to its customer. A provider should have the ability to select the most appropriate means of forwarding (including to account for issues of cost) [and to charge a reasonable fee] [and any such reasonable fee is to be borne by the customer and not the requestor]

Additional questions from the first paragraph: 1) I was asked to focus on the forward may = disclosure issue that I already send a note about.

2) Should vs must in the first line. I was tempted to bold ³must² but am noit sure that I interpreted something correctly.

Additional questions from the second paragraph: The audio issues were a particular roadblock. We still need to address: 1) Must vs should

2) Cost recovery

3) Who pays if there is cost recovery.

A vigorous chat session helped with respect to the audio challenges. Kathy circulated it once but It¹s here also for convenience.

Talk to you on Tuesday.

Don

Adobe Connect chat transcript for Tuesday 02 December 2014:

Terri Agnew:Dear all, welcome to the PPSAI call on the 02 December 2014

Stephanie Perrin:Wow, I am the first one. Mark this on the calendar!

Theo Geurts:we shall remember

Theo Geurts:good afternoon all.

Chris Pelling:afternopon all

Chris Pelling:afternoon *

Graeme Bunton:Good morning all

Bladel:Good morning.

Bladel:Who is "D"?

Graeme Bunton:someone is a little heavy on the breathing

D:"D" is Don B. I'll fix in a minute.

Bladel:Ah, ok. :)

Graeme Bunton:Sounds ok to me

Sarah Wyld - Aplus.net:Thanks Terri!

Kiran Malancharuvil:Its okay to me

Chris Pelling:its breaking y up here in the uk

Michele Neylon:I'm hearing stutter on both lines

Sarah Wyld - Aplus.net:James sounds choppy to me as well.

Holly Raiche:Yes

Sarah Wyld - Aplus.net:graeme, sorry.

Kiran Malancharuvil:You're fine on adobe connect and bad on the audio bridge

Bladel:James isn't speaking. :)

Chris Pelling:I rang in and both sound like daleks

Susan kawaguchi:dialed in and both are choppy

Alex Deacon:sounds loud and clear on the adobe

Osvaldo Novoa:Hello all

Michele Neylon:I'm dialled in nad it's bad

Michele Neylon:the lady speaking is fine

Michele Neylon:not sure who that is

Michele Neylon:Mary maybe?

Chris Pelling:agree with michele

Christian Dawson:Agree - fine on adobe connect and bad on the audio bridge

Michele Neylon:everyone else sounds terrible

Mary Wong:@Michele, that's Terri with the soothing clear speaking voice.

Christian Dawson:Michele is that a value judgment?

Carlton Samuels:Morning all

Michele Neylon:Mary - so Terri is a woman?

Michele Neylon:I didn't know that

Chris Pelling:don sounds non existant now

Michele Neylon:and now we have silence

Chris Pelling:yes terry

Carlton Samuels:Yes, loud and clear

Chris Pelling:don is crackling

Michele Neylon:Don's line is bad

Alex Deacon:don sounds great on the adobe

Michele Neylon:Graeme try please

Kiran Malancharuvil:Don needs to call back into the audio bridge

Carlton Samuels:I'm hearing Don very clearly!

Bladel:Graeme is also breaking up., I think.

Holly Raiche:The rest of us are fine - Don's still fuzzy

Mary Wong:@Graeme, that seems to be right.

Kathy:Full House!

Don Blumenthal:still working on connection

Bladel:Graeme is also a bit choppy.

Bladel::)

Carlton Samuels:Hi Terri, Please record me present and count

Mary Wong:@Carlton, thank you - we will.

Michele Neylon:I'm connected via both phone and adobe

Christian Dawson:Just switched to Adobe only it's much better today

Terri Agnew:@ Carlton, will do

Michele Neylon:switched to adobe

Michele Neylon:audio is clearer

Holly Raiche:AGree with Michele - the audio on adobe is fine - the dial in is fuzzy

Chris Pelling:someone now has an echo

Terri Agnew:apologies everyone for the choppy audio on telephone. We are working on trying to get this to clear up

Michele Neylon:I've muted my microphone so it's not me

vicky sheckler:apologies - i need to drop off

Chris Pelling:I cant hear anything on adoibe :/

Chris Pelling:so will listen to the crackle and make best do

Terri Agnew:adobe seems to be the best option at the moment

Don Blumenthal:I'm back on line.

Graeme Bunton:Great

Mary Wong:Please note that the options (still under discussion) have been placed in square brackets in the current text.

Volker Greimann:apologies, previous meeting was running into overtime

Carlton Samuels:Yes we have control

Holly Raiche:We have control of the screen

Kathy:We may still be in our turkey-induced coma (for those who celebrated Thanksgiving)

Stephanie Perrin:Graeme while your technique is excellent, it might be rash to move on to the next item yet...:-)

Kathy:No, can't hear Steve

Kathy:yes

Chris Pelling:I can hear steve

Carlton Samuels:Yes can hear Steve

Paul McGrady:Sorry I'm late.

Michele Neylon:can't hear him

Michele Neylon:oh I can now

Chris Pelling:this is very flakey today

Chris Pelling:I cant hear on adobe connect Steve

Chris Pelling:but I can on the phjone

Chris Pelling:Just because there is no response, does not show contact issue

steve metalitz:Is there objection to "must" in the first paragraph?

Chris Pelling:thus even if you pay for it, its not to get a guranteed response

Holly Raiche:Can't hear very well

Kathy:@Steve, yes.

Chris Pelling:on 1 Steve it should be stating Electronic

Kiran Malancharuvil:+1 Steve

David Cake:so Steve, you are talking about a hard bounce in email, or equivalent?

steve metalitz:old hand

Kathy:@All, yes being that "should" should be kept as "should" in the first sentence

Bladel 2:Thanks for clarifying,. Can we get some language inserted in to the text somewhere so it is clear?

Chris Pelling:Kathy is clear

Holly Raiche:Isn't the problem if the message is sent - and may be received, but is not responded to . Aside from a hard bounce, could a registrar p/p KNOW the email has or has not been sent

Mary Wong:All, the WG agreed previously to NOT use language like "hard " or "soft" bounces.

David Cake:Did we? What language did we agree to use instead?

Luc Seufer:twerking was porposed if memory serves

Mary Wong:@David, that is what our notes show. Hence the chairs/staff suggestion in Cat E when it was first circulated, about delivery failure.

Holly Raiche:It is still hard to hear on a dial out

Chris Pelling:But not all servers do that

Carlton Samuels:There are standard repsonses for email! Maybe we should eliminate the ones we don't mean

Kathy:@Alex, I hate to ask, but couldyou summarize your comments in the Chat?

Chris Pelling:Not all servers are setup to rspond like that

Kathy:You are breaking up a bit...

Philip Corwin:The phone bridge audio is atrocious. Brittle, crackling, fading in and out with an intermittent echo.

Michele Neylon:if the wording is broad enough to alllow for the technical realities of some mail servers being dumb ..

Kiran Malancharuvil:Not technically difficult from MarkMonitor's perspective

Kiran Malancharuvil:Agree with Alex's point

Michele Neylon:Kiran - how many mail users do you host?

Alex Deacon:@don - agree that should be the focus.

Terri Agnew:as a reminder, please mute microphone and telephone when not speaking. We are still troubling shooting the audio issue

Kiran Malancharuvil:Good question, can relay back to Matt Serlin who gave me quick answers rather than detailed analysis.

Holly Raiche:Isn't the suggestion that if the p/p becomes aware that the message has not been delivered, they MUST use another method of communication

Michele Neylon:+1 to James

Carlton Samuels:@Don: Just say that - 'when there is an affirmative notice that an email has not reached the intended address'

Michele Neylon:we often see issues with mail being sent to Hotmail / Gmail etc.,

Michele Neylon:and you often don't get back any useful errors

Holly Raiche:@ Carlton - agree with suggestion

Susan kawaguchi:only getting every other word that james is saying

Carlton Samuels:@Michele: Then there is no affirmative notice so we do the next thing after time t; escalate

Chris Pelling:dropped phone bridge

Stephanie Perrin:Agree with James. Elegantly put.

Bladel 2:Dang. I even dropped off the phone bridge to fix audio issue.

Holly Raiche:@ James - could you put something in the chat - you were hard to hear

Bladel 2::(

Christian Dawson:It's not you James it's the bridge in general

steve metalitz:@James what about Carlton's proposal: 'when there is an affirmative notice that an email has not reached the intended address'?

Christian Dawson:By the way, +1 to your comments, James

Bladel 2:The simplest summary: Reliability of detecting delivery failure breaks down as scale increases.

val sherman:I think we are talking about situations where the Provider is aware of repeated delivery failures. One other point: since the Customer is required to provide a working email by the terms of service, would it not be in violation of the terms if there is repeated failure of delivery that the Provider is aware of? If so, might an additional option for the Provider be to terminate the service?

Holly Raiche:What is the conclusion please

Holly Raiche:@ Val - I was thinking along those lines

Bladel 2:@Steve: that works, if we can clarify in the text.

Stephanie Perrin:James said that costs should follow the benefitting user.

Kathy:@Mary, I think this is this is the time for some clarification

Kathy:It seems to be warranted at this point in the discussion...

Holly Raiche:@ Kathy - agree - could we clarify please

Mary Wong:The idea for the notification is that it is triggered when the provider "becomes aware".

Stephanie Perrin:From a consumer protection standpoint, to avoid harrassment, there should be some costs accruing to the party making the allegations of abuse.

Kathy:@Mary: that seems very loose at this point in time

Kathy:it is not bounded by time, system response, etc

Chris Pelling:+1 Stephanie

steve metalitz:@Don, yes, provider should offer alternative upon request after specified number of hard bounces.

Kristina Rosette:+1 to Steve

Bladel 2:It could be offered as a premium service.

Susan kawaguchi:If the allegations of abuse are found to be accurate would you advocate the registrant then be charged for the communication

Holly Raiche:@ Steve - isn't he problem that there may not be 'hard' bounces?

Graeme Bunton:Lots of questiosn to addresss in the above

Graeme Bunton:Lots of questiosn to addresss in the above\

steve metalitz:@Don I was responding to the question you posed.....

Mary Wong:Just for clarity - are we going back to the "hard bounce" language?

Luc Seufer:do we have a definition for it?

Stephanie Perrin:If I put someone who is bugging me into my spam filters, do you define that as a hard bounce? Innocent (read naive) question.

Carlton Samuels:@Don: The rule should compel a declared form of communication for the provider. It is the responsibility of the communicators to keep that communication line open and available. We should just say if the response is not fortcoming within a certain time then next step is the defined escalation for communication to be effected. That then becomes the cost of the beneficial user, this time the requestor.

Michele Neylon:Stephanie - your spam filter probably wouldn't bounce

Chris Pelling:if the bridge worked we might

Susan kawaguchi:can barely hear anything on audio

Holly Raiche:@ Stephanie - agree

Mary Wong:@Luc, that was one of the issues when the language was first discussed, including at the WG F2F in LA.

Darcy Southwell:The audio is very hard to follow

Luc Seufer:Or the preferred way of communication of lawyers: sending 20 meg attachements, does this constitute hard bounce?

Bladel 2:I'm starting to think we abandon this call. :(

Susan kawaguchi:should we just go to email and forget the call

Holly Raiche:Actually in most cases

Kathy:@Susan: impossible tracking problem for registrars. how long would it take until the abuse is proven - and then if it is appealed?

Chris Pelling:also as Luc pointed out

Chris Pelling:A lot of mail servers have message limits

Susan kawaguchi:@Kathy so why punish the requestor only when most requests are very targeted to domain names with bad behavior

Stephanie Perrin:@Michele, ok so then the message is received but tucked into my spam filters. What does the upset sender do then and is there any reason why the service provider should pick up the cost of sending by registered mail.

Susan kawaguchi:dropping off call will follow in chat as I cannot understand anything on audio

Michele Neylon:Stephanie - if it doesn't bounce then I don't know about it ..

Luc Seufer:@Mary thanks so we don't

Terri Agnew:Apologies everyone, we are working with Tech Support but not having much luck to clear up audio

Stephanie Perrin:@Michele but you will hear from the sender again....at which point you tell them the message was received. Go away. Correct?

Holly Raiche:@ Dave - I think the issue won't go away until it is clear what amounts to a p/p provider becoming aware

Kathy:@Susan, we were talking earlier (original discussion) about charging a small/reasonable charge to the requestor who wants followup. The analogy was the per-page cost for faxes or copies that law firms charge their clients. That's a charge my clients pay regardless of whether they are ultimately in the right -- or not.

Kiran Malancharuvil:There is quality discussion happening in the chat, would prefer to move this to email

Holly Raiche:@Don - look at the chat for discussion

Michele Neylon:Stephanie - the email has been sent as far as I'm concerned. If you're not getting it due to yuour spam filter I can't know that

Graeme Bunton:Absolutely, carry on email

Michele Neylon:So yeah - "go away"

Christian Dawson:We should also keep in mind that a hard bounce is sometimes a very temporary issue. As a hoster I have 1.4 million domains under my control. My customers go over their disk space or don't pay their bills aoccasionally and hard bounce for an hour, a day, a week. . They just bounce temporarily in a nonmalicious, unintentional way.

Michele Neylon:+1 Christian

Kathy:@Mary and Terri: perhaps a special note to the List that the Chat had a lot of important discussion due to problems with the bridge?

Alex Deacon:@christian - you are describing a "soft bounce" (i.e. one that may clear up at some point in the future)

Stephanie Perrin:It would take only five minutes to read the chat into the transcript.

Holly Raiche:Too hard to hear

Bladel 2:Thanks, Don and appreciate everyone making the effort.

Susan kawaguchi:reading it won't help if you can't hear

Kiran Malancharuvil:We should adjourn so that we don't miss anything

Frank Michlick:no problems hearing via adobe connecgt

Frank Michlick:-g

Holly Raiche:Barely

Christian Dawson:Just goes to show we definitely need definitions and explanations.

Kathy:+1 Christian

steve metalitz:@Stephanie, no it is allocating the cost for failure to give the provider an e-mail address at which you can be reached.

val sherman:+1 Michele. Yes, Stephanie -- Message received but ignored OR message caught by spam filter is not a delivery failure. Delivery failure is not the same as failure to respond.

Holly Raiche:I don't think the issue is cost so much as working through what is meant by becoming aware - once we have done that, then we can talk about cost

Chris Pelling:totally agree

Holly Raiche:The chat is working well

Mary Wong:@Holly, yes that was the idea behind this language, particularly following the hard v soft bounce discussion some time ago.

Chris Pelling:cant hear anything

Kathy:I thought we had narrowed the "non-response" issues down to technical ones...

Chris Pelling:Is Steve talking ?

Stephanie Perrin:Yes Steve is talking

Kathy:yes Chris

Holly Raiche:@ Kathy - agree -

Chris Pelling:ok, well cant hear him here in :(

Chris Pelling:Im listening to adobe connect only now

Holly Raiche:Go to Val's earlier comment -

Christian Dawson:Steve, email is not reliable enough for that to be at all practical.

Don Blumenthal:Steve is talking. To summarize, hard bounce = bad address. P?P registrant should pay

Mary Wong:Steve is comng across clearly on the audio bridge, other speakers not so much. Apologies, all - we are having IT look into this. We've never had this problem to Terri's and my knowledge.

Chris Pelling:ok, sorry, disagree

Chris Pelling:for so many reasons, but the main one is not all servers are setup to reply with a hard bounce

Chris Pelling:I can hear Don aok

Stephanie Perrin:So we need a volunteer to rewrite this so that it is clear, without using the word hard bounce.

Chris Pelling:But Don we might not get ANY response so as far as we know it has gone

Chris Pelling:perfect

Kathy:who is talking?

Chris Pelling:I think Christina

Luc Seufer:I think it's Chrisitian

Holly Raiche:@ Stepanie - and Kathy - we need language to agree with wha are now technical descriptions of the difficulties is KNOWING if he message has reached its desination

Chris Pelling:Christian *

Mary Wong:The idea of conditioning provider notification upon a provider actually getting a notice of a bounce was discussed but rejected including for reasons that Christian is describing.

Don Blumenthal:To my world? Yeah, I can see why ICANN might be seen as being an alien conspiracy.

Kathy:@Holly, agree

Chris Pelling:soft is temorary where a server will try every 4 hours, whereas hard bounce is permanent delivery failure

Kathy:@Christian, I think we should use the technical terminology -- and define it very clearly as well.

David Cake:That terminology sounds OK.

Kathy:@Don: "timely, affirmative notice of nondelivery"?

David Cake:We should not use hard bounce because that is quite specific to email

Holly Raiche:I like Kathy's suggestion

Chris Pelling:might be useful for a read : http://en.wikipedia.org/wiki/Bounce_message

Stephanie Perrin:How would one define "permanent Non-delivery"

Chris Pelling:I cant

Chris Pelling:cant hear anyone a

Chris Pelling:Don I could hear

Don Blumenthal:Timely affirmative notice of permanent nondelivery?

Frank Michlick:bounces aren't always realtime

Holly Raiche:Thanks Chris - we may need to go to that language

Chris Pelling:define timely

Frank Michlick:some include temporary and multiple delivery attemptsxz

Frank Michlick:-xz

Chris Pelling:Don = loud and clear

Stephanie Perrin:My question is how does my provider know that non-delivery is permanent?

Holly Raiche:Maybe we can use the list to define things like bounce and timely

Luc Seufer:ICANN definition of timely should be funny

Carlton Samuels:@David: I'm trying to make a distinction between what happens when we know a communication did not reach its destination and when there is no actionable evidence. The first instance is addressed by the language 'when there is an affirmative.....'. With no actionable evidence of delivery, then we might have a provision that says what next. My what next is that there is a time factor that should be invoked and once that runs out, a definite response required by the provider. If the customer is AWOL then compel cancellation.

Chris Pelling:Stephanuie, if your server company set the server up correctly you SHOULD get a response stating that the message had a permanent delivery failure

Kathy:I just can't see it being a permanent liability for proxy/privacy providers...

Chris Pelling:but its down to the company that set it up

Luc Seufer:the launch of the new gTLD program was operated in a timely manner

Chris Pelling:cant hear anything

Chris Pelling:Luc - its still going :p

Chris Pelling:and what are we up to now, a year ? :p

Don Blumenthal:Luc, applicants might argue that "timely" point. :)

Holly Raiche:@ Mary and Don - try to capture the language that is being suggested in the chat

Chris Pelling:no one talking ?

Don Blumenthal:Mary is

Chris Pelling:ok I can hear Don not Mary

Chris Pelling::(

Kathy:Tx Don - great leadership under rought circumstances!

Kathy:Tx All!

steve metalitz:thanks Don

Holly Raiche:Tks