ICANN/DNSO
DNSO Mailling lists archives

[council]


<<< Chronological Index >>>    <<< Thread Index >>>

[council] RE: Registrars Questions for VGRS Regarding the revised WLS proposal


Rick,

VGRS' responses to questions received regarding the revised Wait Listing
Proposal have been posted at:

	http://verisign-grs.com/wls_responses.pdf

Chuck

> -----Original Message-----
> From: Rick H Wesson [mailto:wessorh@ar.com]
> Sent: Friday, February 08, 2002 8:01 PM
> To: Gomes, Chuck; wls@verisign.com
> Cc: council@dnso.org; Registrars List; halloran@icann.org; 
> Louis Touton;
> execs@registrars.dnso.org
> Subject: Registrars Questions for VGRS Regarding the revised WLS
> proposal
> 
> 
> 
> Chuck,
> 
> Please find attached .PS and .PDF files which contains lists 
> of questions
> the Registrars would like answered as requested in your 
> amended proposal
> dated January 28, 2002.Please use the following address as a point of
> contact for our group: execs@registrars.dnso.org which we 
> will forward to
> the Registrars list.
> 
> A text copy of the attached document is below for easy 
> cut/paste in e-mail
> discussions.
> 
> Regards,
> 
> -rick
> 
> Rick Wesson
> CTO, Registrars constituency
> 
> 
> 
> 
>    February 8, 2002
> 
> 
>    To Chuck Gomes,
> 
> 
>    The Registrar Constituency (RC) is hereby responding to 
> the VeriSign
>    Global Registry Services (VGRS) regarding its revised proposal to
>    manage a Wait Listing Service (WLS), the subscription service for
>    deleted domain names.  After issuing its original WLS proposal on
>    December 30, 2001, VGRS issued a revised proposal on January 29,
>    2002.  VGRS allowed ICANN constituencies and other 
> interested parties
>    to submit comments and questions regarding the revised WLS.
> 
> 
>    Having opposed the original WLS proposal (see January 18, 2002 RC
>    letter to VRSN) and considered the revised proposal, the RC remains
>    significantly concerned with this proposal.  Therefore, 
> the RC opposes
>    the revised WLS proposal in its current form.  Following 
> are the RC's
>    key concerns regarding the WLS, which focus on price and 
> transparency:
> 
> 
> 
>    a) We continue to view the proposed $35.00 price point for WLS
>       (which is in addition to the $6.00 registry fee) as exorbitant,
>       even with the potential proposed rebate program.  VGRS has never
>       justified its price point with cost requirements.  The fact that
>       VGRS easily lowered its original price begs the question of what
>       costs actually justify the proposed WLS price.
> 
> 
>       As we had previously stated, the WLS would significantly raise
>       the price point for the end consumer, effectively undermining
>       competitive registrars' financial wherewithal.  It is highly
>       unlikely that registrars would be able to increase their margins
>       in proportion to the increased margin charged by VGRS.  In
>       contrast, competitive registrars would have to dramatically
>       lower, or eliminate, their current margins in order to compete
>       for WLS names.  This would undermine competitive registrars'
>       revenues and jeopardize their ability to remain profitable.
> 
> 
>       The one registrar that may be able to take effectively advantage
>       of this price is the VeriSign registrar, which continues to
>       enjoy the largest market share. It would be able to use the new
>       higher margin to price below wholesale, as it has in the past
>       with the $6.00 fee.  The result is to unfairly undermine
>       competitor registrars.
> 
> 
>    b) The additional concern that continues to exist with the revised
>       WLS proposal is regarding the lack of transparency if VGRS runs
>       the primary registry, the largest registrar, and the
>       subscription service.  As long as the same company is operating
>       this vertically powerful chain of companies, it may be possible
>       for it to shift domain names from the $6.00 registry to the
>       $41.00 WLS.  In fact, only the registry would know all of the
>       WLS subscriptions and the timing for deleting names.  Such
>       information could be abused by its registrar.  Considering that
>       there is a history - some of it still unresolved - of VeriSign
>       not deleting expired names, the RC is doubly concerned that
>       VGRS' operating the WLS provides new opportunities for domain
>       name hoarding.
> 
> 
>    c) An additional inherent unfairness is the ability of the largest
>       registrar to "game" the WLS system.  Since VeriSign's registrars
>       delete over 50% of domain names, they can offer (to potential
>       WLS subscribers) a WLS subscription on its customers' names that
>       only VeriSign knows are to be deleted.  Other registrars'
>       customers would be buying a WLS name without the benefit of
>       knowing that a particular name will actually be deleted.  The
>       current system does not provide registrars an advantage based
>       its size.  With the WLS system what is to prevent a registrar
>       extending its advantage over the other registrars by not
>       deleting the names that have expired without renewal and thereby
>       offering less risky WLS subscriptions on those names compared to
>       the other registrars?
> 
> 
>    In terms of the RC's questions, they are as follows:
> 
>    a) What are the costs that justify the WLS price? What is the
>    intellectual property that SnapNames is providing?
> 
>    b) The introduction to the VGRS' "Justification" document states
>    that the WLS "is not a solution for the deleted names issue."  If
>    the WLS system does not solve the batch pool problem (per VGRS
>    explanation), why should it be adopted?
> 
>    c) What is the phase in procedure (ie: landrush)?
> 
>        a.  The SnapNames Parallel Registries Proposal (Sept 21, 2001)
>        identifies 25,000 deletions/day. This means that the annual
>        demand on deleted names is roughly 4.5 million.  It is very
>        likely that there will be a landrush. Since VGRS could not
>        handle the load of 160,000 domain release back in Aug ust 2001
>        (which led us to today's condition), how will it deal with a
>        WLS landrush?
> 
>        b.  There will be competition amongst speculators to be the
>        first to get the WLS on the best names about to be deleted,
>        which may extend the landrush effect.
> 
>    d) What is the proposed length of the trial?
> 
>        a.  Page 8, Paragraph g of the proposal states, "Subscriptions
>        continuing beyond the end of the trial period would continue to
>        be serviced by VGRS and registrars."  Thus, the trial period
>        lasts for two years, not one.
> 
>    e) How many expired names does VGRS Registrar have that have not
>    been deleted?  When will they be deleted?  Will these names be
>    deleted before the start of the WLS?
> 
>    f) Pages five and six of the proposal discuss registration data to
>    be submitted to the WLS, which suggest potential added data.  Would
>    any data in addition to the current Whois information be required?
> 
>    g) The proposal allows registrars to delete a domain with full
>    refund if it is less than 120 hours old.  Is there a grace period
>    for deletion (cancellation) of a WLS subscription, which will
>    refund the fee?
> 
>    h) How would VGRS handle "charge-backs" for subscriptions?
> 
>    i) What are the criteria for a evaluation of a successful testbed?
> 
>       a.  These should be specified prior to launch.
> 
>       b.  Do those criteria take into account the existing competitive
>       landscape?  What are market measurements to serve as the basis
>       for a comparison?
> 
>       c.  Under what metrics will the WLS test be considered a
>       failure?
> 
>    j) What are the technical impacts of the protocol (epp) ?
> 
>    k) Does the WLS proposal expose current registrars of record to
>    different and/or additional risks than other registrars?
>    Alternatively, do registrars of record have any advantage over
>    other registrars?
> 
>    l) Many have requested an opt-out capability.  Does the protocol
>    used for registering a WLS subscription have this capability, if
>    not why could this not be added?
> 
>    m) Registrars have requested several improvements and fixes of
>    registry operations in order to address problems that are impeding
>    registrations.  When will the Registry fix these issues and why has
>    the WLS taken precedent above fixing these issues?
> 
>    n) Why has VGRS not implemented batch deletions in several weeks?
> 
>    o) Has VGRS investigated the antitrust, auction law and commodity
>    futures law implications of WLS?  If it is deemed to be illegal or
>    raise civil law implications, will VGRS indemnify affected
>    registrars, resellers, registrants and other market participants?
>    Does VGRS offer any assurances from legal staff that the WLS
>    product is legal?
> 
>    p) Will VGRS release the actual proposed service agreement for
>    comments at some point before a decision to deploy the WLS is made?
> 
>    q) Will VGRS release criteria for what constitutes an acceptable
>    education program for the WLS service?
> 
>    r) Have financial projections been created for the WLS, for both
>    the registry and the registrars? If so, we would like to see them.
> 
>    s) When will technical and operational documents of the WLS be made
>    available?
> 
> 
>    While the RC continues to oppose the WLS in its current form, it
>    recognizes the need for a permanent solution to the apparent
>    problem of deleted names not being released or being released in a
>    manner that undermines other registry functions.  Therefore, the RC
>    welcomes the Names Council's consideration of alternate ideas for
>    addressing these issues, many of which have been discussed by the
>    RC.
> 
> 
>    Regards,
> 
> 
>    Rick Wesson
>    Registrar Constituency
>    Chief Technical Officer
> 
> 
>    cc:  Louis Touton
>         Dan Halloran
>         Names Council
> 
> 


<<< Chronological Index >>>    <<< Thread Index >>>