I like the random aspect, but I think the penalty for anyone trying
it may lead to an indefinite postponement of them being released.
-mark
On Thu, 30 Aug 2001, Duane Cook wrote:
> This is the perfect solution to the problem.
>
> There is no need for any snapname gimicks
>
>
> At 04:38 PM 8/30/2001 -0400, you wrote:
> >Earlier I suggested a solution which would remove all incentive for
> >hammering the registry. Drop any name, at any time of the day or night, but
> >wait until it has been 5 minutes (or some other appropriate delay) since
> >anyone has tried to register it. Everyone would have an equal chance, and
> >those who try too often would never be rewarded.
> >
> >
> >----- Original Message -----
> >From: "Loren Stocker" <[EMAIL PROTECTED]>
> >To: "Robert L Mathews" <[EMAIL PROTECTED]>
> >Cc: <[EMAIL PROTECTED]>
> >Sent: Thursday, August 30, 2001 3:54 PM
> >Subject: Re: [No Registry Agents Available...]
> >
> >
> > > Hi Robert,
> > >
> > > Not a stupid idea at all. I've suggested this myself, only I don't think
> >you
> > > need to wait more than 15 minutes unless you drop them all at once like
> >they
> > > may have today (24 hours might be necessary after this mess!).
> > >
> > > The fact is that the great domains are snapped up within miliseconds --
> >the
> > > leftovers could then be released for general consumption.
> > >
> > > The other issue is dropping as many as they did all at once. Verisign is
> > > ASKING for it.
> > >
> > > Best, Loren
> > >
> > >
> > > Robert L Mathews <[EMAIL PROTECTED]> wrote:
> > > I note that Verisign's new system to prevent hammering of the registry
> > > during drops wasn't working around 2:15 EDT:
> > >
> > > No Registry Agents to service domain: [rfertefdfs.com]
> > >
> > > Although I can't tell who has all the connections, this makes me even
> > > more suspicious of domain back-ordering services that open many
> > > connections. Obviously some registrars are doing automated batch
> > > processing in the overflow pool (which is supposedly not allowed), and
> > > some OpenSRS resellers are hammering OpenSRS (since the guaranteed
> > > OpenSRS pool is full).
> > >
> > > Can I ask what OpenSRS is doing to solve the latter problem of OpenSRS
> > > resellers using too many connections? If resellers do more than a certain
> > > number of lookups in a given time period, are subsequent lookups
> > > restricted to the registry automated batch pool for a while? (They should
> > > be.)
> > >
> > > BTW, here's my (probably stupid) idea to solve the registry problem: for
> > > the first 24 hours after a name is released, it should only be able to be
> > > registered through the automated pool, and not the guaranteed or overflow
> > > pools. Then the feeding frenzy would occur only in the automated pool.
> > > The next day, the domains could be registered through any pool. In this
> > > scenario, registrars would allow speculators to set a flag saying they
> > > want to try the automated pool, and speculators could have at it... while
> > > the rest of us would have normal connections.
> > >
> > > Even better would be some sort of first-come, first-served back ordering
> > > system, implemented by the registry, as someone else described. Not
> > > likely in the near future, I guess.
> > >
> > > --
> > > Robert L Mathews, Tiger Technologies
> > >
> > >
>
--
mark jeftovic
http://www.easydns.com
http://mark.jeftovic.net
