44/8

Owen DeLong owen at delong.com
Wed Aug 28 15:54:39 UTC 2019



> On Aug 27, 2019, at 23:50 , Bryan Fields <Bryan at bryanfields.net> wrote:
> 
> On 8/27/19 11:52 PM, Owen DeLong wrote:
>>> On Jul 26, 2019, at 21:59 , Doug Barton <dougb at dougbarton.us> wrote:
> 
> <snip>
> 
>>> and I have known two of the ARDC board members and one of
>>> the advisors listed at https://www.ampr.org/amprnet/
>>> <https://www.ampr.org/amprnet/> for over fifteen years. I consider them
>>> all friends, and trust their judgement explicitly. One of them I've known
>>> for over 20 years, and consider a close and very dear friend.
>>> 
>>> There have been a number of points over the past 30 years where anyone
>>> who genuinely cared about this space could have used any number of
>>> mechanisms to raise concerns over how it's been managed, and by whom.
> 
> <snip>
> 
> I will say most people ignored them, as TBQH, nothing changes in amateur radio
> until people die.  ARDC finally allocated space, got reverse DNS working, and
> basically did nothing else.  It was a technical org doing nothing of any real
> value.
> 
> I had brought up the issues of governance numerous times, and said it didn't
> look right to have people on the board with conflicts, or even licensed
> amateurs.

I confess I never saw any of the questions about governance prior to the announcement
of the sale. If I had, I might have been spurred to look closer at least.

>  There were other personnel issues brought up as well and no action
> was taken.  If was an org doing the bare minimums and we got what we needed
> from it, so why rock the boat?

Well, there’s also the fact that the people involved in the org that I was aware of were
people that I trusted and would not have expected to take radical action absent some
outreach and consent of the community.

>> However, most of my anger is around the fact that: 1.	It never in a million
>> years would have occurred to me that these people who I also consider
>> friends and also trust explicitly would take this particular action without
>> significant prior (and much wider) consultation with the amateur radio
>> community.
>> 
>> 2.	I believe this was done quietly and carefully orchestrated specifically
>> to avoid any risk of successful backlash by the time the community became
>> aware of this particular intended action.
> 
> Bingo.
> 
>> If you want to say shame on us for trusting these people and not noticing
>> the severe corporate governance problems with ARDC until they took this
>> particular action, then I suppose that’s a fair comment.
> 
> Many know these people, and you cannot let that acquaintance cloud your
> judgment here.  If these were people you did not know and they did this, you'd
> call it what it is.  If an acquaintance does the same action, is it not the same?

An acquaintance is different from a trusted colleague and friend. While I don’t
have particularly close relationships with the parties involved, I did consider them
trusted colleagues and friends.

This doesn’t cloud my judgment about calling this action what it was, but it did
cloud my judgment in terms of anticipating this action. When we trust people,
we don’t expect them to act outside of certain bounds, as I did not expect that
these particular people would act in this particular manner.

It is a denial of human nature to claim that we can not allow our trust to cloud
our judgment when it comes to expectations of how people will act.

> Does it pass the smell test that 44/8 was used, with no benefit to ARDC by
> CAIDA?  This use held back deployment of 44/8 for years by the amateur users.
> Does it smell funny that the majority of the board members of ARDC were CAIDA
> board members?

It does pass the smell test, actually, when I understand the manner in which it was
“used”.

CAIDA is not some evil spammer researching ways to infiltrate more networks or
hide more snowshoeing. They are a public benefit research organization that has
provided a lot of valuable information to the internet over the years.

Their “use” of 44/8 was limited to a passive feed of otherwise unroutable packets
destined for unregistered addresses within 44/8. Since they were adjacent to
the router where this was occurring. I’m not sure how or why you claim that
this held back deployment for years. You’ll need to provide more information
and/or evidence to support that claim. At the moment, as I understand things,
that claim doesn’t pass the smell test.

>>> So let's talk a little about what "stewardship" means. Many folks have
>>> complained about how ARDC has not done a good job of $X function that
>>> stewards of the space should perform. Do you think having some money in
>>> the bank will help contribute to their ability to do that? Has anyone
>>> looked at how much of the space is actually being used now, and what
>>> percentage reduction in available space carving out a /10 actually
>>> represents? And nowadays when IPv6 is readily available essentially "for
>>> free," how much is the amateur community actually being affected by this?
>>> 
>>> 
>> All of those are good questions. I don’t have data to answer any of them
>> other than that removing a /10 from a /8 is obviously a 25% reduction in
>> the total space, so clearly a somewhat larger (though I don’t know by how
>> much) reduction in available space since available space is some fraction
>> of the remaining 1.5 /9s.
> 
> Based on the way this was handled we suspect it was part of an unsolicited
> offer by the buyer.

I have no doubt that Amazon (and likely a broker acting on behalf of Amazon)
probably approached the ARDC board to initiate the sale, but it’s pretty clear
that ARDC took multiple deliberate steps over years that I suspect preceded
any such approach that were clearly aimed at making this possible.

> If an organization decided to sell off 1/4 of it's assets and start a
> charitable giving process, the first thing done would be define the charitable
> giving areas and process.  Get your house in order, build up a board full of
> talented people aligned with this new mission, and then effect the sale, no?
> 
> Considering ARDC is only giving to IRS approved 501c3 organizations, the sale
> of the part of the space dedicated to non-US use (44.128.0.0/9) doesn't seem
> right.  Seems like if you're going to sell space not for use in the US, you
> should have a plan to benefit those who it was taken from, no?
> 
> The first inkling of any of this was reverse DNS for 44/8 users was broken.
> Mail was rejected, and people started to ask questions.  There was no
> consultation of anyone with technical clue on this.  Had there been, we could
> have prevented a 5+ day outage.
> 
> I think we all get that '44.in-addr.arpa. NS $SERVERS' would need to move to
> the next byte boundary in configuration.  Sure, it's 192 new records that need
> to be made at ARIN, and even if they can't script it, it's an hour or two of
> typing in the ARIN portal.

I would have expected at least one ARDC board member in particular to have
significantly more DNS clue than the average DNS administrator.

> Based on the absolute secrecy this was done with and the lack of process or
> thinking that went into it, I have to think it was an unsolicited bid, and
> likely negotiated by ARDC personal lacking real-world business savvy.

That would not surprise me.

>>> And with all due respect to Jon (and I mean that sincerely), what did
>>> it/does it really mean that "Jon gave $PERSON the space for $REASON" 30
>>> years later? Jon was a brilliant guy, but from what I've been told would
>>> also be one of the first to admit when he made a mistake. One of which,
>>> and one that he actively campaigned to fix, was the idea of classful
>>> address space to start with, and particularly the idea that it was OK to
>>> hand out massive chunks of it to anyone who asked. As a former ham I
>>> definitely appreciate the concept of them having space to play ... errr,
>>> experiment with. But did they ever, really, need a /8? Historically, what
>>> percentage of that space has ever actually been used? And as Dave Conrad
>>> pointed out, given all of the "historical" allocations that have been
>>> revisited and/or repurposed already, is taking another look at 44/8
>>> really that far out of line?
>>> 
>> Taking another look is not at all out of line. Discarding 25% of it before
>> letting the community in question on a broader scale take a look is
>> absolutely very far out of line IMHO.
> 
> While I never knew Jon personally, I can't think he would have expected the
> value of IPv4 space either.  I can only think his allocation policies would
> have been much more different.

I never actually met Jon personally. However, what I know from other people’s
accounts and from reading various documents written by him in the early days
of the internet, I think that anyone approaching him with the idea of IPv4
addresses having monetary value would have been viewed as a strange
attempt at dark humor.

> <snip>
> 
>> In actual fact, had the ARDC board approached the broader amateur radio
>> community with a plan to sell the space, it’s entirely likely that I would
>> have lent my support to the plan. That does not change the fact that I feel
>> it was beyond their mandate and out of line for them to take the action
>> first and neither consult the community before nor after.
> 
> This has been my position all along.  A discussion should have been had.
> There was no pressing reason it had to be done right now.

Other than the buyer’s fear that they might not get everything they wanted. Likely
this resulted in the buyer insisting on secrecy and applying pressure to close the
transaction before any risk of publicity.

> Based on my personal discussion with Brian Kantor in 2014, he was 100% against
> selling, or even leasing it as he stated ARDC merely was the custodian of the
> space for amateur use.  I can't think what changed, other than his employment
> situation.
> 
> "They drove a dump-truck full of money up to my house, I'm not made of stone!"
>  -- Herschel Krustofsky

Sounds about right.

Unfortunately, I don’t think the amateur radio community is strong enough to put
meaningful pressure on Amazon. I’m not 100% sure I wouldn’t have supported
the action in a community consultation, either. I resent that this was done without
consultation, I definitely would have insisted on better governance structure for
ARDC prior to approving such consultation. At this point, I think we have lost.
I think the best we can do is try to adapt the governance structure of ARDC to
empower the Amateur Radio community and ensure that there isn’t a repeat of
this.

Owen




More information about the NANOG mailing list