nominet drops, stability, sustainability & the hive

On the 13th of September 2022, .uk dropcatching changed forever. Find out what changes were made, and what it means for the future of the catching industry.

A hive mind surrounded by millions of ants in the style of vaporwave
A hive mind surrounded by millions of ants in the style of vaporwave

A new dawn

On the 13th of September, Nominet rolled out (in one fell swoop) their new drop system for expiring domain names and for the first time ever, public drop lists via Nominet directly. Needless to say this change has shook the industry and knocked a few apples out of the tree. The change was an expected change, after years of deliberation through UKRAC consultations and meetings, if you're a member of Nominet you can read more about his here on the Nominet community forum.

The change to Nominet's domain release system, means that the DAC (Domain Availability Checker) has now been retired for use with drop-catching. The DAC previously gave registrars 432,000 queries a day to poll through expiring domain names, in order to rapidly register domain names when they dropped back into the registry. This was charged at a fee of £25 (exc. VAT) per annum. Will registrars see these fees refunded? Nominet is yet to confirm, but has so far given no indication that these fees for a DAC that cannot be used for drop-catching, will be refunded.

Latency & Timing inconsistencies

The new drop system has ripped through the industry causing waves of frustration at apparent technical nuances, between the published drop times, and the actual time the domain name drops back into the registry.

There have been no technical details released by Nominet, on how to use the new system. Meaning that, yet again, it's up to those dropcatching to figure out just how many create commands to send down the pipe and when. Each member of Nominet gets a quote of 1000 create commands to send per rolling 24 hour period. Plenty, you'd think? Maybe so if you were registering unregistered and uncontested domain names. But this is the dropcatching industry. We're talking about hundreds of TAG holders trying to compete for one single domain at the exact time it drops. Nobody knows how many creates to send to be competitive, so a lot of people are sending >100 per domain.

100(creates)*100(TAGs) = 10000 commands from 100 different IP addresses. In less than a 100 milliseconds. Of course, not everybody is sending 100 creates per domain, this may be an extreme example. But looking at the latency between each domain, it appears to be random.

Can Nominet's systems handle the load?

Each create request has a unique ID number, the successful ID is a key indicator of how early or late another person is sending. Let's take a look at a recent contested domain name that we successfully caught, and the differences between the 10 ID's that came before it.

26634433190 // Ending ID
...
26634432550 // Winning ID
...
26634432464 // First send

Long-term feasibility & the hive mind

We attended Nominet's recent EGM and posed the question to them, whether they will look into the inconsistencies in drop time. Nominet will review this within 6 months for any technical problems. Meanwhile, we're stuck guessing — I suppose many of us will compare this to the randomness of DAC, a sort of lucky draw, and that would be the case if the current landscape wasn't being dominated by a few larger organised parties.

Statistics via BRIGHTWORK on an average of 10-15 domains per day queued over 7 days

20 or so of the TAGs on that list chase from a universal droplist as a hive mind. We believe this to be a breach of Nominet's Anti-Avoidance AUP, but Nominet wouldn't agree with us:

You have provided sufficient evidence for us to investigate the matter and we have done so in accordance with the standards listed in our Acceptable Use Policy, which can be found here: Acceptable Use Policies – Registrar Resources (nominet.uk)
Of course if you have any additional information that contravenes our Acceptable Use Policy, please share this with us. I would ask that any further submissions made relate to new evidence only and you make clear reference to the section of the AUP you feel is non compliant.

When investigating claims of collusion, we are looking to see if any of the domain query and domain management tool principles are being compromised by people working together by most likely pooling their individual systems allocations /quotas. If evidence of this activity is found relevant actions are taken and, where appropriate, sanctions applied. This will usually be to record the tags, subscriptions etc. as being linked, and therefore subject to a lower overall limit than would have been applied to a similar number of non-connected persons.

The aim is for our AUP to be clear and transparent and I'd be more than happy to help clarify any further points on the policy if that would be helpful but I cannot comment nor speculate on other registrars business model or practices in the same way we wouldn't disclose any of your activities with a 3rd party.
I hope this clarifies our position.
Kind Regards
Matthew Brown
Registry Compliance Advisor

Evidently there is an advantage to having multiple tags, so long as you have 20+ of them to do your bidding. If you throw 20 darts at a board, one of them will hit the bullseye more often than not. However, if you only have one dart, your chances of hitting that bullseye dramatically reduce, whether you have the best software or not.

What do you think? Are you finding the new system hard to compete in? Let us know in the comments.

💡
If you liked this post, subscribe to us for more.

Like what you see? 👀

Don’t miss out on the latest posts. Sign up now to get access to more content.
hallie@example.uk
Subscribe