/ Interview with R&D head Alexander Mayrhofer

Jul 29

/ nic.at News - 29.07.2019 12:43
Interview with R&D head Alexander Mayrhofer

Alexander Mayrhofer has managed the R&D department since 2002 - this is, where the many innovative research projects, ideas and numbers that surround nic.at originate. Among other activities, he participates in the Internet Engineering Task Force (IETF) and often travels to speak at events, where both his talks and his eye-catching illustrations (see below) are always popular. In this interview, he also reveals on which projects he is currently working on, what awaits us in the future and how to keep yourself innovative.



To begin with, a general question: what are the responsibilities of the R&D department at nic.at?
Our activities can be divided into three areas: firstly, we work to develop and improve existing products, such as our gTLD registry, or RcodeZero DNS. The second area is focused on community and standardisation. At the international level, we play an active role in CENTR, and in the development of the internet by participating in the Internet Engineering Task Force (IETF). We organise events for the national community, like the DNSheads Vienna meetups. Our third focus area is research work – like, for example, gaining new insights from our data, as well as research on current topics such as identity management.

Our job is to work with highly complex things so that we can make use of the insights they offer and pass these on to our registrars and partners.

What projects are you working on at the moment?
Currently I’m working a lot on DNS Magnitude, a kind of statistical measure for the popularity of domains. Using a special algorithm, each domain is placed on a scale from 0 to 10 – providing a simple way to assess how intensively a domain is used. This can then be used by registrars, for example, for upselling – a popular domain should of course be appropriately safeguarded by products such as Rcode Zero DNS. For instance, it would be very frustrating for both the customers and the operator if a popular online shop was temporarily unavailable.
Another major topic is encrypted DNS. Almost every action on the internet begins with DNS queries. This information is usually transmitted over networks in plain text. Two new protocols, DNS over TLS and DNS over HTTPS, aim to change this. However, DNS over HTTPS in particular also opens up the possibility that in the future the majority of DNS queries will only be responded to by a small number of large operators – a development that many are, at best, sceptical about.

Where do the ideas that are later turned into projects come from?
Many arise out of daily work, or simply when we have a little time to spend on new things. Talks with like-minded people who want to solve similar problems – so ideas from our community – are also often inspiring. But innovation does not always mean inventing something entirely new – most of the time, existing things are combined in new ways. An example of this is our data warehouse. We are just now starting to combine internal insights in new ways, and to offer them to registrars and the public domain.

How many of these ideas are actually implemented?
It’s helpful here to use the concept of an innovation pyramid, based on a factor of 10: out of 3,000 ideas, 300 will be explored in more detail, of which 30 will become significant, specific projects, which can result in 3 completely new strategic directions for a business. So it’s about a lot of trying things out, discarding ideas that don’t work, and picking up something else to try.

What motivates you to stay curious? How does one stay innovative?
I like things that are initially very complicated, that I can simplify, so that they are useful and – this will sound a bit corny – just elegant. Our motivation comes from producing something useful that in the end can actually be applied and will contribute to the company’s success. And I think that you stay innovative if you ask yourself, many times a day, questions like, “Can I make that better? Am I solving a problem? What is the benefit – how will users, registrars and our company profit?”
I’m proud that Google is currently implementing a new protocol for encrypted DNS that I standardised.

Is there not a danger of spreading yourself too thin?
Yes, definitely. You can’t tackle too many things at once – the temptation is great, of course, because there are so many new ideas, and that’s exciting. Innovation is always accompanied by the risk of “stranded work” – have courage and think outside the box, but be able to let projects go if they aren’t heading in a useful direction.

Can you give us a glimpse of the future?
I can at least try. Machine learning will soon play an everyday role in ordinary businesses. We are already using it for forecasting. The challenge here will be how to make machines answerable, like people, for decisions they make – if an algorithm doesn’t allow me to buy a train ticket, I will want to hear a justifiable reason.

Identity management will also be an important issue. The general public will be much more sensitive when it comes to data security and data sovereignty – a lot will happen in this area in the near future.


Graphics: Domain Lagoon

Alexander Mayhofer’s legendary “domain lagoon” illustrates the vastness of the .at namespace – and that the number of domains actually registered can hardly get close to filling it. The “Isle of the Blessed” of 1.3 million registered .at domains is surrounded by the atoll of deleted domains, and the lagoon of WHOIS and EPP queries about free domains. This is encircled by the DNS sea of 300 million NXDOMAINs a month – these are domain name queries received by our DNS servers for domains that do not exist. Out on the high seas lurk sea monsters in an ocean of unregistered domains which are never queried – making it clear that the potential for different .at domains is nowhere near exhausted.


The latest nic//report

In case you want to find out more about the R&D department and their exciting projects, just read our latest nic//report:

Download nic//report