Yahoo proposes ‘really ugly hack’ to DNS
Network World - ANAHEIM, Calif. — Network engineers from Yahoo are pitching what they admit is a “really ugly hack” to the Internet's Domain Name System, but they say it is necessary for the popular Web content provider to support IPv6 …
Context & Ripple Effects
Yahoo's pitch is plumbing, not product: while the company's March has been dominated by deal-making — Justice Department clearance for the Microsoft search agreement, the Citizen Sports acquisition, the Facebook Connect rollout — its network engineers are publicly floating a self-described "really ugly hack" to DNS because the standard system alone cannot get IPv6 clients to Yahoo content.
The framing is characteristically Yahoo, a company whose own blog celebrated "hacks come to life" back in 2007, but this time the hack targets core Internet infrastructure. That Network World carried the story alongside its reporting on Google, Microsoft and Netflix weighing a shared list of IPv6 users shows the transition problem is being worked provider-by-provider rather than through one coordinated upgrade.
First-order effects
- Yahoo's proposal asks DNS operators and resolver-software maintainers to accommodate a non-standard resolution behavior so yahoo.com stays reachable to IPv6 visitors during the transition — acceptance puts real implementation work on everyone serving Yahoo-bound lookups.
Second-order effects
- If peers such as Google, Microsoft and Netflix each pursue their own transition mechanism, as the contemporaneous talk of a shared IPv6-user list suggests, resolver operators end up supporting a patchwork of provider-specific behaviors instead of a single standard.
Third-order effects
- A transition assembled from large providers' individual hacks turns ad-hoc engineering into de facto standards, shifting influence from the DNS standards process to whichever content companies deploy first — and raising the eventual cost of unwinding those hacks once native IPv6 arrives.
The trend: Major content providers are improvising DNS-level workarounds to bridge IPv4 and IPv6, making ad-hoc engineering rather than coordinated standards-setting the de facto migration path.