BSwitzerland built an alternative to BGP. Nobody noticed.
https://www.theregister.com/2026/03/17/switzerland_bgp_alternative/
Bruce Simpson, Ph.D.@blainsmith Er, sorry. I have already criticised this article which seems to be uncritically promoting the SCION architecture, when RFC 8205 BGPSec already reached Proposed Standard status at the IETF. There's been no clear explanation of how they essentially differ and it looks like sloppy journalism to me. My replies at their LinkedIn thread: https://www.linkedin.com/posts/the-register_switzerland-built-an-alternative-to-bgp-activity-7439949368562028545-8o1z?utm_source=share&utm_medium=member_desktop&rcm=ACoAAABpaDkB0xMW8ZMjseSzGs467EN-oUC7cao
@blainsmith "How does SCION differ from IETF RFC 8205 BGPSec? It has Proposed Standard status, SCION does not. Your article hasn't addressed that; it just gets a name-check. For isolation, VRF already exists in most NOS. For multipath, ECMP and exit discriminators exist. BGPsec also provides cryptographic validation of AS-Paths. This is the first I've ever heard of SCION, but I've been on sabbatical. My PhD is partly about obviating the need for BGP for some use cases."
@blainsmith "Another thing: The gold standard for network-layer failover is typically 50ms. Usually BFD is employed with BGP for this. G.8032 can achieve this, but it's layer 2. On the upcoming front, SRv6 TI-LFA only tries to guarantee sub-100ms and is utterly dependent on IGP state, usually IS-IS, to achieve this. I'm not necessarily "defending" BGP vs SCION here; that's not my issue and it may not even be a "fight", as I have stated, I have worked on tech to partially deprecate BGP."
@blainsmith If I'm reading between the lines, there might have been financial interest behind the article and its appearance, but I can't be 100% sure of that. It would not be the first time The Register has been found suspect in this way as others elsewhere have commented.