Comment on FR Doc # 2012-11775

Regarding: Question 39: What standard of availability, if any, is appropriate? and commenting on the statements preceding it.. We consider high availability to mean near 24 hours a day, 7 days a week availability. In other words, to demonstrate compliance with this CTE, an NVE would need to ensure its services would be available at all times, except for very limited, scheduled periods of time. We believe such a requirement is necessary because the need to engage in electronic exchange may occur at any time. We chose SMTP for the Direct Exchange Protocol because it was tolerant to underlying network downtime. SMTP is just fine with a network that is up 75% of the time. You actually do need this kind of uptime to do REST based data exchange which is why that model is so backwards. Your presumption regarding what uptime is is laughably ignorant of any real experience with mail servers. Mail servers go down. The mail system works because SMTP does not care if mail servers go down... it just trys again later.. Even if mail servers do not go down constantly, rural networks still do. Your notion of high availability is based on bad assumptions about how good urban broadband is. But rural sites simply might not have the internet access required to meet these availability requirements. We (the other Direct Project designers and I) gave you a fault tolerant protocol chosen specifically so that you did not have to make these types of silly requirements. Then you decided "REST is good too". Wake up. -FT -FT