A layman's proposal for addressing the fragility of DNS
joelglovier.comIndeed, layman's. This idea introduces so many problems that it's hard to pick what to start with. The most obvious are volume of changes and which changeset is the current one.
You really didn't need to boast about it.
Also, FTR there is no intent to boast here, merely sharing an idea and trying to elicit discussion about a problem which I admittedly do not fully understand.
I had hoped that starting with an admission of plausible fallacy would deter such judgements, but I guess you never can underestimate the power of HN to bring out the moral superiority of others.
Git has a very scalable solution for a large volume of changes and establishing change set as current (via SHA). What is different about the DNS registry than any other codebase that Git could not handle?