@AlecMuffett @pvineetha @SteveBellovin So — and really not trying to pick a fight — the issue is that some of us folks on the outside are genuinely trying to understand this stuff. That means, put actual numbers and metrics to it. 1/
@AlecMuffett @pvineetha @SteveBellovin And when we do this, we run into an obstacle. Which is that corporations like Facebook tend not to publish useful numbers. On costs, bandwidth, failure rates etc. (see yesterday’s conversation.). So we’re doing our best to process this with limited data. 2/
@AlecMuffett @pvineetha @SteveBellovin For the simple reason that answering important questions with data seems a lot more valuable then answering them with feelings. 3/
@AlecMuffett @pvineetha @SteveBellovin Into this attempt at understanding we get people inside the companies telling us *not that we’re wrong* in our suppositions and estimates. More like: to outsiders like us, these things are fundamentally unknowable. 4/
@AlecMuffett @pvineetha @SteveBellovin And the problem with this is that it’s fundamentally unhelpful. If we’re wrong, explain why we’re wrong. Hell, if you can’t do it publicly — do it privately in confidence. Folks do this to me all the time and I correct myself. 5/
@AlecMuffett @pvineetha @SteveBellovin So when I crunch the published cost numbers on 2014 WhatsApp and come up with a per-user cost for operating an e2e back-end and then say “well let’s add a fudge-factor of 10x just in case WhatsApp in 2014 was alarmingly efficient”, and then... 6/
@AlecMuffett @pvineetha @SteveBellovin ... even after doing that I *still* get people telling me that operating a business at scale is just so hard that I can’t reason about it until I start my own 1bn+ user service 🙄 7/
@AlecMuffett @pvineetha @SteveBellovin I guess what I’m saying is that if people have something to contribute, do it. But this “I have secret info that tells me you’re not wrong exactly, but just in a place where you can’t be right” then what can we do with it. /Fin
