nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqta3cuyrlw828clw5n0dg7efxzjfh80frr4zxmzjq0lgqcdfaertq7t5pjr (nprofile…5pjr) nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqln5q8np5aezhtt7ztv6tah86xk4t3smjuchdvxp0u6uta056204qf4h82d (nprofile…h82d) nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqyz5hmul3qjuvzgsgh666xenzalxnn9ed55pcduspjfznzqvhwraq4jpvn3 (nprofile…pvn3) I am curious, when Cloudfront delivers 60% of all podcast episodes, how affordable HLS would be. Cloudfront charges are, roughly:
$0.75 for a million requests
$0.085 per GB of data
HLS typically uses chunks of ten seconds. Each of those are a separate request. So, a two hour show is the same amount of data vs a normal download, but 7,200 requests vs just one.
I know that this pricing model has scared plenty of people away from chunked media in the past. What’s different here?