Jeffrey M. Barber

Tombstone

Career: Amazon S3

Mar 2011 โ€“ Dec 2014 ยท age 29โ€“32

What it was

Three years and eight months as a software engineer, then senior software engineer, on Amazon S3, joining when the store held around four hundred billion objects. The public record is small and deliberate: a blog post, a LinkedIn page and six AWS launches Jeff helped ship. Server-side encryption with S3-managed keys, in October 2011. Object expiration through lifecycle rules, that December. Cross-origin resource sharing for browsers, website redirects stored on the object, lifecycle transition to Glacier, and website routing rules with root-domain hosting, all in 2012. After that, in the post's words, "stuff that I couldn't talk about." Jeff left on December 1, 2014, eighteen days after S3 event notifications and Lambda launched on the same day. The store could finally call code, and that was closure.

Wins, for the age

What it taught

Genealogy

Ancestors: none on record. The personal code goes quiet in these years; what cannot be shown was not kept. Descendants: Wake: a static site generator that deploys to S3 by delta upload. Jswam: the website redirect and root-domain features, used from the outside. Adama Caravan: cloud storage and a hand-rolled request signer, written knowing the sharp edges. Goat Bot: a release pointer in S3 that refuses to overwrite a version.

Epitaph

Here lies a tour of duty. Jeff joined at four hundred billion objects, helped the store learn to expire, redirect, encrypt and move to Glacier, and left once it learned to call code.