The commentary was terse (hiking in the mountains, limited IP access) but accurate. I have designed interesting algorithms and software on large-scale database and HPC systems for much of my adult life. As a consequence, I am much more familiar with these domains than most casual programmers or academics in these fields.
The important point was that ideas like MapReduce and Spanner are literally rehashes of very old ideas that never fell out of use in some domains. But they are so old that no one that uses those ideas would dream of publishing a paper on it because it has been in continuous use for ages. In most cases, a deep enough dive in the dead tree literature will often turn up examples of these in real implementations. I learned most of it from graybeards working on cool stuff when I was young, not from the literature.
I am not exactly young but both MapReduce and Spanner predate my time in computer science. BigTable was actually a novel composition of existing but obscure bits of computer science so I give Google full credit for that.
Google has shown an exceptional ability to exploit ignored or forgotten computer science. However, to the extent they exploit it there is often little invention beyond those that they borrow the ideas from. In that respect, I think the credit is somewhat misplaced with respect to invention.
The important point was that ideas like MapReduce and Spanner are literally rehashes of very old ideas that never fell out of use in some domains. But they are so old that no one that uses those ideas would dream of publishing a paper on it because it has been in continuous use for ages. In most cases, a deep enough dive in the dead tree literature will often turn up examples of these in real implementations. I learned most of it from graybeards working on cool stuff when I was young, not from the literature.
I am not exactly young but both MapReduce and Spanner predate my time in computer science. BigTable was actually a novel composition of existing but obscure bits of computer science so I give Google full credit for that.
Google has shown an exceptional ability to exploit ignored or forgotten computer science. However, to the extent they exploit it there is often little invention beyond those that they borrow the ideas from. In that respect, I think the credit is somewhat misplaced with respect to invention.