I think 1960-1990 would definitely be seen as the golden age of systems. Unix, C, threads and concurrency, the web, etc. I don’t have the feeling like systems people were ever shunned or anything like that?
I can’t make heads or tails of what he’s saying there. It sounds like he’s upset that there aren’t more operating systems? He keeps naming systems projects that were newly developed, but then says that the industry stagnated. He uses the web as an example of stagnation - I couldn’t disagree more. Think of all the purely technical things the web has introduced even since this post was written: XHR requests, WebAssembly, tons of work on JavaScript and its frameworks, distributed database technologies.
So that slide deck is 22 years old, and you have to take it in the context of the time.
Nevertheless, it seems very clear what he’s saying, whether you agree with it or not. He’s basically saying that there’s no place to build holistic software systems, and thus few people building them. Everything is constrained by the past – by profitability (incremental acquisition of users) and publishability (focus on performance).
I happen to think the criticisms are still true if you look at say how the Linux kernel supports containers (haphazardly, and without security properties), the state of browsers, recent changes to OS X and Windows, etc.
And yeah even though I like the web, I would not point out XHR and JS as good parts of it …
The community must accept and explore unorthodox ideas.
The community must separate research from market capitalization.
Yea, I must be too blinded by what’s happened since this talk. I guess I was moreso surprised by a commenter in the present day saying that systems people are not appreciated. I’ve long felt that the industry is dominated by systems research and development. Sure, the research ends up actually being used, but I thought that’s what Mr. Pike was saying that he wanted.
I’m glad PLT people finally have their place in the sun. I’m hoping systems people will have theirs next.
I think 1960-1990 would definitely be seen as the golden age of systems. Unix, C, threads and concurrency, the web, etc. I don’t have the feeling like systems people were ever shunned or anything like that?
Unix and C killed any theoretical golden age - even Pike would agree.
I can’t make heads or tails of what he’s saying there. It sounds like he’s upset that there aren’t more operating systems? He keeps naming systems projects that were newly developed, but then says that the industry stagnated. He uses the web as an example of stagnation - I couldn’t disagree more. Think of all the purely technical things the web has introduced even since this post was written: XHR requests, WebAssembly, tons of work on JavaScript and its frameworks, distributed database technologies.
Seriously, what is he talking about?
So that slide deck is 22 years old, and you have to take it in the context of the time.
Nevertheless, it seems very clear what he’s saying, whether you agree with it or not. He’s basically saying that there’s no place to build holistic software systems, and thus few people building them. Everything is constrained by the past – by profitability (incremental acquisition of users) and publishability (focus on performance).
I happen to think the criticisms are still true if you look at say how the Linux kernel supports containers (haphazardly, and without security properties), the state of browsers, recent changes to OS X and Windows, etc.
And yeah even though I like the web, I would not point out XHR and JS as good parts of it …
Yea, I must be too blinded by what’s happened since this talk. I guess I was moreso surprised by a commenter in the present day saying that systems people are not appreciated. I’ve long felt that the industry is dominated by systems research and development. Sure, the research ends up actually being used, but I thought that’s what Mr. Pike was saying that he wanted.