Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

This is one area Google does things relatively well, I think. Especially when they branch out and design a new thing with heavy CS content, the lead designers are often in their 40s or 50s and have experience with significant past systems in related areas: Rob Pike, Lars Bak, Ken Thompson, Peter Norvig, etc.

I think one reason is that people of that age are mostly the ones who even remember various useful solutions to problems and tradeoffs that have come up in the past. CS is strangely bad at digitizing its history: I can get a philosophy journal issue from 1890 on Jstor, but many CS journals and proceedings from as recently as 1980 are still dead-tree only. And many are still relevant, e.g. the V8 js engine is strongly based on 1980s Smalltalk compiler research. A handful of important papers (say, by Turing) have been piecemeal scanned and put online by interested people, and a handful of university classes cover historical aspects, but overall I don't think history transmission is taken particularly seriously in CS circles. So, it ends up being more of a living-memory sort of deal, where if you want knowledge from the 1970s/80s to influence your designs, you have to hire people who personally remember it.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: