Hacker News new | past | comments | ask | show | jobs | submit login

I'm sort of in the same boat, although I didn't do as much C. (And my interest in getting back into it is more hypothetical.)

Aside from understanding how the language itself has changed, maybe something else to put on the list is how to apply more modern programming practices in C.

In the 90s, I don't think I ever saw C code with unit tests. Any kind of automated testing was pretty rare. I've become convinced that testing in some form is a good thing. If I were going back to C, I'd want to understand the best way to go about that.

People also didn't care (or know) much about security back then. C has some obvious pitfalls (buffer overflows, etc.), and it is pretty important to know good ways to minimize risk. I'd want to understand best practices and techniques for this.

Also, back then build tools were very simple, and some of them were not my favorite things to use (Imake, I'm looking at you). Build tools have advanced a lot since then. Features like reliable, deterministic incremental builds exist now. Some things could be less tedious to configure and maintain. There are probably best practices and preferred choices in build tools, but what exactly they are is another thing I'd want to know.

These are probably not questions that necessarily need an answer from people whose expertise is the language itself, though, so I guess this is a tangent.




People did know about security back then, since it was one of the driving design factors of Burroughs created in 1961, and still sold by Unisys as ClearPath MCP for highly secured deployment environments.

And there were plenty of security related papers and OSes from other companies like IBM, Xerox and DEC.




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

Search: