Re: State of the Art

"Aleksey Demakov" <ademakov@gmail.com> writes:
>Is there anything relatively new (say not described by Muchnik) that
>became or going to become widely used? Something that clearly wins
>over older algorithms.


Ten years may be too short for most techniques to become widely used.
Even parser generators and SSA form, which are probably as successful
as any technique, have taken longer to catch on. E.g., SSA form has
only been used in GCC a few years ago, and lots of people still write
parsers by hand.


Concerning trends, I see a few:


- Multi-core CPUs (and SMT) require better support for multi-threaded
    programs (but with language support, auto-parallelization is too
    hard). In that vein, there has been a lot of work published on
    software transactional memory in recent years; we'll see if that
    catches on.


- Using static analysis and type checking techniques for determining
    additional properties of programs, e.g., wrt security.


    - anton
--
M. Anton Ertl
anton@mips.complang.tuwien.ac.at
http://www.complang.tuwien.ac.at/anton/

Return to the comp.compilers page.
Search the comp.compilers archives again.