Re: Parser Generators for Multiple Protocols in an Embedded Device

iddw@hotmail.com (Dave Hansen) wrote:
>"Embedded Systems" is probably the least specific term in all of
>computer science. I personally have worked on an embedded system that
>required more than 3 times the memory of your "WWW server" example.
>The code was pretty much standard C++ (as close as the available
>compiler could get us, anyway).
>
>I just finished another project a few weeks back with less than 1K of
>program memory, using less than 32 bytes of RAM. This project used
>mostly standard C (as close as the available compiler could get us,
>anyway).


One of the paradoxes when computers become more powerful is that one
can do more of everything: For example, one might apply more advanced
compression techniques to actually make things smaller.


Therefore, I think that languages like C++ should be extended to
include both levels, but downwards towards the portable assembler
level, and upwards towards more structured programming admitting
better implementation of dynamic structures.


This will picture will only change when the CPU's start to graduate
and becoming capable of treating a more advanced (structured) binary
model than currently. If that would happen, then there would be no
point in low level programming anymore (except for implementing the
CPU's) as programming costs will be too high. (Low level programming
is very time consuming, you know.) But this kind of development, we
haven't seen anything from as of yet.


    Hans Aberg * Anti-spam: remove "remove." from email address.
                                    * Email: Hans Aberg <remove.haberg@member.ams.org>
                                    * Home Page: <http://www.matematik.su.se/~haberg/>
                                    * AMS member listing: <http://www.ams.org/cml/>
[CPUs used to have all sort of higher level features. They turned out
not to be useful, which is why we have all sorts of RISC chips now but
nothing like the Intel 432. -John]

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