You're bringing back some good memories, Ralph--thanks! Most of the GoF patterns (as exemplified) introduce flexibility through polymorphism. Depending on your system, you may be able to take advantage of polymorphism, even if you don't allocate memory dynamically. You might be able to construct all your polymorphic objects at initialization time, perhaps using pools of reusable objects, or, assuming C++, take advantage of placement new. Using virtual methods, you'll take a small (but predictable) time hit going through the v-table, and add ram for the v-table pointer itself, but these might be quite a bit more acceptable than dynamic memory allocation.
Here's an introductory article you might find useful:
Effective C++ Memory Allocation by Aaron Dailey http://www.embedded.com/1999/9901/9901feat2.htm
If you think this is a profitable direction for you to explore, let me know and I'll try to remember the sources that helped me when I was learning this stuff (but I suspect others on the list will have more current suggestions than mine).
- Russ
Ralph Malph wrote:
Hello all, looking at archives I'm not sure how "alive" this list is, but someone out there might be listening.
I'm interested in people's opinions as to whether
design patterns would be useful if you could not
dynamically allocate memory.
I am new to patterns and find them quite interesting. I am considering putting in some effort to comprehend
them, but my projects at work include embedded
safety-critical real-time systems, and one of the
things that we do not use is dynamic memory
allocation.
If anyone has an opinion as to whether I should delve into patterns for this type of software (i.e. you have considered doing them and/or done them in similar systems before) I'd be interested to hear it.
Thanks!
__________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site!
http://smallbusiness.yahoo.com/resources/
_______________________________________________ patterns-discussion mailing list [email protected] http://mail.cs.uiuc.edu/mailman/listinfo/patterns-discussion
_______________________________________________ patterns-discussion mailing list [email protected] http://mail.cs.uiuc.edu/mailman/listinfo/patterns-discussion
