Subj : Re: Lockable objects
To   : comp.programming.threads
From : Torsten Robitzki
Date : Thu May 19 2005 08:27 pm

Uenal Mutlu wrote:

> A mutex can be realized in just 8 bytes (counter and threadid).
> These methods would be like any other method of the object:
> just use which you want/need/like, or don't use at all: you have the choice.

I don't have the choice if I want to reserve 8 bytes for every container 
for no benefit or not. And 8 bytes can be a hugh amount of memory if you 
consider that an empty vector usually just claims the space for 3 
pointers. And there are applications that have hugh amounts of empty 
containers.

> The advantage of having such an "architecture" in the language
> gives far more possibilities for the users.

And now a new buzzword: "architecture". Calling having a mutex in every 
container an "architecture" is just ...

What can a user do with a container that have a buildin mutex that the 
user can not do with a container plus a mutex? What is the difference of:
mutex m; std::vector<int> v;

to your approach and to something like std::pair<mutex, std::vector<int> > ?

> And: multithreading has become a de facto standard in programming,
> esp. in application programming.

Really? Did your teacher told you that? Why especial for application 
programming?

.