[ruby-talk:5888] Re: [RFC] Towards a new synchronisation primitive

At Fri, 27 Oct 2000 06:13:32 +0900,
hipster <[e-mail]> wrote:

[...]
> Every instance of class Object (and derivatives) can have a lock, or binary
> semaphore, associated with it. When a method is called on an object that is
> declared synchronized, or a method declared synchronized is called on an
> object, or an object lock is explicitly obtained from within a member
> method, the object becomes locked. 

hmm..  I thought Ruby is avoiding explicit declaration.  of cause,
'never done' doesn't necessary mean bad, but kills consistency.

[...]
> Backward compatibility with the current semantics could be achieved by
> defining `Thread.critical = false|true' as a method obtaining or releasing
> the object lock for the instance that is in scope at the time of the call.
> (But as the notion of a critical section is abstracted away from a thread,
> some code `refactoring' ;) might be a better idea.)

so, basically your proposal is based on async messaging model, right?
I just found out it a few days ago, so I guess I'm not qualified to
discuss about it.  but let me write a few, anyway ;)

I think we should first discuss what model Ruby should use.  The
current model is as I quote Matz on -list "a model from ML interpreter
called CSL."  which I don't know anything about ;P. [ruby-list:1767] 

dW recently had a java's thread article:
http://www-4.ibm.com/software/developer/library/j-king.html?dwzone=java

when I asked about async messaging model on -dev, Yanagawa told me
that "the model is clean and simple, but implementation usually cost
much more". [ruby-dev:11288] 

He also mentioned that "It's still arguable that which model is easy
to understand".  As I see Ruby, "easy to understand/program" is one of
core feature we should keep.

On the other side, Shugo and others are thinking about
re-implementation of thread, too.  Their idea is to keep current
thread model and extend the thread premitives so that mutex and cv
works with thread scheduler.  Isn't it? > whoever reading ;)

anyway, i guess it's time for thread novice to fade out and let others
discuss.  I hope I'll learn more and be able to comment on some.

regards,
--
        yashi