From: Peter Dimov (pdimov_at_[hidden])
Date: 2005-10-03 13:31:28
Steven E. Harris wrote:
> "Peter Dimov" <pdimov_at_[hidden]> writes:
>> I appreciate any comments on the interface
> I have a related two lock queue class I wrote a few years ago that
> uses a special set of nested handles to grant certain visibility and
> modification privileges to clients.
> These handles, called "front" or similar variations, are necessary to
> see values at the front or head of the queue. Creating one
> encapsulates blocking and waiting for at least one element to be
> present, holding of the head lock to prevent concurrent popping or
> mutation of the element, and optionally having the privilege to pop
> the current element and move to the next one before some other waiting
> client can.
Thanks, I'll take a look at your rationale link.
One argument against your enhancements is that (to the untrained eye) they
seem to mandate a lock-based implementation. I wanted to permit a lock-free
implementation, if possible.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk