[cpp-threads] RE: Ping on memory model and concurrency

[Kevlin Henney]

In message 
<65953E8166311641A685BDF71D8658262B70E7 at cacexc12.americas.cpqcorp.net>, 
"Boehm, Hans" <hans.boehm at hp.com> writes
>
>3) I haven't seen much progress with respect to a threads API (thread
>creation,
>locking, etc.)  I personally still think this is a difficult political
>problem
>whose solution is far less critical than the memory model issues.  If
>the
>memory model issues were solved, it would be easy to fix the
>specifications
>for pthreads, win32 threads, boost threads, etc.  But the committee
>definitely
>appeared to be interested in addressing it.

A belated reply to this.

Other demands on my time in the last few months have prevented me from 
making substantial progress on this. However, I managed to find enough 
time today before the deadline to put together a discussion paper based 
on some postings and documents I have put together in the past. It is 
attached below. I won't be able to fold any comments in before the 
mailing goes out (the official deadline was Friday, but Monday morning 
PDT was the practical cut off). I may be able to add some material for 
circulation at the meeting (sent directly to Howard Hinnant), but I do 
not anticipate the paper growing to include more worked examples and 
definitional text until the Berlin meeting next year.

A Boost implementation effort, based on the design I am proposing, is 
now underway. I understand Lawrence Crowl will be submitting a proposal 
on the API. I am hoping that the work will be complementary or 
sufficiently aligned and that there will be sufficient common ground to 
make co-operative rather than competitive progress. The evolutionary 
view between the model I am proposing and the current Boost.Threads 
approach also hopefully highlights enough commonality that those 
favouring a straight adoption of Boost.Threads can see this is as a 
viable alternative.

>Re: Attendance
>
>I'm currently planning on being there in person.  It would be great if
>others
>could make it, too.

Unfortunately, I will be unable to attend, so I have not been as 
motivated to flesh out the current document beyond the framework and 
rationale stage. However, I plan on attending the spring meeting next 
year.

Kevlin

-------------- next part --------------
A non-text attachment was scrubbed...
Name: n1883.pdf
Type: application/octet-stream
Size: 157025 bytes
Desc: n1883.pdf
Url : http://shadbolt.decadentplace.org.uk/pipermail/cpp-threads/attachments/20050829/bfa34014/n1883.obj
-------------- next part --------------

-- 
____________________________________________________________

   Kevlin Henney                   phone:  +44 117 942 2990
   mailto:kevlin at curbralan.com     mobile: +44 7801 073 508
   http://www.curbralan.com        fax:    +44 870 052 2289
   Curbralan: Consultancy + Training + Development + Review
____________________________________________________________