Looks like Joyent, the developer of node.js, is also getting support
directly from Microsoft in their implementation. I'm assuming the IOCP work
is going to go into libuv.
http://blog.nodejs.org/2011/06/23/porting-node-to-windows-with-microsoft%E2%80%99s-help/
On Fri, Jun 24, 2011 at 2:28 AM, Pie
On 06/24/2011 03:26 PM, Martin Pales wrote:
>
> On Fri, Jun 24, 2011 at 12:23 PM, Anton Yemelyanov
> mailto:anton.yemelya...@gmail.com>> wrote:
>
> I just wanted to mention boost::asio, which is rather highly
> regarded as all boost things are and sits on top of IOCP on windows
> as wel
On Fri, Jun 24, 2011 at 12:23 PM, Anton Yemelyanov <
anton.yemelya...@gmail.com> wrote:
> I just wanted to mention boost::asio, which is rather highly regarded as
> all boost things are and sits on top of IOCP on windows as well.
On the other hand, I like the fact that libzmq is a lightweight li
I just wanted to mention boost::asio, which is rather highly regarded as all
boost things are and sits on top of IOCP on windows as well.
ASY
On Fri, Jun 24, 2011 at 5:30 AM, Martin Sustrik wrote:
> The question is whether is whether adjusting 0MQ to use this library is
> in any way simpler tha
The question is whether is whether adjusting 0MQ to use this library is
in any way simpler that adjusting it to use IOCP proper.
Martin
On 06/24/2011 11:28 AM, Pieter Hintjens wrote:
> Looks very useful, and the license is BSD as far as I can see, so fine
> for use with libzmq.
>
> -Pieter
>
> O
Looks very useful, and the license is BSD as far as I can see, so fine for
use with libzmq.
-Pieter
On 24 Jun 2011 10:37, "Dhammika Pathirana" wrote:
> Hi Martin,
>
> On Wed, Jun 22, 2011 at 2:12 PM, Martin Sustrik
wrote:
>> On 06/22/2011 11:07 PM, Steven McCoy wrote:
>>> Following on from withi
Hi Martin,
On Wed, Jun 22, 2011 at 2:12 PM, Martin Sustrik wrote:
> On 06/22/2011 11:07 PM, Steven McCoy wrote:
>> Following on from within a VM, here is native Windows results.
>
> Pretty bad.
>
> Nobody ever done much optimisation on Win platform :|
>
> Btw, I recall there's the poll function i
On 22 June 2011 17:22, Pieter Hintjens wrote:
> On Wed, Jun 22, 2011 at 11:07 PM, Steven McCoy
> wrote:
>
> > Following on from within a VM, here is native Windows results.
> > ./local_lat tcp://*: 1 10
>
> Steve,
>
> You're running tests with 100K messages, which is too low for sensible
On 22 June 2011 17:12, Martin Sustrik wrote:
> On 06/22/2011 11:07 PM, Steven McCoy wrote:
>
>> Following on from within a VM, here is native Windows results.
>>
>
> Pretty bad.
>
> Nobody ever done much optimisation on Win platform :|
>
> Btw, I recall there's the poll function in Win7. What abo
On Wed, Jun 22, 2011 at 11:07 PM, Steven McCoy wrote:
> Following on from within a VM, here is native Windows results.
> ./local_lat tcp://*: 1 10
Steve,
You're running tests with 100K messages, which is too low for sensible
results. I'd raise this high enough that each test runs for at
On 06/22/2011 11:07 PM, Steven McCoy wrote:
> Following on from within a VM, here is native Windows results.
Pretty bad.
Nobody ever done much optimisation on Win platform :|
Btw, I recall there's the poll function in Win7. What about trying with
ZMQ_FORCE_POLL defined so that it's used instead
Following on from within a VM, here is native Windows results.
*Cygwin*:
./local_lat tcp://*: 1 10
./remote_lat tcp://10.208.0.104: 1 10
message size: 1 [B]
roundtrip count: 10
average latency: 140.245 [us]
./remote_thr tcp://10.208.0.104: 1 10
./local_thr tcp://*:555
12 matches
Mail list logo
