On 07/18/2011 08:24 AM, yy l wrote:
> So, what's the problem? Did I use zmq correctly?
You are using sockets from multiple threads which can lead to undefined
behaviour.
Martin
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zerom
Hi Mikko,
Let me know whether the patch is OK.
Thanks!
Martin
>From bf9a923c59e5ebe1907ec5ede52f834751ea993b Mon Sep 17 00:00:00 2001
From: Martin Sustrik
Date: Mon, 18 Jul 2011 09:06:01 +0200
Subject: [PATCH] zmq_sendmsg(3) and zmq_recvmsg(3) added to RPM spec file
Signed-off-by: Martin Sustr
>From 89daa57faab728204447c3bb708e753a2aeffa55 Mon Sep 17 00:00:00 2001
From: Martin Sustrik
Date: Mon, 18 Jul 2011 09:28:59 +0200
Subject: [PATCH 1/2] Missing files for GENERIC socket implementation added
Signed-off-by: Martin Sustrik
---
src/generic.cpp | 265 +++
Hi Martin
Big thank for your support.
I have to study more deeply your guide. Up to now I read chapter 3 and 4. I
found out in chapter 3 that there is router dealer example
There are 2 things that are interesting for me.
1. Asynchronous Client-Server
Clients connect to the server and send req
>From ac20e17cbd1471b9eb3eef79c32cb43420cbc516 Mon Sep 17 00:00:00 2001
From: Martin Sustrik
Date: Mon, 18 Jul 2011 10:04:05 +0200
Subject: [PATCH] Reference to C++ binding removed from zmq(7)
Signed-off-by: Martin Sustrik
---
doc/zmq.txt | 11 +--
1 files changed, 1 insertions(+), 1
On 07/17/2011 06:37 PM, Steven McCoy wrote:
> Fix signaler on Cygwin.
Applied to master. I've also added the same patch to zmq_poll() which
was not supported on Cygwin up to now.
Thanks!
Martin
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
ht
On Sun, Jul 17, 2011 at 8:23 AM, Martin Sustrik wrote:
> Can you please sign-off the patch?
Sure, attached.
> AFAIK, FreeBSD has this option as well. If should be presumably used with
> both ZMQ_HAVE_OSX and ZMQ_HAVE_FREEBSD.
If you like. I'd prefer to only enable it for FreeBSD when we actual
>From 7298b5742f749a2947c9e6460c4a083be84506ed Mon Sep 17 00:00:00 2001
From: Martin Sustrik
Date: Mon, 18 Jul 2011 11:39:38 +0200
Subject: [PATCH] Includes in zmq.cpp cleaned-up
Signed-off-by: Martin Sustrik
---
src/zmq.cpp | 52 +++-
1 files
On 07/18/2011 11:29 AM, Pieter Hintjens wrote:
> Fixed issue 230
Applied to master.
Thanks!
Martin
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
On Sun, Jul 17, 2011 at 1:53 AM, Marc Ilgen wrote:
> OK I tried this very same code with release 2.1.7 of 0mq and it works fine,
> no crashes.
3.0.0 is a first packaging of a new version, so it's fair to assume
there are some issues there.
At this stage, 3.0.x is really only for people who want
Steve,
Do you have an issue in Jira for this? We back all changes to stable
releases with issues, and (where possible), reproducible test cases.
Cheers
Pieter
On Sun, Jul 17, 2011 at 6:37 PM, Steven McCoy wrote:
> Missed one platform on the signaler. Note poll is available on Cygwin it is
> ju
Hmm, sorry, it's not this patch. I'm looking for the origin of this change:
https://github.com/zeromq/libzmq/commit/f97f50c252adc814145a0a482bc2a25229882459
-Pieter
On Mon, Jul 18, 2011 at 1:12 PM, Pieter Hintjens wrote:
> Steve,
>
> Do you have an issue in Jira for this? We back all changes to
On Sun, Jul 17, 2011 at 8:10 AM, Martin Sustrik wrote:
> There's a plan to create a new low-level socket type in the future that
> would replace ROUTER and allow to circumvent the messaging patterns and
> use 0mq as a dumb networking framework. You can help with the work, if
> you are interested.
On Sun, Jul 17, 2011 at 9:36 AM, Marc Ilgen wrote:
> If I remember correctly, the filename was ctx.cpp. The assert was rc == 0. I
> don't remember the line number.
> But it works fine in v2.1.7.
Are you using the 2.1.7 package, or the zeromq2-1 git master?
-Pieter
__
OK, you were testing 3.0, not 2.1.7. Sorry...
On Mon, Jul 18, 2011 at 1:55 PM, Pieter Hintjens wrote:
> On Sun, Jul 17, 2011 at 9:36 AM, Marc Ilgen wrote:
>
>> If I remember correctly, the filename was ctx.cpp. The assert was rc == 0. I
>> don't remember the line number.
>> But it works fine in
Hi all,
A question to the list about release policies[1] and the 3.x development.
Background: we are removing ZMQ_IDENTITY from the product, it's a
necessary step to continuing to improve the product. (Incidentally, if
you are using ZMQ_IDENTITY, let us know your use case).
The current release p
On Jul 18, 2011, at 7:26 AM, Pieter Hintjens wrote:
> Hi all,
>
> A question to the list about release policies[1] and the 3.x development.
>
> Background: we are removing ZMQ_IDENTITY from the product, it's a
> necessary step to continuing to improve the product. (Incidentally, if
> you are usi
Yes, this patch seems to solve the problem.
I think I'll still use v2.1.7 until v3 becomes more stable, but I'm glad
this particular issue got fixed.
Thanks.
--
Marc Ilgen
m...@westofpluto.com
-Original Messag
On Mon, Jul 18, 2011 at 5:26 AM, Pieter Hintjens wrote:
> we are removing ZMQ_IDENTITY from the product, it's a
> necessary step to continuing to improve the product. (Incidentally, if
> you are using ZMQ_IDENTITY, let us know your use case).
Our use case for ZMQ_IDENTITY is this: We have a few
Hello
Suppose I set up a client/server system with SUB and PUB sockets. I want the
client to have to authenticate its ability to subscribe and I want the
server to have the ability to unsubscribe a specific named client in the
case that the client's subscription credentials expire or become invalid
PH> Does people (and especially binding authors) prefer:
PH> (a) starting a 4.0 version for this work (which means that 3.x will
PH> stop when 3.0.x is mature)
PH> (b) doing this work in 3.1, and changing the release policies to allow this.
I don't have a strong investment in IDENTITY, but this h
On Mon, Jul 18, 2011 at 8:32 PM, Marcus Cavanaugh
wrote:
> Our use case for ZMQ_IDENTITY is this: We have a few stateful backend
> services that need to communicate with other servers. We need to direct
> requests to specific services, so round-robin balancing doesn't work. (We
> can't easily mak
Did one of the patches just add this? I'm just noticing a lot of conflicts
with a bundled version for C99 compatibility.
/*
* Copyright (c) 1992-2009 by P.J. Plauger. ALL RIGHTS RESERVED.
* Consult your license regarding permissions and restrictions.
V5.20:0009 */
--
Steve-o
If you use a payload encryption, you could pass out shared decryption keys
using a req/rep. If the filtering needed to change, you could detect it
(decryption fails) and the clients would then request a new decryption key. The
ones that are not allowed to see that data would be rejected.
Joshu
I don't think the numbering actually matters (unless you really like the number
3). Numbering seems to be more of a marketing/political issue if it ever does
matter. I suspect the topic of breaking backwards compatibility and how to
handle it is more important. We will have this when we move fro
>From my perspective, the bottom line is: "Just decide what you guys
want to do and stick to it"
Because that will be easier to explain in my docs for my users.
So in that sense, while I don't care which way or the other engineering-wise,
if you've already released 3.x with IDENTITY (whoops), I'd
On Mon, Jul 18, 2011 at 12:37 PM, Marc Ilgen wrote:
> Hello
> Suppose I set up a client/server system with SUB and PUB sockets. I want
> the
> client to have to authenticate its ability to subscribe and I want the
> server to have the ability to unsubscribe a specific named client in the
> case t
27 matches
Mail list logo
