Re: [zeromq-dev] Moving list to Google groups?

[Jim Idle]

> On 14.04.2016 09:41, Pieter Hintjens wrote:
> >> So an added criteria for any new platform is ability to import our
> archives.
> >>
> >> Free is a plus iff the hosting business is long term reliable.
> >>
> >> Conventional email interface is a plus.
> >>
>

https://developers.google.com/admin-sdk/groups-migration/v1/guides/manage-email-migrations#group_migrate_media_upload

Seems to be mixed emotions on adopting Google groups, but it does have a
conventional email interface, you do not have to create a google account to
use the email, and it supports imports.
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Moving list to Google groups?

[alex.]

I have to take back my vote for librelist. By chance I read some Mails 
from Flask (a python web-framework) which used to host their 
mailing-lists there and decided to move to mail.python.org. Turns out 
librelist didn't use mailman and some users were unable to subscribe to 
the list, while Zed Shaw seems to have abandoned his librelist project.[1]

So bugs, non-conventional software and halting development throws 
librelist out of the race.

Another thing: Personally I couldn't find a mailman hoster that's free 
for opensource projects. Unless you'd count riseup.[2]

[1]: 
http://librelist.com/browser//flask/2015/7/14/switch-mailing-list-software-away-from-librelist/
[2]: https://lists.riseup.net/

Am 14.04.2016 um 11:54 schrieb alex.:
> Also a 'nay' on Google Groups from me.
>
> And I also concur with Lionel Orry: Librelist is just awesome.
> Personally I've had the best mailing list experience with lists hosted
> their - although that might just say something about the quality of the
> projects hosted there.
>
> Afaik it also uses mailman and from their help page as a list owner you
> seem to have good access to archive data, so porting them over is
> probably easily doable. But you can simply send an E-Mail to
> m...@librelist.com and ask, the guy running the show is really nice.
>
> best,
> alex.
>
> On 14.04.2016 09:41, Pieter Hintjens wrote:
>> So an added criteria for any new platform is ability to import our archives.
>>
>> Free is a plus iff the hosting business is long term reliable.
>>
>> Conventional email interface is a plus.
>>
>>
>> On Thu, Apr 14, 2016 at 9:36 AM, Indradhanush Gupta
>>  wrote:
>>>
>>>
>>> On Thu, Apr 14, 2016 at 1:02 PM, Jim Idle  wrote:

 Also usually a lurker here, but I am all in favor of Google Groups.
 Generally much better than the mailing list software. Can be a change for
 some people of course. But +1 from me for what it's worth.

 Jim

 On Thu, Apr 14, 2016 at 3:05 PM, Dinu Gherman
  wrote:
>
> Hi!
>
> I'm a silent lurker here, but I would most likely stop following this
> list after a move to google groups.
>>>
>>>
>>> Agree with Dinu here. I'll probably forget about the list on google groups.
>>>
>
> I don't like imposing the use of an account already connecting to lots of
> data where any email address anywhere would be just what is needed. Sure,
> you can user other identities, but that's a pain.
>
> Unrelated to ideology, I've tried using google groups for other groups
> imposing it on their members and I did not like the "user experience" at
> all, and after a while I gave up. I don't understand how people can like 
> it.
> If it would be more like google mail, I
> might like it.
>
> I would give Mailman 3 a try which should now have an interface decades
> better than Mailman 2, which is probably literally true if you skip the
> plural s.
>
> Cheers,
>
> Dinu
>
> --
> Sent on the move.
> Von unterwegs gesendet.
>
>> Am 14.04.2016 um 00:45 schrieb Pieter Hintjens :
>>
>> Hi all,
>>
>> I'd like to raise this question for discussion. The problem is that
>> zeromq-dev is running on an iMatix server, and I'd like to remove
>> iMatix infrastructure from our community, over time. It's a matter of
>> long term sustainability.
>>
>> (Apart from the list, iMatix also hosts downloads.zeromq.org which we
>> can start to move to Github release attachments IMO.)
>>
>> Anyhow, re. lists, let's weigh pros and cons of Google groups versus
>> other options. I'll make the decision based on this thread and other
>> factors.
>>
>> Thanks
>> Pieter
>> ___
>> zeromq-dev mailing list
>> zeromq-dev@lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev



 ___
 zeromq-dev mailing list
 zeromq-dev@lists.zeromq.org
 http://lists.zeromq.org/mailman/listinfo/zeromq-dev

>>>
>>>
>>>
>>> --
>>> Indradhanush Gupta
>>>
>>>
>>> ___
>>> zeromq-dev mailing list
>>> zeromq-dev@lists.zeromq.org
>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>>
>> ___
>> zeromq-dev mailing list
>> zeromq-dev@lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/l

Re: [zeromq-dev] Moving list to Google groups?

[Jim Idle]

You just add your email address to the google list - you don't actually
need a Google account. Thre is really no difference to now.

On Thu, Apr 14, 2016 at 5:46 PM, Jeff Shanab  wrote:

> I am also a mostly silent lurker and will probably not follow a google
> group. :-(
>
> On Thu, Apr 14, 2016 at 3:05 AM, Dinu Gherman  > wrote:
>
>> Hi!
>>
>> I'm a silent lurker here, but I would most likely stop following this
>> list after a move to google groups. I don't like imposing the use of an
>> account already connecting to lots of data where any email address anywhere
>> would be just what is needed. Sure, you can user other identities, but
>> that's a pain.
>>
>> Unrelated to ideology, I've tried using google groups for other groups
>> imposing it on their members and I did not like the "user experience" at
>> all, and after a while I gave up. I don't understand how people can like
>> it. If it would be more like google mail, I
>> might like it.
>>
>> I would give Mailman 3 a try which should now have an interface decades
>> better than Mailman 2, which is probably literally true if you skip the
>> plural s.
>>
>> Cheers,
>>
>> Dinu
>>
>> --
>> Sent on the move.
>> Von unterwegs gesendet.
>>
>> > Am 14.04.2016 um 00:45 schrieb Pieter Hintjens :
>> >
>> > Hi all,
>> >
>> > I'd like to raise this question for discussion. The problem is that
>> > zeromq-dev is running on an iMatix server, and I'd like to remove
>> > iMatix infrastructure from our community, over time. It's a matter of
>> > long term sustainability.
>> >
>> > (Apart from the list, iMatix also hosts downloads.zeromq.org which we
>> > can start to move to Github release attachments IMO.)
>> >
>> > Anyhow, re. lists, let's weigh pros and cons of Google groups versus
>> > other options. I'll make the decision based on this thread and other
>> > factors.
>> >
>> > Thanks
>> > Pieter
>> > ___
>> > zeromq-dev mailing list
>> > zeromq-dev@lists.zeromq.org
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> ___
>> zeromq-dev mailing list
>> zeromq-dev@lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
>
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] build problem after mlm_client.xml change

[Matjaž Ostroveršnik]

In the mean time I found out that it is not necessary to change 
anything. Just do gsl on mlm_client.xml (started from src folder as 
required by Pieter).
I am working on ostri/malamute, but it should be in sync with 
zeromq/malamute.


On 14.4.2016 16:33, Osiris Pedroso wrote:

Hi Matjaz,

Can you provide a repository with the modified file checked in?

I would like to try it as well.

Thanks,
Osiris

On Thu, Apr 14, 2016 at 9:12 AM Matjaž Ostroveršnik 
mailto:matjaz.ostrovers...@gmail.com>> 
wrote:


Hi,

I tried to add a new method to the mlm_client.xml and then
regenerate all dependant files.
Generation of include/mlm_client.h is unsuccesful (i.e. one gets
xml garbage within the c header sources)

Am I doing something wrong or this is a bug?
Guys how do you regenerate when you change some xml file (ok I
know for project.xml)
I think it is something wrong with an inclusion of custom file

It seems that gsl is including header tag instead of source tag.

How to reproduce:
*git reset --hard*
HEAD is now at c097dcf Merge pull request #164 from
opedroso/WIN_CRLF_REGEN
*git status*
On branch master
Your branch is up-to-date with 'origin/master'.
nothing to commit, working directory clean
*cd src*
*gsl mlm_client.xml **
*GSL/4.1c Copyright (c) 1996-2016 iMatix Corporation
gsl/4 I: Processing mlm_client.xml...
*cd ..**
**make**
*Making all in doc
make[1]: Entering directory '/home/ostri/zstuff/malamute/doc'
make[1]: Nothing to be done for 'all'.
make[1]: Leaving directory '/home/ostri/zstuff/malamute/doc'
make[1]: Entering directory '/home/ostri/zstuff/malamute'
  CC   src/src_libmlm_la-mlm_msg.lo
In file included from src/../include/mlm_library.h:66:0,
 from src/../include/malamute.h:17,
 from src/mlm_classes.h:24,
 from src/mlm_msg.c:20:
src/../include/mlm_client.h:157:5: error: expected identifier or
‘(’ before ‘<’ token
 
 ^
src/../include/mlm_client.h:159:61: error: identifier "or" is a
special operator name in C++ [-Werror=c++-compat]
 Returns 0 if OK, -1 if failed due to lack of memory or
other error.
^
src/../include/mlm_client.h:167:61: error: identifier "or" is a
special operator name in C++ [-Werror=c++-compat]
 Returns 0 if OK, -1 if failed due to lack of memory or
other error.
^
src/../include/mlm_client.h:176:61: error: identifier "or" is a
special operator name in C++ [-Werror=c++-compat]
 Returns 0 if OK, -1 if failed due to lack of memory or
other error.
^
src/../include/mlm_client.h:184:27: error: identifier "and" is a
special operator name in C++ [-Werror=c++-compat]
 Receive a subject and string content from the server. The
content may be
   ^
src/../include/mlm_client.h:185:11: error: identifier "or" is a
special operator name in C++ [-Werror=c++-compat]
 1 or more string frames. This method is orthogonal to the
sendx methods.
   ^
src/../include/mlm_client.h:186:58: error: identifier "not" is a
special operator name in C++ [-Werror=c++-compat]
 End the string arguments with NULL. If there are not
enough frames in
  ^
src/../include/mlm_client.h:188:38: error: identifier "or" is a
special operator name in C++ [-Werror=c++-compat]
 of string contents received, or -1 in case of error. Free
the returned
  ^
src/../include/mlm_client.h:189:17: error: identifier "and" is a
special operator name in C++ [-Werror=c++-compat]
 subject and content strings when finished with them. To
get the type of
 ^
cc1: all warnings being treated as errors
Makefile:1027: recipe for target 'src/src_libmlm_la-mlm_msg.lo' failed
make[1]: *** [src/src_libmlm_la-mlm_msg.lo] Error 1
make[1]: Leaving directory '/home/ostri/zstuff/malamute'
Makefile:1260: recipe for target 'all-recursive' failed
make: *** [all-recursive] Error 1

*gsl -v**
*GSL/4.1c
Built from http://github.com/imatix/gsl.git master
Copyright (c) 1996-2016 iMatix Corporation
Compiler: gcc -c -O2 -D_REENTRANT -D_GNU_SOURCE -Wall -Wno-unused
-fno-strict-aliasing -DBASE_THREADSAFE -I. gsl.c



Best regards

Matjaž

___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org 
http://lists.zeromq.org/mailman/listinfo/zeromq-dev



___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev


___
zeromq-dev m

Re: [zeromq-dev] build problem after mlm_client.xml change

[Osiris Pedroso]

Hi Matjaz,

Can you provide a repository with the modified file checked in?

I would like to try it as well.

Thanks,
Osiris

On Thu, Apr 14, 2016 at 9:12 AM Matjaž Ostroveršnik <
matjaz.ostrovers...@gmail.com> wrote:

> Hi,
>
> I tried to add a new method to the mlm_client.xml and then regenerate all
> dependant files.
> Generation of include/mlm_client.h is unsuccesful (i.e. one gets xml
> garbage within the c header sources)
>
> Am I doing something wrong or this is a bug?
> Guys how do you regenerate when you change some xml file (ok I know for
> project.xml)
> I think it is something wrong with an inclusion of custom file
> 
> It seems that gsl is including header tag instead of source tag.
>
> How to reproduce:
> *git reset --hard*
> HEAD is now at c097dcf Merge pull request #164 from opedroso/WIN_CRLF_REGEN
> *git status*
> On branch master
> Your branch is up-to-date with 'origin/master'.
> nothing to commit, working directory clean
> *cd src*
> *gsl mlm_client.xml *
> GSL/4.1c Copyright (c) 1996-2016 iMatix Corporation
> gsl/4 I: Processing mlm_client.xml...
> *cd ..*
> *make*
> Making all in doc
> make[1]: Entering directory '/home/ostri/zstuff/malamute/doc'
> make[1]: Nothing to be done for 'all'.
> make[1]: Leaving directory '/home/ostri/zstuff/malamute/doc'
> make[1]: Entering directory '/home/ostri/zstuff/malamute'
>   CC   src/src_libmlm_la-mlm_msg.lo
> In file included from src/../include/mlm_library.h:66:0,
>  from src/../include/malamute.h:17,
>  from src/mlm_classes.h:24,
>  from src/mlm_msg.c:20:
> src/../include/mlm_client.h:157:5: error: expected identifier or ‘(’
> before ‘<’ token
>  
>  ^
> src/../include/mlm_client.h:159:61: error: identifier "or" is a special
> operator name in C++ [-Werror=c++-compat]
>  Returns 0 if OK, -1 if failed due to lack of memory or other
> error.
>  ^
> src/../include/mlm_client.h:167:61: error: identifier "or" is a special
> operator name in C++ [-Werror=c++-compat]
>  Returns 0 if OK, -1 if failed due to lack of memory or other
> error.
>  ^
> src/../include/mlm_client.h:176:61: error: identifier "or" is a special
> operator name in C++ [-Werror=c++-compat]
>  Returns 0 if OK, -1 if failed due to lack of memory or other
> error.
>  ^
> src/../include/mlm_client.h:184:27: error: identifier "and" is a special
> operator name in C++ [-Werror=c++-compat]
>  Receive a subject and string content from the server. The content
> may be
>^
> src/../include/mlm_client.h:185:11: error: identifier "or" is a special
> operator name in C++ [-Werror=c++-compat]
>  1 or more string frames. This method is orthogonal to the sendx
> methods.
>^
> src/../include/mlm_client.h:186:58: error: identifier "not" is a special
> operator name in C++ [-Werror=c++-compat]
>  End the string arguments with NULL. If there are not enough
> frames in
>   ^
> src/../include/mlm_client.h:188:38: error: identifier "or" is a special
> operator name in C++ [-Werror=c++-compat]
>  of string contents received, or -1 in case of error. Free the
> returned
>   ^
> src/../include/mlm_client.h:189:17: error: identifier "and" is a special
> operator name in C++ [-Werror=c++-compat]
>  subject and content strings when finished with them. To get the
> type of
>  ^
> cc1: all warnings being treated as errors
> Makefile:1027: recipe for target 'src/src_libmlm_la-mlm_msg.lo' failed
> make[1]: *** [src/src_libmlm_la-mlm_msg.lo] Error 1
> make[1]: Leaving directory '/home/ostri/zstuff/malamute'
> Makefile:1260: recipe for target 'all-recursive' failed
> make: *** [all-recursive] Error 1
>
> *gsl -v*
> GSL/4.1c
> Built from http://github.com/imatix/gsl.git master
> Copyright (c) 1996-2016 iMatix Corporation
> Compiler: gcc -c -O2 -D_REENTRANT -D_GNU_SOURCE -Wall -Wno-unused
> -fno-strict-aliasing -DBASE_THREADSAFE -I. gsl.c
>
>
>
> Best regards
>
> Matjaž
>
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Using GSL to design an FSM protocol and translate it into a CSP protocol

[alex.]

Thanks Bjorn,

that seems exactly like what I was looking for. And if I do it
correctly, I should be able to get a GSL template working that
translates (simple) FSMs into TLA+. If I ever come to that stage I'd be
glad to share it.

Seems kinda cool in my head: building nice networking protocols with
GSL, ready to be integrated into ZeroMQ with zproto, and checked for
validity and security by the TLA toolbox.

Well, we'll see.
Best,
alex.

On 14.04.2016 12:21, Bjorn Reese wrote:
> On 04/14/2016 11:30 AM, alex. wrote:
> 
>> Anyway after doing a little more research some papers indicated that a
>> better approach for writing communication protocols would be
>> communicating sequential processes (CSP). But they are somewhat of an
>> antithesis to FSMs: their basic elements are processes whereas in FSMs
>> the basic elements are states. Now I could brush CSPs aside and just go
> 
> You may want to look into Lamport's TLA+.
> 
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> 
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

[zeromq-dev] build problem after mlm_client.xml change

[Matjaž Ostroveršnik]

Hi,

I tried to add a new method to the mlm_client.xml and then regenerate 
all dependant files.
Generation of include/mlm_client.h is unsuccesful (i.e. one gets xml 
garbage within the c header sources)


Am I doing something wrong or this is a bug?
Guys how do you regenerate when you change some xml file (ok I know for 
project.xml)

I think it is something wrong with an inclusion of custom file

It seems that gsl is including header tag instead of source tag.

How to reproduce:
*git reset --hard*
HEAD is now at c097dcf Merge pull request #164 from opedroso/WIN_CRLF_REGEN
*git status*
On branch master
Your branch is up-to-date with 'origin/master'.
nothing to commit, working directory clean
*cd src*
*gsl mlm_client.xml **
*GSL/4.1c Copyright (c) 1996-2016 iMatix Corporation
gsl/4 I: Processing mlm_client.xml...
*cd ..**
**make**
*Making all in doc
make[1]: Entering directory '/home/ostri/zstuff/malamute/doc'
make[1]: Nothing to be done for 'all'.
make[1]: Leaving directory '/home/ostri/zstuff/malamute/doc'
make[1]: Entering directory '/home/ostri/zstuff/malamute'
  CC   src/src_libmlm_la-mlm_msg.lo
In file included from src/../include/mlm_library.h:66:0,
 from src/../include/malamute.h:17,
 from src/mlm_classes.h:24,
 from src/mlm_msg.c:20:
src/../include/mlm_client.h:157:5: error: expected identifier or ‘(’ 
before ‘<’ token

 
 ^
src/../include/mlm_client.h:159:61: error: identifier "or" is a special 
operator name in C++ [-Werror=c++-compat]
 Returns 0 if OK, -1 if failed due to lack of memory or other 
error.

 ^
src/../include/mlm_client.h:167:61: error: identifier "or" is a special 
operator name in C++ [-Werror=c++-compat]
 Returns 0 if OK, -1 if failed due to lack of memory or other 
error.

 ^
src/../include/mlm_client.h:176:61: error: identifier "or" is a special 
operator name in C++ [-Werror=c++-compat]
 Returns 0 if OK, -1 if failed due to lack of memory or other 
error.

 ^
src/../include/mlm_client.h:184:27: error: identifier "and" is a special 
operator name in C++ [-Werror=c++-compat]
 Receive a subject and string content from the server. The 
content may be

   ^
src/../include/mlm_client.h:185:11: error: identifier "or" is a special 
operator name in C++ [-Werror=c++-compat]
 1 or more string frames. This method is orthogonal to the 
sendx methods.

   ^
src/../include/mlm_client.h:186:58: error: identifier "not" is a special 
operator name in C++ [-Werror=c++-compat]
 End the string arguments with NULL. If there are not enough 
frames in

  ^
src/../include/mlm_client.h:188:38: error: identifier "or" is a special 
operator name in C++ [-Werror=c++-compat]
 of string contents received, or -1 in case of error. Free the 
returned

  ^
src/../include/mlm_client.h:189:17: error: identifier "and" is a special 
operator name in C++ [-Werror=c++-compat]
 subject and content strings when finished with them. To get 
the type of

 ^
cc1: all warnings being treated as errors
Makefile:1027: recipe for target 'src/src_libmlm_la-mlm_msg.lo' failed
make[1]: *** [src/src_libmlm_la-mlm_msg.lo] Error 1
make[1]: Leaving directory '/home/ostri/zstuff/malamute'
Makefile:1260: recipe for target 'all-recursive' failed
make: *** [all-recursive] Error 1

*gsl -v**
*GSL/4.1c
Built from http://github.com/imatix/gsl.git master
Copyright (c) 1996-2016 iMatix Corporation
Compiler: gcc -c -O2 -D_REENTRANT -D_GNU_SOURCE -Wall -Wno-unused 
-fno-strict-aliasing -DBASE_THREADSAFE -I. gsl.c




Best regards

Matjaž

___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Moving list to Google groups?

[Kyle Kelley]

I'm in favor of Google groups, relative to the alternatives.

On Thursday, April 14, 2016, Greg Young  wrote:

> Google groups is a mediocre solution given all the alternatives but
> its also quite commonly used so most people are willing to deal with
> its warts. The history of the group would be important to bring over
> if possible (I do not know of an api for bringing it over).
>
> On Thu, Apr 14, 2016 at 1:45 AM, Pieter Hintjens  > wrote:
> > Hi all,
> >
> > I'd like to raise this question for discussion. The problem is that
> > zeromq-dev is running on an iMatix server, and I'd like to remove
> > iMatix infrastructure from our community, over time. It's a matter of
> > long term sustainability.
> >
> > (Apart from the list, iMatix also hosts downloads.zeromq.org which we
> > can start to move to Github release attachments IMO.)
> >
> > Anyhow, re. lists, let's weigh pros and cons of Google groups versus
> > other options. I'll make the decision based on this thread and other
> > factors.
> >
> > Thanks
> > Pieter
> > ___
> > zeromq-dev mailing list
> > zeromq-dev@lists.zeromq.org 
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
>
> --
> Studying for the Turing test
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org 
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>


-- 
Kyle Kelley (@rgbkrk ; lambdaops.com)
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Moving list to Google groups?

[Greg Young]

Google groups is a mediocre solution given all the alternatives but
its also quite commonly used so most people are willing to deal with
its warts. The history of the group would be important to bring over
if possible (I do not know of an api for bringing it over).

On Thu, Apr 14, 2016 at 1:45 AM, Pieter Hintjens  wrote:
> Hi all,
>
> I'd like to raise this question for discussion. The problem is that
> zeromq-dev is running on an iMatix server, and I'd like to remove
> iMatix infrastructure from our community, over time. It's a matter of
> long term sustainability.
>
> (Apart from the list, iMatix also hosts downloads.zeromq.org which we
> can start to move to Github release attachments IMO.)
>
> Anyhow, re. lists, let's weigh pros and cons of Google groups versus
> other options. I'll make the decision based on this thread and other
> factors.
>
> Thanks
> Pieter
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev



-- 
Studying for the Turing test
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Memory management

[Jeff Shanab]

Thanks I think I have some Ideas to try,

On Thu, Apr 14, 2016 at 7:08 AM, Auer, Jens  wrote:

> Hi,
>
> there is zmq_msg_init_data which gets a callback which is called when the
> message is released. However, it is not fully zero-copy because some
> copying is done when receiving messages. With 4.2, this is reduced, but
> until that, the data is received into a fixed buffer and then copied into
> the message.
>
> That all reminds me that I wanted to add the possibility to use custom
> allocator functions to be able to use a memory pool. I think it should be
> possible, e.g. by setting function pointers as zmq context options, plus
> void* hint that can be set too.
>
> Cheers,
>   Jens
>
> --
> *Jens Auer *| CGI | Software-Engineer
> CGI (Germany) GmbH & Co. KG
> Rheinstraße 95 | 64295 Darmstadt | Germany
> T: +49 6151 36860 154
> *jens.a...@cgi.com* 
> Unsere Pflichtangaben gemäß § 35a GmbHG / §§ 161, 125a HGB finden Sie unter
> *de.cgi.com/pflichtangaben* .
>
> CONFIDENTIALITY NOTICE: Proprietary/Confidential information belonging to
> CGI Group Inc. and its affiliates may be contained in this message. If you
> are not a recipient indicated or intended in this message (or responsible
> for delivery of this message to such person), or you think for any reason
> that this message may have been addressed to you in error, you may not use
> or copy or deliver this message to anyone else. In such case, you should
> destroy this message and are asked to notify the sender by reply e-mail.
> --
> *Von:* zeromq-dev-boun...@lists.zeromq.org [
> zeromq-dev-boun...@lists.zeromq.org]" im Auftrag von "Jeff Shanab [
> jsha...@jfs-tech.com]
> *Gesendet:* Donnerstag, 14. April 2016 12:06
> *An:* ZeroMQ development list
> *Betreff:* [zeromq-dev] Memory management
>
> I am writing an app in C++ and using libzmq directly, not much of the c++
> wrapper stuff.
>
> I am interested in zero copy or at least minimizing the number of copies
> of large buffers.
>
> But am running across an all too common issue, Each library used wants to
> have their own memory management forcing me to copy between the systems.
>
> Is there a way I could use boost shared_ptr with the functions
> so ZMQ could dereference/delete after the send?
>
> In fact is there a way to hook into the system and get a callback when it
> goes to delete with the client pointer so I can delete or return it to a
> pool?
>
>
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Memory management

[Auer, Jens]

Hi,

there is zmq_msg_init_data which gets a callback which is called when the 
message is released. However, it is not fully zero-copy because some copying is 
done when receiving messages. With 4.2, this is reduced, but until that, the 
data is received into a fixed buffer and then copied into the message.

That all reminds me that I wanted to add the possibility to use custom 
allocator functions to be able to use a memory pool. I think it should be 
possible, e.g. by setting function pointers as zmq context options, plus void* 
hint that can be set too.

Cheers,
  Jens

--
Jens Auer | CGI | Software-Engineer
CGI (Germany) GmbH & Co. KG
Rheinstraße 95 | 64295 Darmstadt | Germany
T: +49 6151 36860 154
jens.a...@cgi.com
Unsere Pflichtangaben gemäß § 35a GmbHG / §§ 161, 125a HGB finden Sie unter 
de.cgi.com/pflichtangaben.

CONFIDENTIALITY NOTICE: Proprietary/Confidential information belonging to CGI 
Group Inc. and its affiliates may be contained in this message. If you are not 
a recipient indicated or intended in this message (or responsible for delivery 
of this message to such person), or you think for any reason that this message 
may have been addressed to you in error, you may not use or copy or deliver 
this message to anyone else. In such case, you should destroy this message and 
are asked to notify the sender by reply e-mail.

Von: zeromq-dev-boun...@lists.zeromq.org [zeromq-dev-boun...@lists.zeromq.org]" 
im Auftrag von "Jeff Shanab [jsha...@jfs-tech.com]
Gesendet: Donnerstag, 14. April 2016 12:06
An: ZeroMQ development list
Betreff: [zeromq-dev] Memory management

I am writing an app in C++ and using libzmq directly, not much of the c++ 
wrapper stuff.

I am interested in zero copy or at least minimizing the number of copies of 
large buffers.

But am running across an all too common issue, Each library used wants to have 
their own memory management forcing me to copy between the systems.

Is there a way I could use boost shared_ptr with the functions so 
ZMQ could dereference/delete after the send?

In fact is there a way to hook into the system and get a callback when it goes 
to delete with the client pointer so I can delete or return it to a pool?

___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Server (router) stops receiving messages after awhile

[Luka Napotnik]

I created a minimal test case but couldn't reproduce the problem:
http://pastie.org/private/e9i8ikon6evtxfvbs5b4nq
The code is in go and I tag every request with an ID so the replies
match to the requests. I tried the handover flag but it doesn't help.

On Thu, Apr 14, 2016 at 12:34 PM, Doron Somech  wrote:
> Are you using identities for the clients? If so try enabling router
> handover.
>
> On Apr 14, 2016 12:13, "Pieter Hintjens"  wrote:
>>
>> Can you make a minimal test case that reproduces this?
>>
>>
>> On Thu, Apr 14, 2016 at 10:06 AM, Luka Napotnik
>>  wrote:
>> > Hello,
>> >
>> > my client sends requests (as a DEALER socket) and the server (a ROUTER
>> > socket) handles the request and sends the reply back (taking the route
>> > path into account) . Now, at start, everything works but after awhile,
>> > messages, sent from the client, don't get recv()-ed by the server
>> > anymore. What am I missing? The strangest thing is that the
>> > communication works for awhile. I've installed a monitor on both ends
>> > and there are no events just before things starting to fail.
>> >
>> > Please help.
>> >
>> > --
>> > Luka Napotnik, Platform Developer, Ukmarjeva ulica 2, 1000 Ljubljana,
>> > Slovenia, Europe, www.visionect.com
>> > ___
>> > zeromq-dev mailing list
>> > zeromq-dev@lists.zeromq.org
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> ___
>> zeromq-dev mailing list
>> zeromq-dev@lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



-- 
Luka Napotnik, Lead Platform Developer, Ukmarjeva ulica 2, 1000
Ljubljana, Slovenia, Europe, www.visionect.com
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Server (router) stops receiving messages after awhile

[Doron Somech]

Are you using identities for the clients? If so try enabling router
handover.
On Apr 14, 2016 12:13, "Pieter Hintjens"  wrote:

> Can you make a minimal test case that reproduces this?
>
>
> On Thu, Apr 14, 2016 at 10:06 AM, Luka Napotnik
>  wrote:
> > Hello,
> >
> > my client sends requests (as a DEALER socket) and the server (a ROUTER
> > socket) handles the request and sends the reply back (taking the route
> > path into account) . Now, at start, everything works but after awhile,
> > messages, sent from the client, don't get recv()-ed by the server
> > anymore. What am I missing? The strangest thing is that the
> > communication works for awhile. I've installed a monitor on both ends
> > and there are no events just before things starting to fail.
> >
> > Please help.
> >
> > --
> > Luka Napotnik, Platform Developer, Ukmarjeva ulica 2, 1000 Ljubljana,
> > Slovenia, Europe, www.visionect.com
> > ___
> > zeromq-dev mailing list
> > zeromq-dev@lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Using GSL to design an FSM protocol and translate it into a CSP protocol

[Bjorn Reese]

On 04/14/2016 11:30 AM, alex. wrote:

> Anyway after doing a little more research some papers indicated that a
> better approach for writing communication protocols would be
> communicating sequential processes (CSP). But they are somewhat of an
> antithesis to FSMs: their basic elements are processes whereas in FSMs
> the basic elements are states. Now I could brush CSPs aside and just go

You may want to look into Lamport's TLA+.

___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Moving list to Google groups?

[Alex Bligh]

On 14 Apr 2016, at 08:41, Pieter Hintjens  wrote:

> So an added criteria for any new platform is ability to import our archives.
> 
> Free is a plus iff the hosting business is long term reliable.
> 
> Conventional email interface is a plus.

Additional criterion (satisfied by Google Groups and many others): spam
resistance.

Lists at sourceforge that I'm on have more spam than genuine messages.
So don't go there please.

-- 
Alex Bligh




___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

[zeromq-dev] Memory management

[Jeff Shanab]

I am writing an app in C++ and using libzmq directly, not much of the c++
wrapper stuff.

I am interested in zero copy or at least minimizing the number of copies of
large buffers.

But am running across an all too common issue, Each library used wants to
have their own memory management forcing me to copy between the systems.

Is there a way I could use boost shared_ptr with the functions
so ZMQ could dereference/delete after the send?

In fact is there a way to hook into the system and get a callback when it
goes to delete with the client pointer so I can delete or return it to a
pool?
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Using GSL to design an FSM protocol and translate it into a CSP protocol

[alex.]

Haha, yeah I was expecting no one would understand my request. The
troubles of being short and concise. I also forgot to include the links
I had prepared. :(

So as a short explanation, communicating sequential processes is a model
developed by Tony Hoare in 1978 which is sort of like state machines,
but well different.[3] One book about protocol design I have open right
now is from Robin Sharp "Principles of Protocol Design" (2008) implies
that CSPs have "gained prominence within recent years" because basically
FSMs suck for parallel operations.

Again, I want to use GSL to start off both the protocol and my codebase
using state machines (so big thanks for the links Pieter and Kevin!).
But if it'll be possible to generate a CSP style protocol using GSL from
my state machine, I'd be enormously happy.

Thus my questions: Did anyone ever try something like this remotely?

Best,
alex.


P.S. The links:
[1]: http://www.cs.ox.ac.uk/gavin.lowe/Security/Casper/
[2]: https://www.cs.ox.ac.uk/projects/fdr/
[3]: https://en.wikipedia.org/wiki/Communicating_sequential_processes

On 14.04.2016 11:37, Pieter Hintjens wrote:
> I don't understand half of what you're asking :) But in any case do
> take a look at the tiny FSM / C code generator
> (https://github.com/imatix/gsl/blob/b0aa393a6d00df2859f21aab0ccf6f203e18db96/examples/fsm_c.gsl)
> that I've used happily in a few projects already.
> 
> On Thu, Apr 14, 2016 at 11:30 AM, alex.  wrote:
>> First of all I wasn't really sure if it's okay to talk about the GSL
>> Universal Code Generator on this list, but I am writing anyway since it
>> didn't have its own mailing list. So forgive any insolence. :)
>>
>>
>> I was planning on writing my thesis about a traffic analysis resistant
>> multi-user file-transfer protocol and sure enough I wanted to do the
>> proof-of-concept implementation using ZeroMQ (so, yeah, expect to hear
>> more from me over the next months ;) ). Now I haven't ever written a
>> protocol myself but from the (little) research I did, I came to the
>> conclusion that it would be best to formalize it as best as possible, in
>> particular preventing any indeterminism. So the first idea was to use
>> finite state machines for formalization and for some reason the best
>> tool I could find for writing finite state machines that also lets me
>> write test cases AND generate all kinds of code (including graphics and
>> C-bindings) was made by the same people who created ZeroMQ. What about
>> that, I thought to myself.
>>
>> Anyway after doing a little more research some papers indicated that a
>> better approach for writing communication protocols would be
>> communicating sequential processes (CSP). But they are somewhat of an
>> antithesis to FSMs: their basic elements are processes whereas in FSMs
>> the basic elements are states. Now I could brush CSPs aside and just go
>> for FSMs, but since my thesis is mostly security/cryptography related, I
>> also looked for security protocol analysers and every tool I could find
>> was using CSPs as well. In particular I found Casper[1], which uses a
>> special kind of CSP syntax to generate formalized CSP to be then checked
>> with FDR3[2].
>>
>> In short, the scenario I would love to go for is basically:
>> - writing FSM in iMatix' GSL syntax,
>> - compile it to Casper's CSP syntax,
>> - compile that to general CSP to be checked by FDR3.
>>
>>
>> My question to this fine community is thus: Did anyone ever try to
>> produce a CSP using the GSL Universal Code Generator? Is something like
>> that even achievable (without wasting large amounts of precious time)
>> using GSL?
>> Or in general, did anyone ever try to translate FSMs into CSPs? Or am I
>> just too uninformed to know that something like this is simply absurd?
>>
>> All the best,
>> alex.
>> ___
>> zeromq-dev mailing list
>> zeromq-dev@lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> 
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Moving list to Google groups?

[alex.]

Also a 'nay' on Google Groups from me.

And I also concur with Lionel Orry: Librelist is just awesome.
Personally I've had the best mailing list experience with lists hosted
their - although that might just say something about the quality of the
projects hosted there.

Afaik it also uses mailman and from their help page as a list owner you
seem to have good access to archive data, so porting them over is
probably easily doable. But you can simply send an E-Mail to
m...@librelist.com and ask, the guy running the show is really nice.

best,
alex.

On 14.04.2016 09:41, Pieter Hintjens wrote:
> So an added criteria for any new platform is ability to import our archives.
> 
> Free is a plus iff the hosting business is long term reliable.
> 
> Conventional email interface is a plus.
> 
> 
> On Thu, Apr 14, 2016 at 9:36 AM, Indradhanush Gupta
>  wrote:
>>
>>
>> On Thu, Apr 14, 2016 at 1:02 PM, Jim Idle  wrote:
>>>
>>> Also usually a lurker here, but I am all in favor of Google Groups.
>>> Generally much better than the mailing list software. Can be a change for
>>> some people of course. But +1 from me for what it's worth.
>>>
>>> Jim
>>>
>>> On Thu, Apr 14, 2016 at 3:05 PM, Dinu Gherman
>>>  wrote:

 Hi!

 I'm a silent lurker here, but I would most likely stop following this
 list after a move to google groups.
>>
>>
>> Agree with Dinu here. I'll probably forget about the list on google groups.
>>

 I don't like imposing the use of an account already connecting to lots of
 data where any email address anywhere would be just what is needed. Sure,
 you can user other identities, but that's a pain.

 Unrelated to ideology, I've tried using google groups for other groups
 imposing it on their members and I did not like the "user experience" at
 all, and after a while I gave up. I don't understand how people can like 
 it.
 If it would be more like google mail, I
 might like it.

 I would give Mailman 3 a try which should now have an interface decades
 better than Mailman 2, which is probably literally true if you skip the
 plural s.

 Cheers,

 Dinu

 --
 Sent on the move.
 Von unterwegs gesendet.

> Am 14.04.2016 um 00:45 schrieb Pieter Hintjens :
>
> Hi all,
>
> I'd like to raise this question for discussion. The problem is that
> zeromq-dev is running on an iMatix server, and I'd like to remove
> iMatix infrastructure from our community, over time. It's a matter of
> long term sustainability.
>
> (Apart from the list, iMatix also hosts downloads.zeromq.org which we
> can start to move to Github release attachments IMO.)
>
> Anyhow, re. lists, let's weigh pros and cons of Google groups versus
> other options. I'll make the decision based on this thread and other
> factors.
>
> Thanks
> Pieter
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
 ___
 zeromq-dev mailing list
 zeromq-dev@lists.zeromq.org
 http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>>
>>>
>>>
>>> ___
>>> zeromq-dev mailing list
>>> zeromq-dev@lists.zeromq.org
>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>>
>>
>>
>>
>> --
>> Indradhanush Gupta
>>
>>
>> ___
>> zeromq-dev mailing list
>> zeromq-dev@lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> 
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Moving list to Google groups?

[Alex Bligh]

On 14 Apr 2016, at 10:46, Jeff Shanab  wrote:

> I am also a mostly silent lurker and will probably not follow a google group. 
> :-(  

I've no particular beef for or against Google Groups, but it is entirely 
possible
to follow them like a normal mailing list. The signup feature is hidden a 
little,
but you can get it to send every message to your by email, and you can reply by
email. I participate in several lists that work this way (the raft mailing
list being an example off the top of my head).

The main problem is that if most people use the Google Groups interface, a lot
of them seem to reply trimming all context, which makes following conversations
hard.

-- 
Alex Bligh




___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Using GSL to design an FSM protocol and translate it into a CSP protocol

[Kevin Sapper]

Also for defining protocols there is zproto (
https://github.com/zeromq/zproto) and to kick start your C project you
might check out zproject (https://github.com/zeromq/zproject). Both of them
are GSL tools so if they're not what your looking for the might help you as
an example for your on GSL project.

2016-04-14 11:37 GMT+02:00 Pieter Hintjens :

> I don't understand half of what you're asking :) But in any case do
> take a look at the tiny FSM / C code generator
> (
> https://github.com/imatix/gsl/blob/b0aa393a6d00df2859f21aab0ccf6f203e18db96/examples/fsm_c.gsl
> )
> that I've used happily in a few projects already.
>
> On Thu, Apr 14, 2016 at 11:30 AM, alex.  wrote:
> > First of all I wasn't really sure if it's okay to talk about the GSL
> > Universal Code Generator on this list, but I am writing anyway since it
> > didn't have its own mailing list. So forgive any insolence. :)
> >
> >
> > I was planning on writing my thesis about a traffic analysis resistant
> > multi-user file-transfer protocol and sure enough I wanted to do the
> > proof-of-concept implementation using ZeroMQ (so, yeah, expect to hear
> > more from me over the next months ;) ). Now I haven't ever written a
> > protocol myself but from the (little) research I did, I came to the
> > conclusion that it would be best to formalize it as best as possible, in
> > particular preventing any indeterminism. So the first idea was to use
> > finite state machines for formalization and for some reason the best
> > tool I could find for writing finite state machines that also lets me
> > write test cases AND generate all kinds of code (including graphics and
> > C-bindings) was made by the same people who created ZeroMQ. What about
> > that, I thought to myself.
> >
> > Anyway after doing a little more research some papers indicated that a
> > better approach for writing communication protocols would be
> > communicating sequential processes (CSP). But they are somewhat of an
> > antithesis to FSMs: their basic elements are processes whereas in FSMs
> > the basic elements are states. Now I could brush CSPs aside and just go
> > for FSMs, but since my thesis is mostly security/cryptography related, I
> > also looked for security protocol analysers and every tool I could find
> > was using CSPs as well. In particular I found Casper[1], which uses a
> > special kind of CSP syntax to generate formalized CSP to be then checked
> > with FDR3[2].
> >
> > In short, the scenario I would love to go for is basically:
> > - writing FSM in iMatix' GSL syntax,
> > - compile it to Casper's CSP syntax,
> > - compile that to general CSP to be checked by FDR3.
> >
> >
> > My question to this fine community is thus: Did anyone ever try to
> > produce a CSP using the GSL Universal Code Generator? Is something like
> > that even achievable (without wasting large amounts of precious time)
> > using GSL?
> > Or in general, did anyone ever try to translate FSMs into CSPs? Or am I
> > just too uninformed to know that something like this is simply absurd?
> >
> > All the best,
> > alex.
> > ___
> > zeromq-dev mailing list
> > zeromq-dev@lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Moving list to Google groups?

[Jeff Shanab]

I am also a mostly silent lurker and will probably not follow a google
group. :-(

On Thu, Apr 14, 2016 at 3:05 AM, Dinu Gherman 
wrote:

> Hi!
>
> I'm a silent lurker here, but I would most likely stop following this list
> after a move to google groups. I don't like imposing the use of an account
> already connecting to lots of data where any email address anywhere would
> be just what is needed. Sure, you can user other identities, but that's a
> pain.
>
> Unrelated to ideology, I've tried using google groups for other groups
> imposing it on their members and I did not like the "user experience" at
> all, and after a while I gave up. I don't understand how people can like
> it. If it would be more like google mail, I
> might like it.
>
> I would give Mailman 3 a try which should now have an interface decades
> better than Mailman 2, which is probably literally true if you skip the
> plural s.
>
> Cheers,
>
> Dinu
>
> --
> Sent on the move.
> Von unterwegs gesendet.
>
> > Am 14.04.2016 um 00:45 schrieb Pieter Hintjens :
> >
> > Hi all,
> >
> > I'd like to raise this question for discussion. The problem is that
> > zeromq-dev is running on an iMatix server, and I'd like to remove
> > iMatix infrastructure from our community, over time. It's a matter of
> > long term sustainability.
> >
> > (Apart from the list, iMatix also hosts downloads.zeromq.org which we
> > can start to move to Github release attachments IMO.)
> >
> > Anyhow, re. lists, let's weigh pros and cons of Google groups versus
> > other options. I'll make the decision based on this thread and other
> > factors.
> >
> > Thanks
> > Pieter
> > ___
> > zeromq-dev mailing list
> > zeromq-dev@lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Using GSL to design an FSM protocol and translate it into a CSP protocol

[Pieter Hintjens]

I don't understand half of what you're asking :) But in any case do
take a look at the tiny FSM / C code generator
(https://github.com/imatix/gsl/blob/b0aa393a6d00df2859f21aab0ccf6f203e18db96/examples/fsm_c.gsl)
that I've used happily in a few projects already.

On Thu, Apr 14, 2016 at 11:30 AM, alex.  wrote:
> First of all I wasn't really sure if it's okay to talk about the GSL
> Universal Code Generator on this list, but I am writing anyway since it
> didn't have its own mailing list. So forgive any insolence. :)
>
>
> I was planning on writing my thesis about a traffic analysis resistant
> multi-user file-transfer protocol and sure enough I wanted to do the
> proof-of-concept implementation using ZeroMQ (so, yeah, expect to hear
> more from me over the next months ;) ). Now I haven't ever written a
> protocol myself but from the (little) research I did, I came to the
> conclusion that it would be best to formalize it as best as possible, in
> particular preventing any indeterminism. So the first idea was to use
> finite state machines for formalization and for some reason the best
> tool I could find for writing finite state machines that also lets me
> write test cases AND generate all kinds of code (including graphics and
> C-bindings) was made by the same people who created ZeroMQ. What about
> that, I thought to myself.
>
> Anyway after doing a little more research some papers indicated that a
> better approach for writing communication protocols would be
> communicating sequential processes (CSP). But they are somewhat of an
> antithesis to FSMs: their basic elements are processes whereas in FSMs
> the basic elements are states. Now I could brush CSPs aside and just go
> for FSMs, but since my thesis is mostly security/cryptography related, I
> also looked for security protocol analysers and every tool I could find
> was using CSPs as well. In particular I found Casper[1], which uses a
> special kind of CSP syntax to generate formalized CSP to be then checked
> with FDR3[2].
>
> In short, the scenario I would love to go for is basically:
> - writing FSM in iMatix' GSL syntax,
> - compile it to Casper's CSP syntax,
> - compile that to general CSP to be checked by FDR3.
>
>
> My question to this fine community is thus: Did anyone ever try to
> produce a CSP using the GSL Universal Code Generator? Is something like
> that even achievable (without wasting large amounts of precious time)
> using GSL?
> Or in general, did anyone ever try to translate FSMs into CSPs? Or am I
> just too uninformed to know that something like this is simply absurd?
>
> All the best,
> alex.
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

[zeromq-dev] Using GSL to design an FSM protocol and translate it into a CSP protocol

[alex.]

First of all I wasn't really sure if it's okay to talk about the GSL
Universal Code Generator on this list, but I am writing anyway since it
didn't have its own mailing list. So forgive any insolence. :)


I was planning on writing my thesis about a traffic analysis resistant
multi-user file-transfer protocol and sure enough I wanted to do the
proof-of-concept implementation using ZeroMQ (so, yeah, expect to hear
more from me over the next months ;) ). Now I haven't ever written a
protocol myself but from the (little) research I did, I came to the
conclusion that it would be best to formalize it as best as possible, in
particular preventing any indeterminism. So the first idea was to use
finite state machines for formalization and for some reason the best
tool I could find for writing finite state machines that also lets me
write test cases AND generate all kinds of code (including graphics and
C-bindings) was made by the same people who created ZeroMQ. What about
that, I thought to myself.

Anyway after doing a little more research some papers indicated that a
better approach for writing communication protocols would be
communicating sequential processes (CSP). But they are somewhat of an
antithesis to FSMs: their basic elements are processes whereas in FSMs
the basic elements are states. Now I could brush CSPs aside and just go
for FSMs, but since my thesis is mostly security/cryptography related, I
also looked for security protocol analysers and every tool I could find
was using CSPs as well. In particular I found Casper[1], which uses a
special kind of CSP syntax to generate formalized CSP to be then checked
with FDR3[2].

In short, the scenario I would love to go for is basically:
- writing FSM in iMatix' GSL syntax,
- compile it to Casper's CSP syntax,
- compile that to general CSP to be checked by FDR3.


My question to this fine community is thus: Did anyone ever try to
produce a CSP using the GSL Universal Code Generator? Is something like
that even achievable (without wasting large amounts of precious time)
using GSL?
Or in general, did anyone ever try to translate FSMs into CSPs? Or am I
just too uninformed to know that something like this is simply absurd?

All the best,
alex.
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Server (router) stops receiving messages after awhile

[Pieter Hintjens]

Can you make a minimal test case that reproduces this?


On Thu, Apr 14, 2016 at 10:06 AM, Luka Napotnik
 wrote:
> Hello,
>
> my client sends requests (as a DEALER socket) and the server (a ROUTER
> socket) handles the request and sends the reply back (taking the route
> path into account) . Now, at start, everything works but after awhile,
> messages, sent from the client, don't get recv()-ed by the server
> anymore. What am I missing? The strangest thing is that the
> communication works for awhile. I've installed a monitor on both ends
> and there are no events just before things starting to fail.
>
> Please help.
>
> --
> Luka Napotnik, Platform Developer, Ukmarjeva ulica 2, 1000 Ljubljana,
> Slovenia, Europe, www.visionect.com
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

[zeromq-dev] Server (router) stops receiving messages after awhile

[Luka Napotnik]

Hello,

my client sends requests (as a DEALER socket) and the server (a ROUTER
socket) handles the request and sends the reply back (taking the route
path into account) . Now, at start, everything works but after awhile,
messages, sent from the client, don't get recv()-ed by the server
anymore. What am I missing? The strangest thing is that the
communication works for awhile. I've installed a monitor on both ends
and there are no events just before things starting to fail.

Please help.

-- 
Luka Napotnik, Platform Developer, Ukmarjeva ulica 2, 1000 Ljubljana,
Slovenia, Europe, www.visionect.com
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Moving list to Google groups?

[Pieter Hintjens]

So an added criteria for any new platform is ability to import our archives.

Free is a plus iff the hosting business is long term reliable.

Conventional email interface is a plus.


On Thu, Apr 14, 2016 at 9:36 AM, Indradhanush Gupta
 wrote:
>
>
> On Thu, Apr 14, 2016 at 1:02 PM, Jim Idle  wrote:
>>
>> Also usually a lurker here, but I am all in favor of Google Groups.
>> Generally much better than the mailing list software. Can be a change for
>> some people of course. But +1 from me for what it's worth.
>>
>> Jim
>>
>> On Thu, Apr 14, 2016 at 3:05 PM, Dinu Gherman
>>  wrote:
>>>
>>> Hi!
>>>
>>> I'm a silent lurker here, but I would most likely stop following this
>>> list after a move to google groups.
>
>
> Agree with Dinu here. I'll probably forget about the list on google groups.
>
>>>
>>> I don't like imposing the use of an account already connecting to lots of
>>> data where any email address anywhere would be just what is needed. Sure,
>>> you can user other identities, but that's a pain.
>>>
>>> Unrelated to ideology, I've tried using google groups for other groups
>>> imposing it on their members and I did not like the "user experience" at
>>> all, and after a while I gave up. I don't understand how people can like it.
>>> If it would be more like google mail, I
>>> might like it.
>>>
>>> I would give Mailman 3 a try which should now have an interface decades
>>> better than Mailman 2, which is probably literally true if you skip the
>>> plural s.
>>>
>>> Cheers,
>>>
>>> Dinu
>>>
>>> --
>>> Sent on the move.
>>> Von unterwegs gesendet.
>>>
>>> > Am 14.04.2016 um 00:45 schrieb Pieter Hintjens :
>>> >
>>> > Hi all,
>>> >
>>> > I'd like to raise this question for discussion. The problem is that
>>> > zeromq-dev is running on an iMatix server, and I'd like to remove
>>> > iMatix infrastructure from our community, over time. It's a matter of
>>> > long term sustainability.
>>> >
>>> > (Apart from the list, iMatix also hosts downloads.zeromq.org which we
>>> > can start to move to Github release attachments IMO.)
>>> >
>>> > Anyhow, re. lists, let's weigh pros and cons of Google groups versus
>>> > other options. I'll make the decision based on this thread and other
>>> > factors.
>>> >
>>> > Thanks
>>> > Pieter
>>> > ___
>>> > zeromq-dev mailing list
>>> > zeromq-dev@lists.zeromq.org
>>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>> ___
>>> zeromq-dev mailing list
>>> zeromq-dev@lists.zeromq.org
>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>>
>>
>> ___
>> zeromq-dev mailing list
>> zeromq-dev@lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
>
>
> --
> Indradhanush Gupta
>
>
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Moving list to Google groups?

[Indradhanush Gupta]

On Thu, Apr 14, 2016 at 1:02 PM, Jim Idle  wrote:

> Also usually a lurker here, but I am all in favor of Google Groups.
> Generally much better than the mailing list software. Can be a change for
> some people of course. But +1 from me for what it's worth.
>
> Jim
>
> On Thu, Apr 14, 2016 at 3:05 PM, Dinu Gherman  > wrote:
>
>> Hi!
>>
>> I'm a silent lurker here, but I would most likely stop following this
>> list after a move to google groups.
>
>
Agree with Dinu here. I'll probably forget about the list on google groups.


> I don't like imposing the use of an account already connecting to lots of
>> data where any email address anywhere would be just what is needed. Sure,
>> you can user other identities, but that's a pain.
>>
>> Unrelated to ideology, I've tried using google groups for other groups
>> imposing it on their members and I did not like the "user experience" at
>> all, and after a while I gave up. I don't understand how people can like
>> it. If it would be more like google mail, I
>> might like it.
>>
>> I would give Mailman 3 a try which should now have an interface decades
>> better than Mailman 2, which is probably literally true if you skip the
>> plural s.
>>
>> Cheers,
>>
>> Dinu
>>
>> --
>> Sent on the move.
>> Von unterwegs gesendet.
>>
>> > Am 14.04.2016 um 00:45 schrieb Pieter Hintjens :
>> >
>> > Hi all,
>> >
>> > I'd like to raise this question for discussion. The problem is that
>> > zeromq-dev is running on an iMatix server, and I'd like to remove
>> > iMatix infrastructure from our community, over time. It's a matter of
>> > long term sustainability.
>> >
>> > (Apart from the list, iMatix also hosts downloads.zeromq.org which we
>> > can start to move to Github release attachments IMO.)
>> >
>> > Anyhow, re. lists, let's weigh pros and cons of Google groups versus
>> > other options. I'll make the decision based on this thread and other
>> > factors.
>> >
>> > Thanks
>> > Pieter
>> > ___
>> > zeromq-dev mailing list
>> > zeromq-dev@lists.zeromq.org
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> ___
>> zeromq-dev mailing list
>> zeromq-dev@lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
>
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>


-- 
Indradhanush Gupta
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Moving list to Google groups?

[Jim Idle]

Also usually a lurker here, but I am all in favor of Google Groups.
Generally much better than the mailing list software. Can be a change for
some people of course. But +1 from me for what it's worth.

Jim

On Thu, Apr 14, 2016 at 3:05 PM, Dinu Gherman 
wrote:

> Hi!
>
> I'm a silent lurker here, but I would most likely stop following this list
> after a move to google groups. I don't like imposing the use of an account
> already connecting to lots of data where any email address anywhere would
> be just what is needed. Sure, you can user other identities, but that's a
> pain.
>
> Unrelated to ideology, I've tried using google groups for other groups
> imposing it on their members and I did not like the "user experience" at
> all, and after a while I gave up. I don't understand how people can like
> it. If it would be more like google mail, I
> might like it.
>
> I would give Mailman 3 a try which should now have an interface decades
> better than Mailman 2, which is probably literally true if you skip the
> plural s.
>
> Cheers,
>
> Dinu
>
> --
> Sent on the move.
> Von unterwegs gesendet.
>
> > Am 14.04.2016 um 00:45 schrieb Pieter Hintjens :
> >
> > Hi all,
> >
> > I'd like to raise this question for discussion. The problem is that
> > zeromq-dev is running on an iMatix server, and I'd like to remove
> > iMatix infrastructure from our community, over time. It's a matter of
> > long term sustainability.
> >
> > (Apart from the list, iMatix also hosts downloads.zeromq.org which we
> > can start to move to Github release attachments IMO.)
> >
> > Anyhow, re. lists, let's weigh pros and cons of Google groups versus
> > other options. I'll make the decision based on this thread and other
> > factors.
> >
> > Thanks
> > Pieter
> > ___
> > zeromq-dev mailing list
> > zeromq-dev@lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Moving list to Google groups?

[Dinu Gherman]

Hi!

I'm a silent lurker here, but I would most likely stop following this list 
after a move to google groups. I don't like imposing the use of an account 
already connecting to lots of data where any email address anywhere would be 
just what is needed. Sure, you can user other identities, but that's a pain.

Unrelated to ideology, I've tried using google groups for other groups imposing 
it on their members and I did not like the "user experience" at all, and after 
a while I gave up. I don't understand how people can like it. If it would be 
more like google mail, I
might like it.

I would give Mailman 3 a try which should now have an interface decades better 
than Mailman 2, which is probably literally true if you skip the plural s.

Cheers,

Dinu

--
Sent on the move.
Von unterwegs gesendet.

> Am 14.04.2016 um 00:45 schrieb Pieter Hintjens :
> 
> Hi all,
> 
> I'd like to raise this question for discussion. The problem is that
> zeromq-dev is running on an iMatix server, and I'd like to remove
> iMatix infrastructure from our community, over time. It's a matter of
> long term sustainability.
> 
> (Apart from the list, iMatix also hosts downloads.zeromq.org which we
> can start to move to Github release attachments IMO.)
> 
> Anyhow, re. lists, let's weigh pros and cons of Google groups versus
> other options. I'll make the decision based on this thread and other
> factors.
> 
> Thanks
> Pieter
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev