[Caml-list] ZeroMQ vs Nanomsg

[Caml-list] ZeroMQ vs Nanomsg

[Dario Teixeira]

Hi,

I'm evaluating ZeroMQ and Nanomsg for the middleware of a distributed
application.  Now, I know both projects have had a fair share of drama
in the past, to the point that it's hard for an outside observer to be
up-to-date on the *current* strengths and weaknesses of each one.  
Still,
it's my understanding that though ZeroMQ is more established, Nanomsg 
was
developed later by one of the original authors of ZeroMQ, and supposedly
fixes some of the architectural mistakes of ZeroMQ.  To complicate 
matters,
Nanomsg's author quit the project a while ago, leaving it in limbo for
a while.  In the meantime, it seems development has picked up again, and
Nanomsg recently had its 1.0 release.

Anyway, I'm sure the picture I painted above is incomplete.  Therefore,
I would be much obliged to hear your opinions and experiences with 
either
project.  I will be using the OCaml bindings, obviously, so the maturity 
of
the bindings is also a factor.  (On first glance they seem quite 
complete
and both offer support for LWT, which is a must.)

Thanks in advance for your time!
Kind regards,
Dario Teixeira

Re: [Caml-list] ZeroMQ vs Nanomsg

[Simon Cruanes]

[-- Attachment #1: Type: text/plain, Size: 1566 bytes --]

Hi,

I didn't know development of nanomsg had resumed, but it's good news. I
use rgrinberg's onanomsg binding, which works fine for me (although it's
only for a small program), no memleaks, and the API is nice. I think the
bindings are quite mature and rgrinberg is very nice to interact with
;-).

Le Fri, 15 Jul 2016, Dario Teixeira a écrit :
> I'm evaluating ZeroMQ and Nanomsg for the middleware of a distributed
> application.  Now, I know both projects have had a fair share of drama
> in the past, to the point that it's hard for an outside observer to be
> up-to-date on the *current* strengths and weaknesses of each one.  Still,
> it's my understanding that though ZeroMQ is more established, Nanomsg was
> developed later by one of the original authors of ZeroMQ, and supposedly
> fixes some of the architectural mistakes of ZeroMQ.  To complicate matters,
> Nanomsg's author quit the project a while ago, leaving it in limbo for
> a while.  In the meantime, it seems development has picked up again, and
> Nanomsg recently had its 1.0 release.
> 
> Anyway, I'm sure the picture I painted above is incomplete.  Therefore,
> I would be much obliged to hear your opinions and experiences with either
> project.  I will be using the OCaml bindings, obviously, so the maturity of
> the bindings is also a factor.  (On first glance they seem quite complete
> and both offer support for LWT, which is a must.)

-- 
Simon Cruanes

http://weusepgp.info/
key 49AA62B6, fingerprint 949F EB87 8F06 59C6 D7D3  7D8D 4AC0 1D08 49AA 62B6

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [Caml-list] ZeroMQ vs Nanomsg

[SP]

On Fri, Jul 15, 2016 at 11:59:12AM +0100, Dario Teixeira wrote:
>I would be much obliged to hear your opinions and experiences with
>either project.  I will be using the OCaml bindings, obviously, so the 
>maturity of the bindings is also a factor.  (On first glance they seem
>quite complete and both offer support for LWT, which is a must.)

What I'm curious about is how interoperable they are. Can one switch
between the two back-ends or bindings? Does either of them "wrap" the
other?

-- 
    SP

Re: [Caml-list] ZeroMQ vs Nanomsg

[Dario Teixeira]

Hi,

> I didn't know development of nanomsg had resumed, but it's good news. I
> use rgrinberg's onanomsg binding, which works fine for me (although 
> it's
> only for a small program), no memleaks, and the API is nice. I think 
> the
> bindings are quite mature and rgrinberg is very nice to interact with
> ;-).

Thanks for the feedback!  That's good to hear, as I'm currently leaning
more towards Nanomsg.  Do you also have any positive/negative experience
with ZeroMQ?

Cheers,
Dario

Re: [Caml-list] ZeroMQ vs Nanomsg

[Dario Teixeira]

Hi,

> What I'm curious about is how interoperable they are. Can one switch
> between the two back-ends or bindings? Does either of them "wrap" the
> other?

Someone please correct me if I'm wrong, but my impression is that
there's no interoperability at all between ZeroMQ and Nanomsg.

Cheers,
Dario

Re: [Caml-list] ZeroMQ vs Nanomsg

[Hezekiah M. Carty]

[-- Attachment #1: Type: text/plain, Size: 2133 bytes --]

Hello,

I've been using the zeromq bindings for the last six years or so.  zeromq
has been very nice in my experience - excellent documentation and a helpful
community.  The current and past maintainers of the zeromq bindings have
been very easy to work with.  I wrote the Lwt interface for zeromq -
feedback on that piece is welcome.

I started looking at nanomsg shortly before its original author/maintainer
quit the project.  I stopped seriously considering nanomsg for my projects
at that point.  I'm happy to hear that maintenance has been picked back up
though.  And to echo Simon Cruanes's comment, rgrinberg is also very easy
to work with!

Hez

On Fri, Jul 15, 2016 at 6:59 AM Dario Teixeira <dario.teixeira@nleyten.com>
wrote:

> Hi,
>
> I'm evaluating ZeroMQ and Nanomsg for the middleware of a distributed
> application.  Now, I know both projects have had a fair share of drama
> in the past, to the point that it's hard for an outside observer to be
> up-to-date on the *current* strengths and weaknesses of each one.
> Still,
> it's my understanding that though ZeroMQ is more established, Nanomsg
> was
> developed later by one of the original authors of ZeroMQ, and supposedly
> fixes some of the architectural mistakes of ZeroMQ.  To complicate
> matters,
> Nanomsg's author quit the project a while ago, leaving it in limbo for
> a while.  In the meantime, it seems development has picked up again, and
> Nanomsg recently had its 1.0 release.
>
> Anyway, I'm sure the picture I painted above is incomplete.  Therefore,
> I would be much obliged to hear your opinions and experiences with
> either
> project.  I will be using the OCaml bindings, obviously, so the maturity
> of
> the bindings is also a factor.  (On first glance they seem quite
> complete
> and both offer support for LWT, which is a must.)
>
> Thanks in advance for your time!
> Kind regards,
> Dario Teixeira
>
>
> --
> Caml-list mailing list.  Subscription management and archives:
> https://sympa.inria.fr/sympa/arc/caml-list
> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> Bug reports: http://caml.inria.fr/bin/caml-bugs
>

[-- Attachment #2: Type: text/html, Size: 2861 bytes --]

Re: [Caml-list] ZeroMQ vs Nanomsg

[Dario Teixeira]

Hi,

> I've been using the zeromq bindings for the last six years or so.
> zeromq has been very nice in my experience - excellent documentation
> and a helpful community.  The current and past maintainers of the
> zeromq bindings have been very easy to work with.  I wrote the Lwt
> interface for zeromq - feedback on that piece is welcome.

Thanks for the feedback.  Yes, ZeroMQ has excellent documentation,
which in fact I've used to also learn Nanomsg (there's plenty of
overlap in concepts and patterns).  Nanomsg's docs are unfortunately
a bit lacking, particularly where edge cases and gotchas are concerned.


> I started looking at nanomsg shortly before its original
> author/maintainer quit the project.  I stopped seriously considering
> nanomsg for my projects at that point.  I'm happy to hear that
> maintenance has been picked back up though.  And to echo Simon
> Cruanes's comment, rgrinberg is also very easy to work with!

Did you look at Nanomsg purely out of curiosity, or did it offer
something that you found lacking in ZeroMQ?

Cheers,
Dario

Re: [Caml-list] ZeroMQ vs Nanomsg

[Daniel Bünzli]

Le vendredi, 15 juillet 2016 à 16:58, Dario Teixeira a écrit :
> Did you look at Nanomsg purely out of curiosity, or did it offer
> something that you found lacking in ZeroMQ?


This question is not for me but I'll respond anyways… What I personally found interesting in nanomsg was the fact that the original author wanted to standardize the protocols in rfcs [1]. This would have allowed for pure OCaml implementations without having to reverse engineer the C one and would have provided interesting, standardized, alternatives to sockets APIs to do networking (for example in MirageOS contexts).

Unfortunately the new maintainer said he was not interested in doing this [2] (which is a pity since he's also the author of a pure go implementation of the protocols).

Best,  

Daniel

[1] https://github.com/nanomsg/nanomsg/tree/master/rfc
[2] http://www.freelists.org/post/nanomsg/Coming-back-as-BDFL,7

Re: [Caml-list] ZeroMQ vs Nanomsg

[Hezekiah M. Carty]

[-- Attachment #1: Type: text/plain, Size: 874 bytes --]

On Fri, Jul 15, 2016 at 11:58 AM Dario Teixeira <dario.teixeira@nleyten.com>
wrote:

> > I started looking at nanomsg shortly before its original
> > author/maintainer quit the project.  I stopped seriously considering
> > nanomsg for my projects at that point.  I'm happy to hear that
> > maintenance has been picked back up though.  And to echo Simon
> > Cruanes's comment, rgrinberg is also very easy to work with!
>
> Did you look at Nanomsg purely out of curiosity, or did it offer
> something that you found lacking in ZeroMQ?
>
>
Mostly curiosity.  I was pretty happy with zeromq but wanted to see what
nanomsg had to offer.  There were some intriguing socket/pattern/protocols
in nanomsg (BUS and SURVEY for example) but the better documentation in
zeromq won out because of all the clear examples of the benefits and
pitfalls of zeromq's various socket types.

Hez

[-- Attachment #2: Type: text/html, Size: 1224 bytes --]