Boost logo

Boost :

Subject: Re: [boost] [RPC] Compile-time reflection for RPC and mocking
From: Peter Bindels (dascandy_at_[hidden])
Date: 2014-12-09 09:40:10


On 9 December 2014 at 12:10, Bjorn Reese <breese_at_[hidden]> wrote:

> On 12/08/2014 03:27 PM, Peter Bindels wrote:
>
> - I can work on this idea to work it out into a full RPC mechanism, to be
>> proposed for Boost.RPC, if there is enough interest. Currently it uses
>>
>
> We are seeing a general tendency away from RPC towards either ReST for
> client-server architectures or messaging middlewares for peer-to-peer
> architectures.
>

Who's "We" ? Given the powerful abilities that you can get with async and
future/promise, coupled with easy remoting with some rpc in between that
gives the same interface, would lead me to believe RPC is due for a
comeback. At least, for computing distribution (think threading++), not for
using services.

> If you are interested in working on alternative proposals, then one
> possibility could be to add a client-side for the Boost.Http project:
>
> https://github.com/BoostGSoC14/boost.http
>

Sounds like an interesting proposal, but quite different. Also, it looks
like a client library for that would be more like a separate HTTP client
project, not as a companion project. How's the boost acceptance status of
this project?

We already have Boost.Serialization and Boost.Fusion that are capable of
> serialization.

True, and that would probably be my next guess for integrating it into
boost - switch out some simplified local hacks for battle-tested Boost
code. For the current level of "it's a bit of a hack" that doesn't
necessarily make sense.


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk