Re: [boost] [serialization] proposed improvements: forward-compatibility of serialization
-------- Original Message --------
From: Sohail Somani <sohail@taggedtype.net>
This was a problem that interested me a few years ago so I made a solution that worked with any archive and I presented it at BoostCon (now CppNow). It was inefficient, hacky and required no modifications to existing archives but forward compatibility did work so I know it is completely possible.
Do you have a specific link handy ? or do you perhaps remember project name so I could google it?
My immediate question would be whether you already have released code for the products in question which complicates things a bit.
Fortunately we still have about 4-6 weeks before general release (-;
The proposal itself doesn't seem impossible, at a glance. Thanks - that's promising (-; I believe that this approach should be more efficient and more elegant that Google's Protocol Buffers.
And for us minimizing network traffic is of paramount importance. Andrew _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
On 2014-04-15, 2:59 PM, Andrzej Horoszczak wrote:
-------- Original Message --------
From: Sohail Somani <sohail@taggedtype.net>
This was a problem that interested me a few years ago so I made a solution that worked with any archive and I presented it at BoostCon (now CppNow). It was inefficient, hacky and required no modifications to existing archives but forward compatibility did work so I know it is completely possible.
Do you have a specific link handy ? or do you perhaps remember project name so I could google it?
No I don't but I could probably dig up the presentation if you email me off-list and remind me.
My immediate question would be whether you already have released code for the products in question which complicates things a bit.
Fortunately we still have about 4-6 weeks before general release (-;
The proposal itself doesn't seem impossible, at a glance. Thanks - that's promising (-; I believe that this approach should be more efficient and more elegant that Google's Protocol Buffers.
I don't know enough about protocol buffers to say, TBH.
And for us minimizing network traffic is of paramount importance.
The question is which feature of paramount importance is most important, especially with 4-6 weeks to general release!
Andrew
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
participants (2)
-
Andrzej Horoszczak -
Sohail Somani