On 11/30/19 7:39 AM, Bjorn Reese via Boost wrote:
On 2019-11-24 17:00, Robert Ramey via Boost wrote:
Hmmmm- I'm not seeing how this relates to my point.
I am agreeing the JSON serialization is not very hard, I was responding to your "no one has done it" comment.
Just to re-iterate: My points are:
a) I don't think it would be hard to make a JSON version of a serialization archive class using the XML archive as a model.
Agreed.
b) As far as I know, now one has done it.
Disagree. Please, please, please click on the following link:
https://github.com/breese/trial.protocol/tree/develop/include/trial/protocol...
I took a cursory look at this and it seems interesting. I had never seen it before. I should be easy to a) copy some files over to the serialization library - ojason_archive .. and ijason_archive b) tweak the test setup to include these "new" archive classes c) run the whole serialization librar test suite - which is quite a bit d) with the jam setup, one can restrict tests only to one specific archive (or even one specific test) . So running it locally on one's machine is quite practical. It if passes everything, has any required additional documentation (very little) and it's truely a drop in replacement for any other archive and it meets the "ramey coding standards" (not too anal), and willing to support complaints from users of the json archive (not really that bad), I'll be happy to add it the list of included archives along with xml, et. al. This will make you an official boost developer (if you aren't already) with out going through most of the boost agony. If you don't want to do it but someone else does, the same offer would apply to them. Robert Ramey
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost