On Wed Jul 18 07:34:54 2018, sogerc1@gmail.com wrote:
Show quoted text> All right.
> May I ask why not? Or rather may I ask if there's a reason why I
> shouldn't
> fork JSON::PP and add the functionality myself? AFAICT from a
> superficial
> look I would only need to add one line of code and I can't imagine how
> that
> would hinder the module. Except of course the performance penalty but
> that's pretty much a given when tie()-ing hashes.
Because the feature isn't compatible with JSON::XS (and its fork), and thus doesn't help JSON.pm users in general. I don't want to add a feature that is only available to JSON::PP, to keep things simple.
If you do want to fork JSON::PP to add your feature, well I think it's OK as long as you choose a name that wouldn't conflict with a future version of JSON::PP, but if you do, I'd rather recommend to use much cleaner JSON::Tiny as your base, or, as I said before, use JSON::YAJL as it should be faster and spare you the burden to maintain the fork.
Show quoted text>
> On Tue, Jul 17, 2018 at 5:13 PM, Kenichi Ishigaki via RT <
> bug-JSON-PP@rt.cpan.org> wrote:
>
> > <URL:
https://rt.cpan.org/Ticket/Display.html?id=125865 >
> >
> > On Tue Jul 17 20:48:56 2018, sogerc1@gmail.com wrote:
> > > I've seen a few questions about this floating on the internet and I
> > > also
> > > had this problem a few times, that I would need to modify some json
> > > stuff
> > > but with keeping the order of keys in hashes. Encoding already
> > > works with
> > > tied hashes but would you consider adding a configuration option to
> > > tie
> > > every newly created hash to Tie::IxHash during decoding?
> >
> > Sorry, no.
> >
> > You might want to try
https://metacpan.org/pod/JSON::YAJL::Parser
> > instead.
> >
> >
