Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:88521
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain gmail.com designates 209.85.213.195 as permitted sender)
MIME-Version: 1.0
Sender: jakub.php@gmail.com
In-Reply-To: <CAGa2bXYOpM8_6A8=x3Aoi8KBopyoab3=1fh8E56qeyEwjKi1Yw@mail.gmail.com>
References: <CAGa2bXbBhJtKo2tk+6udr2UHJFe46fbcw2-+yatru4pO1CuXxw@mail.gmail.com>
	<CAGa2bXZfRP_eV954mH+i08jvqtReJtip-oMP6ms8y8J_QNigJA@mail.gmail.com>
	<CAJW__o0J=JeVbTdK3xu6F8uDLn0Psw-_4c1_anz0O6=D+35KBQ@mail.gmail.com>
	<CAGa2bXY3wieKd90ypgYSZ3nZgLOtELcnt+cuuGorm4jDy7U_3A@mail.gmail.com>
	<023701d0f4a4$a0da4f60$e28eee20$@belski.net>
	<CAGa2bXYeMq5MuGiGUqnTZUgAK4xUZe6wkpqcijnU_FjjLiVwsw@mail.gmail.com>
	<CAGa2bXYOpM8_6A8=x3Aoi8KBopyoab3=1fh8E56qeyEwjKi1Yw@mail.gmail.com>
Date: Sun, 27 Sep 2015 18:22:54 +0100
Message-ID: <CAEKnhAG36KNerpn81gJ0hRQarinwyNeWH7vvH7ZbxsKyHx2y-Q@mail.gmail.com>
To: Yasuo Ohgaki <yohgaki@ohgaki.net>
Cc: Anatol Belski <anatol.php@belski.net>, Kalle Sommer Nielsen <kalle@php.net>, 
	Ferenc Kovacs <tyra3l@gmail.com>, "internals@lists.php.net" <internals@lists.php.net>
Content-Type: multipart/alternative; boundary=089e01538362c72da90520bdd4b0
Subject: Re: [PHP-DEV] Re: [RFC] [DISCUSSION] More precise float value
From: bukka@php.net (Jakub Zelenka)

--089e01538362c72da90520bdd4b0
Content-Type: text/plain; charset=UTF-8

Hi Yasuo,

On Fri, Sep 25, 2015 at 1:10 AM, Yasuo Ohgaki <yohgaki@ohgaki.net> wrote:

> Hi all,
>
> On Fri, Sep 25, 2015 at 7:49 AM, Yasuo Ohgaki <yohgaki@ohgaki.net> wrote:
> > No problem. I'll update so that 0 mode is for 7.1.
> > JSON's is better to use larger precision. So this change is targeted
> > to 5.6 and 7.0.
> >
> > Let me know if you have comments on this.
>
> The RFC is updated.
> https://wiki.php.net/rfc/precise_float_value
>
> Patches for 7.1/7.0/5.6 will be updated soon. 7.0/5.6's change will be
> one liner.
> i.e. Simply change to use PG(serialize_precision) instead of EG(precision).
>
>
I don't really like that change in a point release. It will be a huge mess
and it's also quite a big BC break for point release in a function that is
used a lot on production systems. Let me remind you that it might change
considerably generated data size for some cases. It would be probably a
good idea to check with RM's if they are ok with such BC break before
voting.

What worries me even more is changing anything like this in 5.6 where json
ext is not used on the most used Linux distros (Debian based and RHEL based
ones) where json-c is used by default. It means that you would have to ask
Remi to change it too. If it was changed, then it becomes quite difficult
to find out what precision is actually used in which version. I think that
it would be just a huge mess... :).

Also there is no option for voters who would like to use
serialize_precision for json in 7.1.  I would probably use one vote with
more options. Something like this:
JSON precision change
- no change
- use serialize_precision in 5.6
- use serialize_precicion in 7.0
- use serialize_precision in 7.1
- use json_precision in 7.1

The 5.6 and 7.0 should be there of course only if RM's are happy with this
BC break. I wouldn't personally allow that especially for 5.6. I think that
this change can wait till 7.1

In addition, the mode 0 voting is slightly unclear to me. It's one thing to
introduce mode 0 and another make it default. I'm actually for making it
default but it would be good to have that option in the voting. So how
about something like this?:
- do not introduce mode 0
- introduce mode 0 but keep current defaults
- introduce mode 0 and make it default

Cheers

Jakub

--089e01538362c72da90520bdd4b0--