Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:106780
Content-Type: text/plain;
	charset=utf-8
Mime-Version: 1.0 (1.0)
In-Reply-To: <f99cd630-7321-fad5-f584-e108efa4ca0d@gmail.com>
Date: Wed, 28 Aug 2019 19:23:56 -0400
Cc: PHP internals <internals@lists.php.net>
Content-Transfer-Encoding: quoted-printable
Message-ID: <98CDCF47-D262-449B-92BD-1E041A467B21@gmail.com>
References: <CAF+90c9Kc6m_bv=aJxbD3-RmnNXXLqv6mPjHp7U7qxX_ck27qw@mail.gmail.com> <CAKuGYLQC2z+GmOgbqxFWbAGS6B2z=mXTp72c6Be53ju4D6pNRw@mail.gmail.com> <CAJHtznxgkUFZ2os5Hx5rYmFCrATJ+LTOovK6W4J1cDw6oo45=w@mail.gmail.com> <CAKuGYLQfg8CQOORWTO57Cfk92oM3DNmw0h9wt3jP9gSUCx+QWg@mail.gmail.com> <CAJHtznw1POVOtK656fWn+Vdj94RnHWmZ9A5rEHoWjiXXod=Piw@mail.gmail.com> <CAKuGYLR6j173SKUxR0gXnCARZ-Lg4mmZyOLJYD8hRfsKnp9CfQ@mail.gmail.com> <6ed74ae0-5feb-39de-a39a-f9a8f80401b2@gmail.com> <342BB5DE-80D0-405E-8C71-D959F38D17E8@gmail.com> <b024d0fe-ac0f-d9b9-a389-b25804ac40db@gmail.com> <CAKuGYLT17d0pL9taO6M0xUPwu3dPU6nLRNGWFc89k0RtU6BM6Q@mail.gmail.com> <f99cd630-7321-fad5-f584-e108efa4ca0d@gmail.com>
To: Stanislav Malyshev <smalyshev@gmail.com>
Subject: Re: [PHP-DEV] [RFC] Reclassifying engine warnings
From: matthewmatthew@gmail.com (Matthew Brown)

$foo++ becoming 1 when $foo is undefined is not intuitive to me.

To take a very trivial example, that behaviour causes =E2=80=9Cfor ($i =3D 0=
; $i < 10; $I++) {}=E2=80=9D to loop indefinitely.

> On Aug 28, 2019, at 6:52 PM, Stanislav Malyshev <smalyshev@gmail.com> wrot=
e:
>=20
> Hi!
>=20
>> This is where I think PHP may have broken us a little.
>=20
> I think it's in no way "broken" to be able to easily match expectations,
> like $foo++ always do what you meant without clunky boilerplate.
> Also, if you think PHP is the only language I program in daily (and I
> mean every day, except some weekends and vacations maybe) you're wrong
> :) I have something to compare to, so what I say some things are easier
> in PHP that's because I actually compared.
>=20
>> I just asked a few non-PHP developers here what they expect "(function
>> () { $a++; })()" to do, and they agreed it would be some sort of error.
>> Got the same answer for "(function () { $a->bar =3D 5; })() ".
>=20
> I see absolutely no reason for it. Maybe if you're a Java programmer who
> never saw non-statically-typed non-B&D language - but that's not a
> virtue we should be striving to emulate. If we have a tool that already
> does things better and some people don't even know such tools are
> possible, we should educate them, not break our tools so it would be
> more comfortable fit to their experience.
>=20
>> Indeed, anyone who's used another C-like language (JS, TS, Java, C# etc)
>> is used to these things being errors, so it can be disorientating to see
>=20
> I don't think having things just work instead of usual boilerplate that
> you have to declare everything upfront and repeat even obvious things
> numerous times is "disorienting" in any way. If you check how languages
> that are alive evolve (like Java or C++, C is mostly fossilized these
> days), even strict ones, you see they support more and more intuitive
> approaches - like auto variables for example - which make things easier.
> Because human should spend time thinking, not figuring out how to
> satisfy a dumb compiler. That's the direction we should be moving to.
> Not adding more errors and boilerplate in clear cases like $foo++.
>=20
> --=20
> Stas Malyshev
> smalyshev@gmail.com