Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:102290
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain gmail.com designates 209.85.215.53 as permitted sender)
MIME-Version: 1.0
References: <EB6971A9-2344-4AA2-BAE6-C4FDC7CF267A@lvht.net>
 <CAL=_i_mq0jQ_87X8NC21fn=Y3XdWXWfk3ROWG4P0g68mOOEX5A@mail.gmail.com>
 <CAFMT4Npu__xm1i_RXD0wSMju8Kc_C5NvXFW9pgvY8fQRo9g8LQ@mail.gmail.com>
 <CAL=_i_nmiB1Nc00ZXH=bODSL0K-Vf1ggE2HzbeufrFsGnGMMmg@mail.gmail.com> <CALW=WECeBaN+OP80La7SyEu+a9PHMBocOgHtyXVx+gQ9rYSBKw@mail.gmail.com>
In-Reply-To: <CALW=WECeBaN+OP80La7SyEu+a9PHMBocOgHtyXVx+gQ9rYSBKw@mail.gmail.com>
Date: Mon, 18 Jun 2018 09:38:44 +0900
Message-ID: <CAH19jWhZ1ZKXmJyf_D__EY46TVMTGXPPyCFGi3pq+N+QABhcOA@mail.gmail.com>
To: "S.A.N" <ua.san.alex@gmail.com>
Cc: internals <internals@lists.php.net>, i@lvht.net
Content-Type: multipart/alternative; boundary="00000000000075a5d4056edfcaa2"
Subject: Re: [PHP-DEV] [VOTE] Fiber API
From: shinji.igarashi@gmail.com (shinji igarashi)

--00000000000075a5d4056edfcaa2
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hello, I'm a fan of this RFC.

It seems that there is no time to finalize this RFC during the alpha stage.
Maybe the true voting will be postponed to the next version, if no more
works are made within a few days.

Haitao, do you still have the motivation to bring this into core?
If you have no time to brush up the RFC (in english...), how about
committing the RFC text into a repository?
There are some people supporting the concept itself.
Maybe they can help you to make the RFC more presentable by sending PR.
Or, if you don't have reason to continue the RFC anymore, then say so...
Someone can continue your work.

Even if there are people supporting the concept, the current choice of
implementation seems to be disliked.
Haitao, why do you stick to the way  switching VM stack only?
Yes, the way switching also machine (C) stack would increase memory
consumption.
But, to borrow your expression, Ruby's Fiber uses this method.
Ruby switches both VM stack and machine stack in its implementation of
Fiber.
If yielding from internal callback is important to pass the voting phase,
I think the machine stack switching should be considered more seriously.

The machine stack switching also has its own issues to be solved.
- Measuring stack consumptions in real world applications.
- Finding reasonable excuses for lacking portablity. (Is this acceptable to
core feature or bundled extension?)
- Still no sufficient explanations bringing it into core, even if above
problems are cleared.

I want to see the future that traditional ordinal PHP library is
asynchronized reasonably with this kind of feature, like this interesting
work.
https://github.com/joelwurtz/doctrine-async




Cheers,
Shinji IGARASHI


2018/06/13 =E5=8D=88=E5=89=8D8:14 "S.A.N" <ua.san.alex@gmail.com>:

Maybe someone can help in the finalization Fiber.
Haitao Lv, in pull request asked for help from Dmitry Stogov and
Nikita Popov, but no one answered him.
https://github.com/php/php-src/pull/3203

All popular web programming languages (JavaScript, Python, Ruby, Lua,
etc..) already implement userland async runtime.

Thanks.


--=20
PHP Internals - PHP Runtime Development Mailing List
To unsubscribe, visit: http://www.php.net/unsub.php

--00000000000075a5d4056edfcaa2--