Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:65512
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: unknown (pb1.pair.com: domain zend.com does not designate 209.85.214.171 as permitted sender)
References: <5d21b42656d49b4a71d9f808541bd745@mail.gmail.com>
 <1359562177.37091.YahooMailNeo@web133001.mail.ir2.yahoo.com> <1359562237.22084.YahooMailNeo@web133001.mail.ir2.yahoo.com>
In-Reply-To: <1359562237.22084.YahooMailNeo@web133001.mail.ir2.yahoo.com>
MIME-Version: 1.0
Thread-Index: AQIqL8waXVttbugrqkVxisGQEOzMkQHawY2bAc7yT9aXjF8vAA==
Date: Wed, 30 Jan 2013 18:47:13 +0200
Message-ID: <a38a9590ac5c68c454b8340538a03690@mail.gmail.com>
To: hakre <hanskrentel@yahoo.de>
Cc: internals@lists.php.net
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Subject: RE: [PHP-DEV] [RFC] Integrating Zend Optimizer+ into the PHP distribution
From: zeev@zend.com (Zeev Suraski)

> > * In that RFC you write:
> >
> > "the integration won=E2=80=99t happen before late 2014." [if it's not b=
undled
> > with PHP 5.5]
> >
> > Can you please outline why?

Based on an 18 month release cycle, and assuming we release 5.5.0 in mid
2013, 5.6.0 will come out late 2014.

> > Especially does it mean you stop
> > contributing to the PECL development if you don't get this bundled with
> > PHP
> 5.5?

No.  If you take a closer look at the options, the 'No' option reads 'Don=
=E2=80=99t
integrate Optimizer+ to PHP, provide it as an optional component in PECL
only'.  We're going to publish the code as soon as we can, I hope no later
than next week, and it'll be before we have the results of the vote.

By 'integration' I refer to going beyond just including it in PECL, but
including it in core.

> > Also can you please outline why you put obviously so much focus in
> > bundling this to PHP 5.5? Or is my impression wrong?

Optimizer+ has been a free (closed source) component since 2008.  We've bee=
n
talking about open sourcing it numerous times over the years but it was
never prioritized high enough.  With the discussion last week about
integrating an opcode cache into PHP's core, the challenges of using APC fo=
r
that purpose on a short timeline, and the fact Optimizer+ is a significantl=
y
faster implementation than APC - I thought that this could be a good
opportunity to commit ourselves (Zend) into doing this.  Otherwise it would
have probably never happened.

To me, waiting for a couple of months to get a huge performance gain
out-of-the-box is a no brainer.  In fact, it might be a way to convince a
lot of people that migrating is worth it.

> > * With full respect and the best intentions: Are you able and if yes,
> > can you share about the motivation why you decided (quite
> > surprisingly) to contribute at this place in time?

See above answer.

> >You also wrote in an earlier
> > email that you got out of sync with your userbase.

I did not.  Perhaps you read it that way :)
Pierre said something along the lines of 'some people here being
disconnected from our userbase'.  I agreed with him, but obviously, I wasn'=
t
talking about myself.

> > Under these
> > circumstances, the impression could be that it took a little bit too
> > long until this decision was done and I would like to see this
> > impression
> > clarified because there are many loose ends.

Please bring up any loose ends you're spotting and I'll try to address them
as best I can.

> > * Is this surprising and welcomed release related in any way to the
> > Openstack Initiative?

Not at all.

> > * Which benefits does Zend Inc. see in contributing the Opcode cache?

Simply put, this could benefit PHP greatly without negatively affecting our
business in any way.

> > * Last but not least, not related to the opcode cache alone, but
> > related because you want to bundle it with core: If some day the PHP
> > group decides to choose a similar software license, but different in
> > the sense that it is more compatible with existing FLOSS licensing,
> > would you have a problem to re-license as well, e.g. under MIT or Apach=
e
> > 2.0
> for that part?

The plan is to contribute the source code to the PHP project.  It'll be
under the same license as PHP and subject to any changes in the PHP
licensing scheme that we'll agree on.

Zeev