Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:111082
To: =?iso-8859-2?Q?Micha=B3_Marcin_Brzuchalski?=
	<michal.brzuchalski@gmail.com>, PHP Internals List <internals@lists.php.net>
Thread-Topic: [PHP-DEV] Re: [RFC][DISCUSSION] debug_backtrace alternative as
 an array of StackFrame objects
Thread-Index: AQHWXtFgqJP43t0OTE2TPa02/q7m7KkROYlE
Date: Tue, 21 Jul 2020 02:27:50 +0000
Message-ID:
 <DM6PR07MB6618D57BE5B6E246EA112CB8F9780@DM6PR07MB6618.namprd07.prod.outlook.com>
References:
 <CABdc3WpEmoKTYEsDk+K1Jb=kh7nu8gJ3TkhY5639ueJaWjN9AQ@mail.gmail.com>,<CABdc3WrNcPdY03d5oyUSW5nwaTkqd_wGscf2kWwxhD2ykPzRYA@mail.gmail.com>
In-Reply-To:
 <CABdc3WrNcPdY03d5oyUSW5nwaTkqd_wGscf2kWwxhD2ykPzRYA@mail.gmail.com>
Accept-Language: en-CA, en-US
Content-Language: en-CA
x-ms-exchange-messagesentrepresentingtype: 1
x-tmn: [HTX2bC6gvWITokyWCu/HcAh0kcbZxp395xko0f3G0NqsEdBT+CYBcjfeL95Ob0HM]
x-ms-publictraffictype: Email
x-incomingheadercount: 44
x-eopattributedmessage: 0
x-ms-office365-filtering-correlation-id: 63a21b4d-76ad-4583-2a78-08d82d1daa70
x-ms-traffictypediagnostic: BN7NAM10HT003:
x-microsoft-antispam: BCL:0;
x-microsoft-antispam-message-info:
 hvksma4yG6fxqkecn6nOFl15NwYElHq31Y6qKtk8hGfCxCds90sGD44nma3+A5nS6y3NjM56cVnfheuLvuQPr9BCi4qHfU6uL6V5DoXrAZHz4matXCX2heyMAMqlcs9Q6o/DduFm/8AUK+heOxSGXS2vjfdFeJs7x8WMbMWUuCcLkH8zf4I6obZHy2ERYadLqcsTdJYemPHMvRxPTB+1/Ih0Qx0jlouvK5KB5oeu2Y3r3d9kwk8u3qqVeqABshET
x-forefront-antispam-report:
 CIP:255.255.255.255;CTRY:;LANG:en;SCL:0;SRV:;IPV:NLI;SFV:NSPM;H:DM6PR07MB6618.namprd07.prod.outlook.com;PTR:;CAT:NONE;SFTY:;SFS:;DIR:OUT;SFP:1901;
x-ms-exchange-antispam-messagedata:
 6fFh20CmdkkdM+JmXkR7p3T57zBH7uiJ5FOePP5nEy18DzXEUYE2BYWo9NSpsxxNFMVuzYzt5XbkN6S1WC2KUkB15fYAKgvP2Sw5+Xo5WZLznhfqxjr5cvn13tjlrRTn/4fTHicsZa41goEN4hMlmDx9LaZLmrqZ0dII0pCQlbTon6OoJwmUoiPINuK2q7VJ/c5oMF5sazOFkKOhs2ha6w==
x-ms-exchange-transport-forked: True
Content-Type: text/plain; charset="iso-8859-2"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-OriginatorOrg: hotmail.com
X-MS-Exchange-CrossTenant-AuthAs: Anonymous
X-MS-Exchange-CrossTenant-AuthSource: BN7NAM10FT063.eop-nam10.prod.protection.outlook.com
X-MS-Exchange-CrossTenant-RMS-PersistedConsumerOrg: 00000000-0000-0000-0000-000000000000
X-MS-Exchange-CrossTenant-Network-Message-Id: 63a21b4d-76ad-4583-2a78-08d82d1daa70
X-MS-Exchange-CrossTenant-rms-persistedconsumerorg: 00000000-0000-0000-0000-000000000000
X-MS-Exchange-CrossTenant-originalarrivaltime: 21 Jul 2020 02:27:50.8724
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Internet
X-MS-Exchange-CrossTenant-id: 84df9e7f-e9f6-40af-b435-aaaaaaaaaaaa
X-MS-Exchange-Transport-CrossTenantHeadersStamped: BN7NAM10HT003
Subject: Re: [PHP-DEV] Re: [RFC][DISCUSSION] debug_backtrace alternative as an
 array of StackFrame objects
From: tysonandre775@hotmail.com (tyson andre)

Hi Micha=B3 Marcin Brzuchalski,=0A=
=0A=
>=A0Heads up, I plan to open the vote tomorrow morning.=0A=
=0A=
=0A=
Sorry for the late comment - I'd been busy/occupied with various other thin=
gs and this RFC fell off my radar,=0A=
and I'd assumed that anything I'd end up noticing when looking at the RFC =
=0A=
would have already been pointed out by other reviewers.=0A=
(In general, I've only reviewed or locally compiled the code of a small fra=
ction of recent RFCs mentioned on the mailing list.)=0A=
=0A=
I had some comments on the implementation in https://github.com/php/php-src=
/pull/5820#pullrequestreview-452045885=0A=
=0A=
- How much does this change elapsed time in typical use cases? This had com=
pletely skipped my mind when I saw this RFC, but I'd generally assume that =
stack traces aren't kept along for longer than needed to process them, and =
that most stack traces would be at most hundreds of frames in typical appli=
cations. So I'd consider CPU time elapsed more important than RAM.=0A=
  I'd had similar issues with the tradeoffs in PhpToken (and forgot). Those=
 issues were that the applications using the array would use less memory an=
d be faster if I used the returned array multiple times, but less so if I o=
nly processed the array once and immediately freed it.=0A=
  PHP also seems faster at fetching dimension from arrays than properties f=
rom objects.=0A=
=0A=
  Still, though, this might help with a bit with developer productivity (e.=
g. IDEs being better at inferring types of $frame->getFile() in some cases,=
 or providing help text for the method), or in being able to check types wh=
en passing individual frames=0A=
- I found a bug: The ArrayAccess magic methods are implemented, but not the=
 real methods.=0A=
- I found some memory leaks, possibly related to tracking argument arrays i=
n the object being stored=0A=
=0A=
- Casting getFile() to the empty string is unexpected for me ($frame['file'=
] and $frame->file are null in the implementation).=0A=
- Forbidding modification/deletion on the ArrayAccess magic API (e.g. `$fra=
me['trace'] =3D ...`)=0A=
  seems inconsistent when modification and deletion are allowed through the=
 real properties,=0A=
  (and throwing there may affect code written to process the array returned=
 by Throwable->getTrace())=0A=
- The typed property default was incompatible with the typed property decla=
ration (`class StackFrame { string $file =3D null; /*...*/}` for internal s=
tack frames)=0A=
=0A=
With the feature freeze nearby, more of a heads up would have been useful.=
=0A=
Aside: The guidelines in https://wiki.php.net/rfc/howto have been followed,=
=0A=
but I think it'd be useful for those guidelines to say "consider *at least*=
 one days heads up mail on the mailing list" instead of "consider one days"=
 to avoid cases where review comments are made on short notice.=0A=
Normally, any issues found the day before voting starts would have more tim=
e to reschedule the vote to be addressed.=0A=
=0A=
- I don't know how amendments to the howto get proposed or decided on=0A=
=0A=
Regards,=0A=
- Tyson=