# Backporting to 5_3 - Thread: https://externals.io/message/36536 - Original archive: https://news-web.php.net/php.internals/36536 - Started by Felipe Pena on 2008-03-26T13:04:46+00:00 - Messages: 31 - Page 1 of 2: messages 1 to 25 (25 per page, add `?page=N` to the URL) - Next page: https://externals.io/message/36536.md?page=2 - Reply by email to internals@lists.php.net - More for agents: search with https://externals.io/search.md?q={query} · JSON API at https://externals.io/api · documentation at https://externals.io/llms.txt --- ## #36536 — Felipe Pena — 2008-03-26T13:04:46+00:00 URL: https://externals.io/message/36536#36536 Hello, I would like know which itens below needs be backported to 5_3: - Added runtime JIT auto-globals fetching and caching. (Dmitry, Sara) - Added "jump label" operator (limited "goto"). (Dmitry, Sara) - Removed support for "continue" and "break" operators with non-constant operands. (Dmitry) - Removed undocumented and incomplete support for strings in list() operator. (Dmitry) - Changed "instanceof" and "catch" operators, `is_a()` and `is_subclass_of()` functions to not call __autoload(). (Dmitry) Let me know, and then i could post the patch for aprovation. PS: If is whished be backported a item in NEWS file that i not mentioned, let know too. :) -- Regards, Felipe Pena. --- ## #36538 — Jani Taskinen — 2008-03-26T13:41:56+00:00 URL: https://externals.io/message/36536#36538 · In reply to: #36536 All. --Jani > Hello, > > I would like know which itens below needs be backported to 5_3: > > - Added runtime JIT auto-globals fetching and caching. (Dmitry, Sara) > - Added "jump label" operator (limited "goto"). (Dmitry, Sara) > - Removed support for "continue" and "break" operators with non-constant > operands. (Dmitry) > - Removed undocumented and incomplete support for strings in list() > operator. (Dmitry) > - Changed "instanceof" and "catch" operators, `is_a()` and > `is_subclass_of()` functions to not call __autoload(). (Dmitry) > > Let me know, and then i could post the patch for aprovation. > > PS: If is whished be backported a item in NEWS file that i not > mentioned, let know too. :) > > -- > Regards, > Felipe Pena. > > -- Patches/Donations: http://pecl.php.net/~jani/ --- ## #36540 — Stanislav Malyshev — 2008-03-26T16:39:28+00:00 URL: https://externals.io/message/36536#36540 · In reply to: #36536 Hi! > - Added runtime JIT auto-globals fetching and caching. (Dmitry, Sara) Does this change any semantics, etc? Any reason why it wasn't merged in the first place? > - Added "jump label" operator (limited "goto"). (Dmitry, Sara) No strong opinion here, I didn't need it anyway :) > - Removed support for "continue" and "break" operators with non-constant > operands. (Dmitry) I'd wait for 6 with this. May break some scripts. > - Removed undocumented and incomplete support for strings in list() > operator. (Dmitry) Not sure here. What code does it break? > - Changed "instanceof" and "catch" operators, `is_a()` and > `is_subclass_of()` functions to not call __autoload(). (Dmitry) This definitely should be - there's absolutely no reason any kind on instanceof-like operator would call autoload. -- Stanislav Malyshev, Zend Software Architect stas@zend.com http://www.zend.com/ (408)253-8829 MSN: stas@zend.com --- ## #36548 — Cristian Rodriguez — 2008-03-26T16:55:36+00:00 URL: https://externals.io/message/36536#36548 · In reply to: #36540 2008/3/26, Stanislav Malyshev : > > […] > > > I'd wait for 6 with this. May break some scripts. I dont think it will break more a piece of code that doesnt really work. ;) -- "If debugging is the process of removing bugs, then programming must be the process of putting them in." – Edsger Dijkstra http://www.cristianrodriguez.net --- ## #36550 — Pierre Joye — 2008-03-26T16:56:48+00:00 URL: https://externals.io/message/36536#36550 · In reply to: #36540 > Hi! > > > > […] > > Does this change any semantics, etc? Any reason why it wasn't merged in > the first place? It was initially done for the input encoding in head as the runtime encoding may change after RINIT and inside the script. By the way, the JIT fetch&de/encode part is still to be done in HEAD. -- Pierre http://blog.thepimp.net | http://www.libgd.org --- ## #36552 — Stanislav Malyshev — 2008-03-26T17:13:17+00:00 URL: https://externals.io/message/36536#36552 · In reply to: #36548 > > […] > > I dont think it will break more a piece of code that doesnt really work. ;) Af far as I know, breaks with arguments do work in PHP. -- Stanislav Malyshev, Zend Software Architect stas@zend.com http://www.zend.com/ (408)253-8829 MSN: stas@zend.com --- ## #36554 — Hannes Magnusson — 2008-03-26T18:02:42+00:00 URL: https://externals.io/message/36536#36554 · In reply to: #36552 > > […] > > Af far as I know, breaks with arguments do work in PHP. $foo = 2; > while(true) { > while(true) { > echo "Hello world!\n"; > if (true) { > break $foo; > } > } > echo "Never reached\n"; > } > > This is what was removed. not normal "break 2;" According to Google Code Search ( http://www.google.com/codesearch?q=lang%3Aphp+break\s%2B\%24&hl=en&btnG=Search+Code ) there are only two distinct pieces of code out there using dynamic break: "never heard of" XPath.class.php and "also never heard of" some code deep inside this Xayara CMS. Although it seems XPath.class.php is used by some projects (phpmyadmin, phpgroupware, falbum, lifetype) I checked phpmyadmin and at least in 2.9.1 this class is not included anymore. On the other hand, dynamic continie only seems to be used in PhpDocumentor: http://www.google.com/codesearch?hl=en&lr=&q=lang%3Aphp+continue\s%2B\%24&btnG=Search Havent't checked if that one is still used. - Markus --- ## #36578 — Stanislav Malyshev — 2008-03-26T21:05:14+00:00 URL: https://externals.io/message/36536#36578 · In reply to: #36575 Hi! > > […] As I understand, this code works now. > "never heard of" XPath.class.php and "also never heard of" some code > deep inside this Xayara CMS. And used by people. Did I or anybody here on the list actually contacted with these people or not does not matter - we can't contact millions of PHP users anyway. So I think the removal might be held for PHP 6. -- Stanislav Malyshev, Zend Software Architect stas@zend.com http://www.zend.com/ (408)253-8829 MSN: stas@zend.com --- ## #36580 — Lukas Kahwe Smith — 2008-03-26T21:16:38+00:00 URL: https://externals.io/message/36536#36580 · In reply to: #36578 > > > […] > > And used by people. Did I or anybody here on the list actually > contacted with these people or not does not matter - we can't > contact millions of PHP users anyway. > So I think the removal might be held for PHP 6. > yeah .. lets keep this kind of cleanup for php6 regards, Lukas --- ## #36587 — Felipe Pena — 2008-03-26T23:30:00+00:00 URL: https://externals.io/message/36536#36587 · In reply to: #36540 2008/3/26, Stanislav Malyshev : > > […] > > > This definitely should be - there's absolutely no reason any kind on > instanceof-like operator would call autoload. Hehe, curious... See above. + * (PHP_5_1) + NEWS + ZendEngine2/zend_builtin_functions.c + ZendEngine2/zend_compile.c + ZendEngine2/zend_compile.h + ZendEngine2/zend_execute.h + ZendEngine2/zend_execute_API.c + ZendEngine2/tests/catch.phpt + ZendEngine2/tests/instanceof.phpt + ZendEngine2/tests/is_a.phpt: + Changed "instanceof" and "catch" operators, `is_a()` and `is_subclass_of()` + functions to not call __autoload(). This was done a long time! :D (and is documented about the instanceof) --- ## #36589 — johannes@php.net — 2008-03-26T23:34:32+00:00 URL: https://externals.io/message/36536#36589 · In reply to: #36536 > Hello, > > I would like know which itens below needs be backported to 5_3: > > - Added runtime JIT auto-globals fetching and caching. (Dmitry, Sara) Not sure if this has benefits without the encoding stuff which is planned for PHP 6. > - Added "jump label" operator (limited "goto"). (Dmitry, Sara) No objection, while it adds yet another keyword. > - Removed support for "continue" and "break" operators with non-constant > operands. (Dmitry) -1, also just two use cases were found I don't like to break that with 5.3. > - Removed undocumented and incomplete support for strings in list() > operator. (Dmitry) What exactly does that change? > - Changed "instanceof" and "catch" operators, `is_a()` and > `is_subclass_of()` functions to not call __autoload(). (Dmitry) +1, makes sense. > Let me know, and then i could post the patch for aprovation. Please do that. Thanks, johannes --- ## #36594 — Pierre Joye — 2008-03-26T23:48:24+00:00 URL: https://externals.io/message/36536#36594 · In reply to: #36589 > > > > > […] > > Not sure if this has benefits without the encoding stuff which is > planned for PHP 6. If the JIT is actually enabled by default (as far as I remember it is not), it can be useful as a request will simply not process any JIT data if it is not actually used. The current behavior now is to do a compile time JIT (detection of possible GPCES usage). The encoding is useless in 5.3 without unicode support, it is binary as we know it ;) Cheers, -- Pierre http://blog.thepimp.net | http://www.libgd.org --- ## #36595 — johannes@php.net — 2008-03-26T23:54:16+00:00 URL: https://externals.io/message/36536#36595 · In reply to: #36594 > > > […] > > If the JIT is actually enabled by default (as far as I remember it is > not), it can be useful as a request will simply not process any JIT > data if it is not actually used. The current behavior now is to do a > compile time JIT (detection of possible GPCES usage). The encoding is > useless in 5.3 without unicode support, it is binary as we know it ;) And compile time JIT is fine as long as we don't have to care for encoding, right? So this change won't be needed in 5.3. johannes --- ## #36596 — Pierre Joye — 2008-03-27T00:01:45+00:00 URL: https://externals.io/message/36536#36596 · In reply to: #36595 > > And compile time JIT is fine as long as we don't have to care for > encoding, right? So this change won't be needed in 5.3. Except when one sends a xxMB post that you are not expecting and will not use. But that has been so for years, too bad though, 5.3 is the perfect time for such thing. -- Pierre http://blog.thepimp.net | http://www.libgd.org --- ## #36597 — johannes@php.net — 2008-03-27T00:05:30+00:00 URL: https://externals.io/message/36536#36597 · In reply to: #36596 > > > […] > > > […] > > Except when one sends a xxMB post that you are not expecting and will > not use. But that has been so for years, too bad though, 5.3 is the > perfect time for such thing. which is handled by post_max_filesize and as long as we don't add error handling it doesn't make a difference if that happens at compile- or runtime. johannes --- ## #36598 — Pierre Joye — 2008-03-27T00:35:45+00:00 URL: https://externals.io/message/36536#36598 · In reply to: #36597 > > > > […] > > which is handled by post_max_filesize and as long as we don't add error > handling it doesn't make a difference if that happens at compile- or > runtime. Indeed it makes a difference. In one case it is always initialized, in the other only if it is actually used (and not only if a _GET/_POST... exists in the script). That's a pretty big difference. -- Pierre http://blog.thepimp.net | http://www.libgd.org --- ## #36599 — Richard Lynch — 2008-03-27T00:59:47+00:00 URL: https://externals.io/message/36536#36599 · In reply to: #36552 > > […] > > Af far as I know, breaks with arguments do work in PHP. And I even used it once... I could re-factor to work around it in minutes, but there it is... Just one data-point. -- Some people have a "gift" link here. Know what I want? I want you to buy a CD from some indie artist. http://cdbaby.com/from/lynch Yeah, I get a buck. So? --- ## #36601 — Felipe Pena — 2008-03-27T02:22:28+00:00 URL: https://externals.io/message/36536#36601 · In reply to: #36589 2008/3/26, Johannes Schlüter : > > […] > > > No objection, while it adds yet another keyword. > Patch: - http://felipe.ath.cx/diff/goto.diff Patch full (with re2c regenerated files): - http://felipe.ath.cx/diff/goto-with-re2c.diff --- ## #36602 — Felipe Pena — 2008-03-27T02:43:57+00:00 URL: https://externals.io/message/36536#36602 · In reply to: #36589 2008/3/26, Johannes Schlüter : > > […] > > > What exactly does that change? $a = "foo"; list($b) = $a; var_dump($b); Before: string(1) "f" After (HEAD): NULL --- ## #36610 — Lukas Kahwe Smith — 2008-03-27T10:13:34+00:00 URL: https://externals.io/message/36536#36610 · In reply to: #36550 > On Wed, Mar 26, 2008 at 5:39 PM, Stanislav Malyshev > wrote: > > […] > > It was initially done for the input encoding in head as the runtime > encoding may change after RINIT and inside the script. > By the way, the JIT fetch&de/encode part is still to be done in HEAD. I have added this to the php60 todo list. hope you dont mind that i put your name behind this item. at the very least you would know what to do if someone wants to take on this item. regards, Lukas --- ## #36628 — Felipe Pena — 2008-03-27T22:21:04+00:00 URL: https://externals.io/message/36536#36628 · In reply to: #36602 Em Qua, 2008-03-26 às 23:43 -0300, Felipe Pena escreveu: > 2008/3/26, Johannes Schlüter : > > […] > > $a = "foo"; list($b) = $a; var_dump($b); > > Before: > string(1) "f" > > After (HEAD): > `NULL` Hmm, is that being used anywhere at the planet? :D Anyway... Patch: http://felipe.ath.cx/diff/list.diff -- Regards, Felipe Pena. --- ## #36629 — Stanislav Malyshev — 2008-03-27T22:31:54+00:00 URL: https://externals.io/message/36536#36629 · In reply to: #36602 > > […] > > $a = "foo"; list($b) = $a; var_dump($b); This is weird... I wouldn't go as far as saying nobody uses it - since experience shows there's somebody somewhere using any weird combination of PHP features imaginable - but it's definitely a strange code. Even though, I think since it's functionality drop, we may want to wait with it, unless it doesn't work anyway (e.g. produces segfaults in some cases). -- Stanislav Malyshev, Zend Software Architect stas@zend.com http://www.zend.com/ (408)253-8829 MSN: stas@zend.com --- ## #36641 — Hannes Magnusson — 2008-03-28T11:58:12+00:00 URL: https://externals.io/message/36536#36641 · In reply to: #36601 > 2008/3/26, Johannes Schlüter : > > > […] > > Patch: > - http://felipe.ath.cx/diff/goto.diff > > Patch full (with re2c regenerated files): > - http://felipe.ath.cx/diff/goto-with-re2c.diff This should definitely go into 5_3, please commit. -Hannes --- ## #36642 — Hannes Magnusson — 2008-03-28T12:00:54+00:00 URL: https://externals.io/message/36536#36642 · In reply to: #36629 > > […] > > This is weird... I wouldn't go as far as saying nobody uses it - since > experience shows there's somebody somewhere using any weird combination > of PHP features imaginable - but it's definitely a strange code. Even > though, I think since it's functionality drop, we may want to wait with > it, unless it doesn't work anyway (e.g. produces segfaults in some cases). I agree. Any functionality removal, no matter how wtf it might be, shouldn't be merged. If we could however throw `E_DEPRECATED` there, that would be great. -Hannes --- Next page: https://externals.io/message/36536.md?page=2