PHP 5.1 Release Candidate 1

php.internals

Andi Gutmans

21 years ago
Hi all, As planned (a tiny bit delayed), I'd like to RC1 PHP 5.1 within the next few days. I suggest to aim for Tuesday in a week. If there are any critical issues which need addressing please email me. As previously mentioned, I'd like to branch of PHP_5_1 after RC1 in order to allow for the Unicode merge to take place. We'll need about a week of HEAD silence in the core components (Dmitry and Andrei will indicate which ones) so that we can do the merge. Thanks, Andi

Andrei Zmievski

21 years ago
We probably shouldn't merge anything in until after OSCON. When is Dmitriy back? -Andrei On Jul 26, 2005, at 5:50 PM, Andi Gutmans wrote:

Michael Wallner

21 years ago
Hi Andi Gutmans, you wrote:
> As planned (a tiny bit delayed), I'd like to RC1 PHP 5.1 within the next > few days. I suggest to aim for Tuesday in a week. > If there are any critical issues which need addressing please email me.
Is instanceof going to be fixed? If not (without any reason), I'll shut up from now on, but I'd like to suggest to remove the deprecation warning from is_a() as instanceof does not provide the same functionality. Thanks,
-- Michael - < mike(@)php.net >

Jani Taskinen

21 years ago
These should be definately fixed: http://bugs.php.net/bug.php?id=33772&edit=1 (destruct thing) http://bugs.php.net/bug.php?id=33723&edit=1 (apache + php_admin_value issue) These would be nice to get fixed: http://bugs.php.net/bug.php?id=31177&edit=1 (memory leak) http://bugs.php.net/bug.php?id=33598&edit=1 (throw from __destruct -> fatal error) http://bugs.php.net/bug.php?id=30453&edit=1 (a) some engine bug) http://bugs.php.net/bug.php?id=30452&edit=1 (b) some engine bug) http://bugs.php.net/bug.php?id=31341&edit=1 (inconsistency with { escaping) http://bugs.php.net/bug.php?id=25359&edit=1 (array_multisort + global/reference) Some of those I've got bitten by myself too.. :) --Jani On Tue, 26 Jul 2005, Andi Gutmans wrote:
> Hi all, > > As planned (a tiny bit delayed), I'd like to RC1 PHP 5.1 within the next few > days. I suggest to aim for Tuesday in a week. > If there are any critical issues which need addressing please email me. > > As previously mentioned, I'd like to branch of PHP_5_1 after RC1 in order to > allow for the Unicode merge to take place. We'll need about a week of HEAD > silence in the core components (Dmitry and Andrei will indicate which ones) so > that we can do the merge. > > Thanks, > Andi > >
-- Donate @ http://pecl.php.net/wishlist.php/sniper

John Coggeshall

21 years ago
I'd also like to get this in to PHP 5.1+ after we branch before Unicode: http://blog.coggeshall.org/archives/204_Code_Coverage_Support_for_PHP_5.html John

Adam Maccabee Trachtenberg

21 years ago
I can't remember the status of __toString(), but I thought we agreed improving it was a 5.1 feature, pending an investigation to make sure we were able to handle this gracefully in the engine. I don't remember seeing any updates to this code, so I wanted to see if I missed it, or if we decided to push this off. -adam On Wed, 27 Jul 2005, Jani Taskinen wrote:
> > These should be definately fixed: > > http://bugs.php.net/bug.php?id=33772&edit=1 (destruct thing) > http://bugs.php.net/bug.php?id=33723&edit=1 (apache + php_admin_value issue) > > These would be nice to get fixed: > > http://bugs.php.net/bug.php?id=31177&edit=1 (memory leak) > http://bugs.php.net/bug.php?id=33598&edit=1 (throw from __destruct -> fatal error) > http://bugs.php.net/bug.php?id=30453&edit=1 (a) some engine bug) > http://bugs.php.net/bug.php?id=30452&edit=1 (b) some engine bug) > http://bugs.php.net/bug.php?id=31341&edit=1 (inconsistency with { escaping) > http://bugs.php.net/bug.php?id=25359&edit=1 (array_multisort + global/reference) > > Some of those I've got bitten by myself too.. :) > > --Jani > > > On Tue, 26 Jul 2005, Andi Gutmans wrote: > > > Hi all, > > > > As planned (a tiny bit delayed), I'd like to RC1 PHP 5.1 within the next few > > days. I suggest to aim for Tuesday in a week. > > If there are any critical issues which need addressing please email me. > > > > As previously mentioned, I'd like to branch of PHP_5_1 after RC1 in order to > > allow for the Unicode merge to take place. We'll need about a week of HEAD > > silence in the core components (Dmitry and Andrei will indicate which ones) so > > that we can do the merge. > > > > Thanks, > > Andi > > > > > >
-- adam@trachtenberg.com | http://www.trachtenberg.com author of o'reilly's "upgrading to php 5" and "php cookbook" avoid the holiday rush, buy your copies today!

Marcus Börger

21 years ago
Hello Adam, "we accepted that availability of enough excuses to delay it further in the hope we never implement it :-)" To be honest, we all - obviously including you - forgot about the issue completley and only realized it during beta process. And for a thing that need so heavy testing since it must be consistently working all over the evngine it is simply to late. Probably the best is to remind right after the unicode stuff is in and simply commit its basics. During development of 5.2 we can take of it's correctness then. marcus Wednesday, July 27, 2005, 10:10:24 PM, you wrote:

Andi Gutmans

21 years ago
We'll save the discussion for after the Unicode merge but I actually came up with some very good reasons why it's dangerous. Marcus just likes to ignore them :) At 10:51 PM 7/27/2005 +0200, Marcus Boerger wrote:

Andi Gutmans

21 years ago
At 11:59 PM 7/28/2005 +0200, Marcus Boerger wrote:
>Hello Andi, > > na you convinced me. I only wanted to make fun out of the situation and >we really all forgot about the issue, didn't we?
:) Andi

Marcus Börger

21 years ago
Hello Andi, na you convinced me. I only wanted to make fun out of the situation and we really all forgot about the issue, didn't we? regards marcus Thursday, July 28, 2005, 11:45:52 PM, you wrote:
> We'll save the discussion for after the Unicode merge but I actually came > up with some very good reasons why it's dangerous. Marcus just likes to > ignore them :)
> At 10:51 PM 7/27/2005 +0200, Marcus Boerger wrote: >>Hello Adam, >> >> "we accepted that availability of enough excuses to delay it further >> in the hope we never implement it :-)" >> >> To be honest, we all - obviously including you - forgot about the issue >>completley and only realized it during beta process. And for a thing that >>need so heavy testing since it must be consistently working all over the >>evngine it is simply to late. Probably the best is to remind right after >>the unicode stuff is in and simply commit its basics. During development >>of 5.2 we can take of it's correctness then. >> >>marcus >> >>Wednesday, July 27, 2005, 10:10:24 PM, you wrote: >> >> > I can't remember the status of __toString(), but I thought we agreed >> > improving it was a 5.1 feature, pending an investigation to make sure >> > we were able to handle this gracefully in the engine. >> >> > I don't remember seeing any updates to this code, so I wanted to see >> > if I missed it, or if we decided to push this off. >> >> > -adam >> >> > On Wed, 27 Jul 2005, Jani Taskinen wrote: >> >> >> >> >> These should be definately fixed: >> >> >> >> http://bugs.php.net/bug.php?id=33772&edit=1 (destruct thing) >> >> http://bugs.php.net/bug.php?id=33723&edit=1 (apache + >> php_admin_value issue) >> >> >> >> These would be nice to get fixed: >> >> >> >> http://bugs.php.net/bug.php?id=31177&edit=1 (memory leak) >> >> http://bugs.php.net/bug.php?id=33598&edit=1 (throw from >> __destruct -> fatal error) >> >> http://bugs.php.net/bug.php?id=30453&edit=1 (a) some engine bug) >> >> http://bugs.php.net/bug.php?id=30452&edit=1 (b) some engine bug) >> >> http://bugs.php.net/bug.php?id=31341&edit=1 (inconsistency with { >> escaping) >> >> http://bugs.php.net/bug.php?id=25359&edit=1 (array_multisort + >> global/reference) >> >> >> >> Some of those I've got bitten by myself too.. :) >> >> >> >> --Jani >> >> >> >> >> >> On Tue, 26 Jul 2005, Andi Gutmans wrote: >> >> >> >> > Hi all, >> >> > >> >> > As planned (a tiny bit delayed), I'd like to RC1 PHP 5.1 within the >> next few >> >> > days. I suggest to aim for Tuesday in a week. >> >> > If there are any critical issues which need addressing please email me. >> >> > >> >> > As previously mentioned, I'd like to branch of PHP_5_1 after RC1 in >> order to >> >> > allow for the Unicode merge to take place. We'll need about a week >> of HEAD >> >> > silence in the core components (Dmitry and Andrei will indicate >> which ones) so >> >> > that we can do the merge. >> >>-- >>PHP Internals - PHP Runtime Development Mailing List >>To unsubscribe, visit: http://www.php.net/unsub.php
Best regards, Marcus

Andi Gutmans

21 years ago
Is this an extension or changes to the core? Do you already have a patch? In any case, this would have to wait until earliest after the branch and go to HEAD. But I suggest to first make a proposal and show what you're planning on doing here on internals@ Andi At 03:02 PM 7/27/2005 -0400, John Coggeshall wrote:

Marcus Börger

21 years ago
Hello Andi, during apacheCon EU it only needed a few more tweaks which should be done by now (john?). And i can only say it looks really great! But anyway it's sense is to produce more tests where we don't have any yet. And since writing tests requires a lot of work most people hate i guess we can wait a bit longer for it :-/ marcus Thursday, July 28, 2005, 8:31:09 PM, you wrote:
> Is this an extension or changes to the core? Do you already have a patch? > In any case, this would have to wait until earliest after the branch and go > to HEAD. > But I suggest to first make a proposal and show what you're planning on > doing here on internals@
> Andi
> At 03:02 PM 7/27/2005 -0400, John Coggeshall wrote: >>I'd also like to get this in to PHP 5.1+ after we branch before Unicode: >> >>http://blog.coggeshall.org/archives/204_Code_Coverage_Support_for_PHP_5.html >> >>John
Best regards, Marcus

Andi Gutmans

21 years ago
Okay cool. In any case, when it's convenient sending it over to internals@ (maybe after OSCON) makes sense as many here weren't at ApacheCon EU. Andi At 08:37 PM 7/28/2005 +0200, Marcus Boerger wrote:

John Coggeshall

21 years ago
Honestly it's only a patch to the build system to add basically --enable-gcov to the configure (much like --enable-debug). Basically it just checks for the proper libraries and utilities during configure as it should and if everything is okay builds PHP with the proper compile- time flags and libraries. The libgcov takes care of all the instrumenting internally (so no changes to internals code). The difference is that with --enable-gcov PHP produces a ton of .gcda and .gcno files which contain the code coverage data. Enabling at configure also provides two new make commands: make cov (gather up all the .gcda, .gcno files in the build tree from running php and put them in one place then run the lcov utility to process them into a consolidated .phplcov-info.dat file) make cov-html Process the .phplcov-info.dat file and generate a pretty HTML report in the cov_html/ directory It's really a non-intrusive no-code-change patch. Unfortunately I can't seem to get at the CVS right now.. Attached is (basically) the patch. John On Thu, 2005-07-28 at 11:31 -0700, Andi Gutmans wrote:

Nuno Lopes

21 years ago
Today I was trying to reduce the number of tests that fail on my solaris boxes. (because I don't remember to see PHP pass all the tests) I needed to patch a couple of tests that didn't work well on that plataform. Detailed explanation (I hope): ext/date/tests/bug32555.phpt strftime is system/locale dependent. so I've changed it to use date() instead ext/date/tests/bug33532.phpt solaris strftime() doesnt support %z ext/date/tests/date_default_timezone_set-1.phpt ext/date/tests/date_default_timezone_get-1.phpt HAVE_TM_ZONE doesnt get set in solaris, so no "magic" guess is available (and thus no error message is produced) I'm not sure how to fix this. though. (I've tried to use a regex, but it becomes nasty. a simple skip condition would be usefull) ext/iconv/tests/eucjp2iso2022jp.phpt ext/iconv/tests/eucjp2sjis.phpt system's iconv() doesnt support those convertions. I hope somebody review my patch this time, so that it doesnt get lost as the other did (to fix tests on windows). Nuno P.S.: I really like John's idea. I think we could have an automated script on-line to produce those nice pages on a daily basis.

Pierre-Alain Joye

21 years ago
On Tue, 26 Jul 2005 17:50:07 -0700 andi@zend.com (Andi Gutmans) wrote:
> Hi all, > > As planned (a tiny bit delayed), I'd like to RC1 PHP 5.1 within > the next few days. I suggest to aim for Tuesday in a week. > If there are any critical issues which need addressing please > email me.
Mind to create the branche? :)
> As previously mentioned, I'd like to branch of PHP_5_1 after RC1 > in order to allow for the Unicode merge to take place. We'll need > about a week of HEAD silence in the core components (Dmitry and > Andrei will indicate which ones) so that we can do the merge.
I still like to see is_a not deprecated unless we have a valid alternative. As Mike mentioned already, we do not have a replacement for this function, instanceof being useless as it miserably failed when you check against a non existent class. Regards, --Pierre