PHP 5.1 Beta 1

Christopher Kings-Lynne

21 years ago

> I'd like to roll PHP 5.1 Beta 1 very soon.
> Wez, I've been waiting for PDO in order to get this going. Do you feel
> it's mature enough or should I hold off some more time?

Well if someone can help me get sqlstate working in pdo_pgsql I'd be
happy with a release.  At the moment it's just silly that it's not done.
However, I'm finding the PDO code a bit of a nightmare so I cannot
propagate the sqlstates in all cases...

Maybe could I send someone my patch so far?

Chris

Christopher Kings-Lynne

21 years ago

> Well if someone can help me get sqlstate working in pdo_pgsql I'd be
> happy with a release.  At the moment it's just silly that it's not done.
>  However, I'm finding the PDO code a bit of a nightmare so I cannot
> propagate the sqlstates in all cases...
>
> Maybe could I send someone my patch so far?

Also, I need an answer to my post "Concern over ext/pgsql API" on this
list...

Chris

Robert

21 years ago

Hi,

> I would like first to see some bugs closed, like:
> * #30641 - PHP 5.1 doesn't compile on Solaris 9
> * #31725 - PHP is always segfaulting when using sqlite
> * upgrade sqlite and PCRE libraries

Maybe also upgrade GD.

Also, can I humbly suggest we add the Progress Meter / Upload Status
Patch for version 5.1? (http://pdoru.from.ro/upload-progress-meter/)

If the list is ok with this, I'm gonna contact the coder of the patch
and maybe he could add it.


Thanks, Rob

Jan Kneschke

21 years ago

On Sun, Apr 03, 2005 at 03:00:03PM -0700, Robert wrote:

> Hi,
>
> >I would like first to see some bugs closed, like:
> >* #30641 - PHP 5.1 doesn't compile on Solaris 9
> >* #31725 - PHP is always segfaulting when using sqlite
> >* upgrade sqlite and PCRE libraries
>
> Maybe also upgrade GD.
>
> Also, can I humbly suggest we add the Progress Meter / Upload Status
> Patch for version 5.1? (http://pdoru.from.ro/upload-progress-meter/)
>
> If the list is ok with this, I'm gonna contact the coder of the patch
> and maybe he could add it.

If you really need it, take a little help from the webserver:

http://weigon.dyndns.org/upload-progress.html

I'm sure that if necessary this should be possible to add to apache too.

The get an idea how it works:
http://www.lighttpd.net/documentation/upload-progress.html

> Thanks, Rob

Jan

--
Jan Kneschke                                     http://jan.kneschke.de/
Perhaps you want to say 'thank you, jan':    http://jk.123.org/wishlist/

Andi Gutmans

21 years ago

Guys,

Not that I disagree with resolving these problems, but it has nothing to do
with beta 1 which was planned to get wider testing and feedback, mainly
about PDO which is the main new code base in 5.1.
It won't wait for any of these things although I'll be happy if people
resolve them :)

Andi

At 03:00 PM 4/3/2005 -0700, Robert wrote:

Nuno Lopes

21 years ago

But at least you could allow Solaris guys to test PHP 5.1 beta 1 :)

Nuno

----- Original Message -----

Andi Gutmans

21 years ago

Do you have a fix?
Andi

At 03:15 PM 4/3/2005 +0100, Nuno Lopes wrote:

Nuno Lopes

21 years ago

Patch attached (explanation in the bug report).
Nuno


----- Original Message -----

Andi Gutmans

21 years ago

thanks. will check it out.

At 05:16 PM 4/3/2005 +0100, Nuno Lopes wrote:

Wez Furlong

21 years ago

Please send it to me.

--Wez.

On Apr 2, 2005 11:02 PM, Christopher Kings-Lynne
<chriskl@familyhealth.com.au> wrote:

Nuno Lopes

21 years ago

I would like first to see some bugs closed, like:
* #30641 - PHP 5.1 doesn't compile on Solaris 9
* #31725 - PHP is always segfaulting when using sqlite
* upgrade sqlite and PCRE libraries

Thanks,
Nuno


----- Original Message -----

Antony Dovgal

21 years ago

On Sun, 3 Apr 2005 12:48:59 +0100
"Nuno Lopes" <nlopess@php.net> wrote:

> I would like first to see some bugs closed, like:
>  * #30641 - PHP 5.1 doesn't compile on Solaris 9
>  * #31725 - PHP is always segfaulting when using sqlite
> * upgrade sqlite and PCRE libraries

There are also some bugs that will break BC:
#32427 (Interfaces not allowed access modifiers)
#32428	(The @ warning error supression operator is broken) - this one looks like a major problem to me.

And some bugs that I'd like to see fixed before we release anything:

#29210 (is_callable - no support for private and protected methods)
#29015 (Incorrect behavior of class members with empty names, because ZE2 incorrectly
considers that all property names that start with '\0' are mangled).

None of these bugs are too hard to fix, so why not to fix them before the first beta?

--
Wbr,
Antony Dovgal aka tony2001
antony@zend.com

Jani Taskinen

21 years ago

Maybe we should first get all the tests pass before rolling
any kind of release? Here's the list of currently failing tests:
(note: some might not be bugs but just bad tests..)

ZE2 ArrayAccess::offsetGet ambiguties [tests/classes/array_access_003.phpt]
ZE2 ArrayAccess and sub Arrays [tests/classes/array_access_005.phpt]
ZE2 ArrayAccess and ASSIGN_OP operators (.=) [tests/classes/array_access_008.phpt]
ZE2 Serializeable [tests/classes/serialize_001.phpt]
Test for buffering in core functions with implicit flush off [tests/func/008.phpt]
Bug #22836 (returning references to NULL) [Zend/tests/bug22836.phpt]
unset() CV 5 (indirect unset() of global variable in session_start()) [Zend/tests/unset_cv05.phpt]
Bug #24155 (gdImageRotate270 rotation problem). [ext/gd/tests/bug24155.phpt]
Bug #27582 (ImageFillToBorder() on alphablending image looses alpha on fill color) [ext/gd/tests/bug27582_1.phpt]
iconv stream filter [ext/iconv/tests/iconv_stream_filter.phpt]
HTML input/output [ext/mbstring/tests/htmlent.phpt]
mb_output_handler() (Shift_JIS) [ext/mbstring/tests/mb_output_handler_shift_jis.phpt]
mb_send_mail() test 5 (lang=Simplified Chinese) [ext/mbstring/tests/mb_send_mail05.phpt]
mb_send_mail() test 6 (lang=Traditional Chinese) [ext/mbstring/tests/mb_send_mail06.phpt]
mb_send_mail() test 7 (lang=Korean) [ext/mbstring/tests/mb_send_mail07.phpt]
mysql connect [ext/mysql/tests/001.phpt]
Bug #29992 (foreach by reference corrupts the array) [ext/standard/tests/array/bug29992.phpt]
Bug #31213 (Sideeffects caused by bug #29493) [ext/standard/tests/array/bug31213.phpt]
various fputcsv() functionality tests [ext/standard/tests/file/fputcsv.phpt]
User-space streams [ext/standard/tests/file/userstreams.phpt]
Check for mktime with out-of-range parameters [ext/standard/tests/time/003.phpt]
Bug #20382 (strtotime ("Monday", $date) produces wrong result on DST changeover) [ext/standard/tests/time/bug20382.phpt]
Bug #21966 (date() or mktime() returning bad value for mktime month param of '2') [ext/standard/tests/time/bug21966.phpt]
Bug #27719: mktime returns incorrect timestamp for dst days [ext/standard/tests/time/bug27719.phpt]
Bug #32001 (xml_parse*() goes into infinite loop when autodetection in effect) [ext/xml/tests/bug32001.phpt]


--Jani


On Sat, 2 Apr 2005, Andi Gutmans wrote:

Andi Gutmans

21 years ago

Jani,

We're talking beta (I don't mind calling it alpha either) to get PDO and
the new engine architecture tested by more people.
It doesn't mean all tests need to pass. Don't worry, bugs will be addressed
before a 5.1 release which is still very far off. But don't expect the
whole bugs.php.net DB to clean out either.

Andi

At 03:40 PM 4/3/2005 +0300, Jani Taskinen wrote:

Jani Taskinen

21 years ago

Then call it pre-alpha.

--Jani

Marcus Boerger

21 years ago

Hello Jani,


yes that'd much better then calling it beta. Beta would prevent us from
changes and according to my history there are several things just missing
right now.

Sunday, April 3, 2005, 9:32:49 AM, you wrote:

>      Then call it pre-alpha.

>      --Jani

>> We're talking beta (I don't mind calling it alpha either) to get PDO and the
>> new engine architecture tested by more people.
>> It doesn't mean all tests need to pass. Don't worry, bugs will be addressed
>> before a 5.1 release which is still very far off. But don't expect the whole
>> bugs.php.net DB to clean out either.
>>
>> Andi
>>
>> At 03:40 PM 4/3/2005 +0300, Jani Taskinen wrote:
>>
>>>     Maybe we should first get all the tests pass before rolling
>>>     any kind of release? Here's the list of currently failing tests:
>>>     (note: some might not be bugs but just bad tests..)
>>>
>>>     ZE2 ArrayAccess::offsetGet ambiguties
>>> [tests/classes/array_access_003.phpt]
>>>     ZE2 ArrayAccess and sub Arrays [tests/classes/array_access_005.phpt]
>>>     ZE2 ArrayAccess and ASSIGN_OP operators (.=)
>>> [tests/classes/array_access_008.phpt]
>>>     ZE2 Serializeable [tests/classes/serialize_001.phpt]
>>>     Test for buffering in core functions with implicit flush off
>>> [tests/func/008.phpt]
>>>     Bug #22836 (returning references to NULL) [Zend/tests/bug22836.phpt]
>>>     unset() CV 5 (indirect unset() of global variable in session_start())
>>> [Zend/tests/unset_cv05.phpt]
>>>     Bug #24155 (gdImageRotate270 rotation problem).
>>> [ext/gd/tests/bug24155.phpt]
>>>     Bug #27582 (ImageFillToBorder() on alphablending image looses alpha on
>>> fill color) [ext/gd/tests/bug27582_1.phpt]
>>>     iconv stream filter [ext/iconv/tests/iconv_stream_filter.phpt]
>>>     HTML input/output [ext/mbstring/tests/htmlent.phpt]
>>>     mb_output_handler() (Shift_JIS)
>>> [ext/mbstring/tests/mb_output_handler_shift_jis.phpt]
>>>     mb_send_mail() test 5 (lang=Simplified Chinese)
>>> [ext/mbstring/tests/mb_send_mail05.phpt]
>>>     mb_send_mail() test 6 (lang=Traditional Chinese)
>>> [ext/mbstring/tests/mb_send_mail06.phpt]
>>>     mb_send_mail() test 7 (lang=Korean)
>>> [ext/mbstring/tests/mb_send_mail07.phpt]
>>>     mysql connect [ext/mysql/tests/001.phpt]
>>>     Bug #29992 (foreach by reference corrupts the array)
>>> [ext/standard/tests/array/bug29992.phpt]
>>>     Bug #31213 (Sideeffects caused by bug #29493)
>>> [ext/standard/tests/array/bug31213.phpt]
>>>     various fputcsv() functionality tests
>>> [ext/standard/tests/file/fputcsv.phpt]
>>>     User-space streams [ext/standard/tests/file/userstreams.phpt]
>>>     Check for mktime with out-of-range parameters
>>> [ext/standard/tests/time/003.phpt]
>>>     Bug #20382 (strtotime ("Monday", $date) produces wrong result on DST
>>> changeover) [ext/standard/tests/time/bug20382.phpt]
>>>     Bug #21966 (date() or mktime() returning bad value for mktime month
>>> param of '2') [ext/standard/tests/time/bug21966.phpt]
>>>     Bug #27719: mktime returns incorrect timestamp for dst days
>>> [ext/standard/tests/time/bug27719.phpt]
>>>     Bug #32001 (xml_parse*() goes into infinite loop when autodetection in
>>> effect) [ext/xml/tests/bug32001.phpt]
>>>
>>>
>>>     --Jani
>>>
>>>
>>> On Sat, 2 Apr 2005, Andi Gutmans wrote:
>>>
>>>> Hey,
>>>>
>>>> I'd like to roll PHP 5.1 Beta 1 very soon.
>>>> Wez, I've been waiting for PDO in order to get this going. Do you feel it's
>>>> mature enough or should I hold off some more time?
>>>> Also, has someone already integrated PDO configure switches with the old
>>>> --with-db configure switches? (i.e. the idea was that --with-mysql would
>>>> enable pdo-mysql too so that we don't get lost with a million configure
>>>> switches).
>>>>
>>>> Andi
>>>>

--
Best regards,
Marcus                            mailto:mail@marcus-boerger.de

Derick Rethans

21 years ago

On Sun, 3 Apr 2005, Andi Gutmans wrote:

> Jani,
>
> We're talking beta (I don't mind calling it alpha either) to get PDO and the
> new engine architecture tested by more people.

I think we should call it alpha too...

Derick

--
Derick Rethans
http://derickrethans.nl | http://ez.no | http://xdebug.org

Andi Gutmans

21 years ago

I'm fine with calling it Alpha especially as PDO might still change stuff.

At 10:05 AM 4/3/2005 -0400, Derick Rethans wrote:

Sean Coates

21 years ago

Andi Gutmans wrote:

> I'd like to roll PHP 5.1 Beta 1 very soon.
> Wez, I've been waiting for PDO in order to get this going. Do you feel
> it's mature enough or should I hold off some more time?

Before we go beta, I'd like to see php_check_syntax() disappear (as it
seems it will never be "fixed").

It's had function table problems since its inception (IIRC, it runs
php_lint_script -- but php_lint_script was never designed to clean up
after itself; the process was supposed to terminate after a php -l)

I spoke with Ilia (and Rasmus) a few nights ago, and they agreed that it
would be best to get rid of php_check_syntax() altogether (it would take
too much effort to fix, and we already have a mechanism for syntax
checking (php -l)).

Anyway, if we're going Alpha, this isn't an issue, yet -- if we're going
Beta, then I'd like to see it removed.

S

Andi Gutmans

21 years ago

I'll go Alpha. I'll try and roll towards the end of next week.
I agree with Ilia and Rasmus and suggest to nuke it.

Andi

At 11:48 AM 4/3/2005 -0400, Sean Coates wrote:

Marcus Boerger

21 years ago

Hello Andi,

Saturday, April 2, 2005, 10:53:19 PM, you wrote:

> Hey,

> I'd like to roll PHP 5.1 Beta 1 very soon.
> Wez, I've been waiting for PDO in order to get this going. Do you feel it's
> mature enough or should I hold off some more time?
> Also, has someone already integrated PDO configure switches with the old
> --with-db configure switches? (i.e. the idea was that --with-mysql would
> enable pdo-mysql too so that we don't get lost with a million configure
> switches).

> Andi

Can we get some things straight first (i guess that's the idea of an alpha).
I'd like to reintroduce __toString() since 5.1 is supposed to make it work
again. The patching should be easy enough since we left in most of the
necessary code as we planned to reintroduce it in the first place.

Then i think we need to enable return by reference in c functions to get
array overloading fixed. For thta you have a patch from me. As you have for
several other things :-)

However most important seems to be fixing return by reference. Derick and me
already did a lot of analysis, weeks of and and. Maybe you want to help.

regards
marcus

Adam Maccabee Trachtenberg

21 years ago

On Sun, 3 Apr 2005, Marcus Boerger wrote:

> Can we get some things straight first (i guess that's the idea of an alpha).
> I'd like to reintroduce __toString() since 5.1 is supposed to make it work
> again. The patching should be easy enough since we left in most of the
> necessary code as we planned to reintroduce it in the first place.

This is also my own list of things to fix in 5.1.

For example, look at how Rasmus has to use string casting in SimpleXML:

http://toys.lerdorf.com/archives/33-Buzzing-the-Yahoo!-Search-Web-Services.html

IMHO, needing to cast for such basic behavior is not in the spirit of
PHP, and it's certainly not simple.

-adam

--
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!

Andi Gutmans

21 years ago

I agree and that's why we made the opcode changes to make them re-entrant.
Will look into it. I was sure we already uncommented that code :)

At 08:54 PM 4/3/2005 -0400, Adam Maccabee Trachtenberg wrote:

Sebastian Bergmann

21 years ago

Andi Gutmans wrote:

> I'd like to roll PHP 5.1 Beta 1 very soon.

What about the ability to declare type-hinted parameters as optional?

--
Sebastian Bergmann                      http://www.sebastian-bergmann.de/
GnuPG Key: 0xB85B5D69 / 27A7 2B14 09E4 98CD 6277 0E5B 6867 C514 B85B 5D69

Wez Furlong

21 years ago

Hey Andi,

I've been a busy with work (and house hunting), so I've been a bit out of it.
Let me catch up and take stock on PDO again; I'd like to push a PECL
release of the various packages this weekend, before we go with Beta
1.

I can't think of any show-stoppers in PDO right now, so the time to
Beta 1 should/can be sort.

--Wez.

On Apr 2, 2005 10:53 PM, Andi Gutmans <andi@zend.com> wrote:

Marcus Boerger

21 years ago

Hello Wez,

Wednesday, April 6, 2005, 2:48:22 PM, you wrote:

> Hey Andi,

> I've been a busy with work (and house hunting), so I've been a bit out of it.
> Let me catch up and take stock on PDO again; I'd like to push a PECL
> release of the various packages this weekend, before we go with Beta
> 1.

> I can't think of any show-stoppers in PDO right now, so the time to
> Beta 1 should/can be sort.

at the moment the only problem is that the mysql driver is only available
for the old protocol but fortunatley hartmut said he'll take care of that.
To avoid confusion i'd like to see the old driver as 'mysql3' and thave the
new one as 'mysql' something like we have for sqlite.


Best regards,
Marcus                            mailto:mail@marcus-boerger.de

Wez Furlong

21 years ago

The plan is to have only one mysql driver that can be linked against
both version 3 and 4 of the client libraries.

--Wez.

On Apr 6, 2005 9:56 AM, Marcus Boerger <mail@marcus-boerger.de> wrote:

Marcus Boerger

21 years ago

Hello Wez,

i very much dislike that idea because a mysql3 will never be working as
expected. It can never support multiple statements at a time or transactions
or whatnot. Thus i pretty much want something that keeps me from using it
until i am pretty sure i really can live with that. A distingushing by
having to add the version for the old stuff would lead to that.

marcus

Wednesday, April 6, 2005, 4:16:33 PM, you wrote:

> The plan is to have only one mysql driver that can be linked against
> both version 3 and 4 of the client libraries.

> --Wez.

> On Apr 6, 2005 9:56 AM, Marcus Boerger <mail@marcus-boerger.de> wrote:
>> Hello Wez,
>>
>> Wednesday, April 6, 2005, 2:48:22 PM, you wrote:
>>
>> > Hey Andi,
>>
>> > I've been a busy with work (and house hunting), so I've been a bit out of it.
>> > Let me catch up and take stock on PDO again; I'd like to push a PECL
>> > release of the various packages this weekend, before we go with Beta
>> > 1.
>>
>> > I can't think of any show-stoppers in PDO right now, so the time to
>> > Beta 1 should/can be sort.
>>
>> at the moment the only problem is that the mysql driver is only available
>> for the old protocol but fortunatley hartmut said he'll take care of that.
>> To avoid confusion i'd like to see the old driver as 'mysql3' and thave the
>> new one as 'mysql' something like we have for sqlite.
>>
>>
>> Best regards,
>>  Marcus                            mailto:mail@marcus-boerger.de
>>
>>

--
Best regards,
Marcus                            mailto:mail@marcus-boerger.de

Wez Furlong

21 years ago

But neither can most other drivers.
Since the majority of the API is the same, what sense is there in
duplicating the code?

--Wez.

On Apr 6, 2005 10:22 AM, Marcus Boerger <mail@marcus-boerger.de> wrote:

Ilia A.

21 years ago

Marcus Boerger wrote:

>   i very much dislike that idea because a mysql3 will never be working as
> expected. It can never support multiple statements at a time or transactions
> or whatnot. Thus i pretty much want something that keeps me from using it
> until i am pretty sure i really can live with that. A distingushing by
> having to add the version for the old stuff would lead to that.

I agree with Wez, there seems to be little point in having two virtually
identical drivers for MySQL in PDO. One common driver, where in one
instance some functionality is emulated and in other natively supported.
BTW In MySQL 3.23 transaction are possible, but you need to use innodb
tables to "access" them. Having one driver would also make life simpler
for people who need to work with both MySQL 3.23, 4.0 and 4.1

Ilia

Andi Gutmans

21 years ago

OK so keep me posted in the beginning of next week.
I'll also look at some of the other mentioned issues.

At 08:48 AM 4/6/2005 -0400, Wez Furlong wrote: