From ms419 at freezone.co.uk  Mon Jan  1 19:06:20 2007
From: ms419 at freezone.co.uk (Jack Bates)
Date: Mon, 01 Jan 2007 19:06:20 -0800
Subject: php domxml extension
Message-ID: <1167707180.5574.20.camel@fis.lat>

I'm working on an OpenId module for the Gallery PHP photo album:
http://gallery-contrib.svn.sourceforge.net/viewvc/gallery-contrib/trunk/gallery2/modules/openid/

The Yadis library claims to require either the domxml or dom PHP
extensions. These are beyond Gallery's current minimum requirements. Can
anyone comment why the PHP XML extension is not sufficient for OpenID?

Thanks! Jack
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 242 bytes
Desc: This is a digitally signed message part
Url : http://lists.openidenabled.com/pipermail/dev/attachments/20070101/37e76603/attachment-0001.pgp 

From cygnus at janrain.com  Tue Jan  2 10:43:19 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Tue, 2 Jan 2007 10:43:19 -0800
Subject: php domxml extension
In-Reply-To: <1167707180.5574.20.camel@fis.lat>
References: <1167707180.5574.20.camel@fis.lat>
Message-ID: <20070102184319.GI9343@janrain.com>

# The Yadis library claims to require either the domxml or dom PHP
# extensions. These are beyond Gallery's current minimum
# requirements. Can anyone comment why the PHP XML extension is not
# sufficient for OpenID?

The 'xml' (expat) extension does not appear to support Xpath, which
would be why we didn't implement support for it.  (If it does support
Xpath, please let me know where such documentation can be found, and
please feel free to submit a patch to subclass
Services_Yadis_XMLParser to implement a wrapper for it.)  Neither the
PHP documentation nor the expat documentation makes any mention of
Xpath usage.

-- 
  Jonathan Daugherty
  JanRain, Inc.


From mike at nthwave.net  Wed Jan  3 07:48:08 2007
From: mike at nthwave.net (Michael Mell)
Date: Wed, 3 Jan 2007 07:48:08 -0800
Subject: JanRain OpenID ruby lib 1.2?
Message-ID: <F223F0BA-5B5A-4739-B8B9-FA60A87466F9@nthwave.net>

Is the http://www.openidenabled.com/openid/libraries/ruby/download/  
page supposed to contain a link to the JanRain OpenID 1.2 ruby lib? I  
see the php download page has the 1.2 lib and the ruby Heraldry  
Consumer plugin requires it.  I can't find it anywhere.

Also on http://openid.net/ under Developers in the right bar the link  
to Libraries should point to the new location. It currently points to  
a "This page has moved" notice.

Thanks,
Mike



From brian at janrain.com  Wed Jan  3 11:08:37 2007
From: brian at janrain.com (Brian Ellin)
Date: Wed, 3 Jan 2007 11:08:37 -0800
Subject: JanRain OpenID ruby lib 1.2?
In-Reply-To: <F223F0BA-5B5A-4739-B8B9-FA60A87466F9@nthwave.net>
References: <F223F0BA-5B5A-4739-B8B9-FA60A87466F9@nthwave.net>
Message-ID: <9674d5350701031108o6f7465feke07fb405322ce277@mail.gmail.com>

Hi Michael,

You should use the 1.3 version of the library with the Heraldry
consumer plugin.  It is a bugfix release of the 1.2 version.  You may
download the source (it also needs the ruby-yadis package):

http://www.openidenabled.com/resources/downloads/ruby-openid/ruby-openid-1.1.3.tar.gz
http://www.openidenabled.com/resources/downloads/ruby-yadis/ruby-yadis-0.3.3.tar.gz

or use rubygems to install:

gem install ruby-openid --include-dependencies

Hope that helps,
Brian Ellin

On 1/3/07, Michael Mell <mike at nthwave.net> wrote:
> Is the http://www.openidenabled.com/openid/libraries/ruby/download/
> page supposed to contain a link to the JanRain OpenID 1.2 ruby lib? I
> see the php download page has the 1.2 lib and the ruby Heraldry
> Consumer plugin requires it.  I can't find it anywhere.
>
> Also on http://openid.net/ under Developers in the right bar the link
> to Libraries should point to the new location. It currently points to
> a "This page has moved" notice.
>
> Thanks,
> Mike
>
>
> _______________________________________________
> Dev mailing list
> Dev at lists.openidenabled.com
> http://lists.openidenabled.com/mailman/listinfo/dev
>


From mike at nthwave.net  Wed Jan  3 11:40:08 2007
From: mike at nthwave.net (Michael Mell)
Date: Wed, 3 Jan 2007 11:40:08 -0800
Subject: JanRain OpenID ruby lib 1.2?
In-Reply-To: <9674d5350701031108o6f7465feke07fb405322ce277@mail.gmail.com>
References: <F223F0BA-5B5A-4739-B8B9-FA60A87466F9@nthwave.net>
	<9674d5350701031108o6f7465feke07fb405322ce277@mail.gmail.com>
Message-ID: <461fe4100d0385c394c8fd3c0fe115c2@nthwave.net>

Hi Brian,
That does help.
I saw the 1.1.3 version on the download page. I assumed that 1.2 would  
be a later, not an earlier, revision. I see now that when you refer to  
1.3 you mean 1.1.3 and 1.2 refers to 1.1.2. That's confusing for those  
not in the know especially when there really is a 1.2.1 version of the  
OpenID lib for php.


Thanks,
Mike

On Jan 3, 2007, at 11:08 AM, Brian Ellin wrote:

> Hi Michael,
>
> You should use the 1.3 version of the library with the Heraldry
> consumer plugin.  It is a bugfix release of the 1.2 version.  You may
> download the source (it also needs the ruby-yadis package):
>
> http://www.openidenabled.com/resources/downloads/ruby-openid/ruby- 
> openid-1.1.3.tar.gz
> http://www.openidenabled.com/resources/downloads/ruby-yadis/ruby- 
> yadis-0.3.3.tar.gz
>
> or use rubygems to install:
>
> gem install ruby-openid --include-dependencies
>
> Hope that helps,
> Brian Ellin
>
> On 1/3/07, Michael Mell <mike at nthwave.net> wrote:
>> Is the http://www.openidenabled.com/openid/libraries/ruby/download/
>> page supposed to contain a link to the JanRain OpenID 1.2 ruby lib? I
>> see the php download page has the 1.2 lib and the ruby Heraldry
>> Consumer plugin requires it.  I can't find it anywhere.
>>
>> Also on http://openid.net/ under Developers in the right bar the link
>> to Libraries should point to the new location. It currently points to
>> a "This page has moved" notice.
>>
>> Thanks,
>> Mike
>>
>>
>> _______________________________________________
>> Dev mailing list
>> Dev at lists.openidenabled.com
>> http://lists.openidenabled.com/mailman/listinfo/dev
>>
>



From cygnus at janrain.com  Wed Jan  3 11:41:07 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Wed, 3 Jan 2007 11:41:07 -0800
Subject: JanRain OpenID ruby lib 1.2?
In-Reply-To: <461fe4100d0385c394c8fd3c0fe115c2@nthwave.net>
References: <F223F0BA-5B5A-4739-B8B9-FA60A87466F9@nthwave.net>
	<9674d5350701031108o6f7465feke07fb405322ce277@mail.gmail.com>
	<461fe4100d0385c394c8fd3c0fe115c2@nthwave.net>
Message-ID: <20070103194106.GJ9343@janrain.com>

# That's confusing for those not in the know especially when there
# really is a 1.2.1 version of the OpenID lib for php.

I guess this isn't the first time someone thought that.  The versions
of our various libraries are not related, for the record.  It's
impossible for us to *keep* them the same (at least in terms of minor
releases), as some libraries require bugfix releases that others do
not.

-- 
  Jonathan Daugherty
  JanRain, Inc.


From norman at rasmussen.co.za  Thu Jan  4 01:29:03 2007
From: norman at rasmussen.co.za (Norman Rasmussen)
Date: Thu, 4 Jan 2007 11:29:03 +0200
Subject: JanRain OpenID ruby lib 1.2?
In-Reply-To: <20070103194106.GJ9343@janrain.com>
References: <F223F0BA-5B5A-4739-B8B9-FA60A87466F9@nthwave.net>
	<9674d5350701031108o6f7465feke07fb405322ce277@mail.gmail.com>
	<461fe4100d0385c394c8fd3c0fe115c2@nthwave.net>
	<20070103194106.GJ9343@janrain.com>
Message-ID: <5b698f5a0701040129n6bfb5252xf160c5f0be1a6d22@mail.gmail.com>

iirc tortoisesvn numbers it's version : upstreamver_productver.

eg: 1.2.1-1.3 could be used for the new ruby lib

On 1/3/07, Jonathan Daugherty <cygnus at janrain.com> wrote:
> # That's confusing for those not in the know especially when there
> # really is a 1.2.1 version of the OpenID lib for php.
>
> I guess this isn't the first time someone thought that.  The versions
> of our various libraries are not related, for the record.  It's
> impossible for us to *keep* them the same (at least in terms of minor
> releases), as some libraries require bugfix releases that others do
> not.
>
> --
>   Jonathan Daugherty
>   JanRain, Inc.
>
> _______________________________________________
> Dev mailing list
> Dev at lists.openidenabled.com
> http://lists.openidenabled.com/mailman/listinfo/dev
>


-- 
- Norman Rasmussen
 - Email: norman at rasmussen.co.za
 - Home page: http://norman.rasmussen.co.za/


From mike at nthwave.net  Thu Jan  4 07:14:07 2007
From: mike at nthwave.net (Michael Mell)
Date: Thu, 4 Jan 2007 07:14:07 -0800
Subject: JanRain OpenID ruby lib 1.2?
In-Reply-To: <20070103194106.GJ9343@janrain.com>
References: <F223F0BA-5B5A-4739-B8B9-FA60A87466F9@nthwave.net>
	<9674d5350701031108o6f7465feke07fb405322ce277@mail.gmail.com>
	<461fe4100d0385c394c8fd3c0fe115c2@nthwave.net>
	<20070103194106.GJ9343@janrain.com>
Message-ID: <2e31a5dde3b4f30f9f094cd0de27c99d@nthwave.net>


On Jan 3, 2007, at 11:41 AM, Jonathan Daugherty wrote:

> # That's confusing for those not in the know especially when there
> # really is a 1.2.1 version of the OpenID lib for php.
>
> I guess this isn't the first time someone thought that.  The versions
> of our various libraries are not related, for the record.  It's
> impossible for us to *keep* them the same (at least in terms of minor
> releases), as some libraries require bugfix releases that others do
> not.

Thanks Jonathan. That's understandable. My main point is that  
documentation and conversation from JanRain has been dropping the "1"  
in 1.1.3. For example, the README from the Heraldry OpenID Consumer  
Plugin [1] says
<quote>
PRE-REQUISITES
--------------
* JanRain's Yadis and OpenID 1.2 libraries in Ruby.
</quote>

and yet there is *no* JanRain OpenID 1.2 library in Ruby.

Mike

[1]  
http://svn.apache.org/repos/asf/incubator/heraldry/libraries/ruby/ 
openid_consumer_plugin/trunk/openid_consumer/README



From cygnus at janrain.com  Thu Jan  4 10:38:58 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Thu, 4 Jan 2007 10:38:58 -0800
Subject: JanRain OpenID ruby lib 1.2?
In-Reply-To: <2e31a5dde3b4f30f9f094cd0de27c99d@nthwave.net>
References: <F223F0BA-5B5A-4739-B8B9-FA60A87466F9@nthwave.net>
	<9674d5350701031108o6f7465feke07fb405322ce277@mail.gmail.com>
	<461fe4100d0385c394c8fd3c0fe115c2@nthwave.net>
	<20070103194106.GJ9343@janrain.com>
	<2e31a5dde3b4f30f9f094cd0de27c99d@nthwave.net>
Message-ID: <20070104183858.GP9343@janrain.com>

# PRE-REQUISITES
# --------------
# * JanRain's Yadis and OpenID 1.2 libraries in Ruby.
# </quote>
# 
# and yet there is *no* JanRain OpenID 1.2 library in Ruby.

That's a documentation bug; that is referring to the OpenID *protocol*
version 1.2, not the library version.  It can be confusing.  (But
generally speaking, any version number following the word "OpenID"
refers to the version of OpenID supported by the implementation in
question.)

-- 
  Jonathan Daugherty
  JanRain, Inc.


From norman at rasmussen.co.za  Thu Jan  4 12:37:25 2007
From: norman at rasmussen.co.za (Norman Rasmussen)
Date: Thu, 4 Jan 2007 22:37:25 +0200
Subject: JanRain OpenID ruby lib 1.2?
In-Reply-To: <20070104183858.GP9343@janrain.com>
References: <F223F0BA-5B5A-4739-B8B9-FA60A87466F9@nthwave.net>
	<9674d5350701031108o6f7465feke07fb405322ce277@mail.gmail.com>
	<461fe4100d0385c394c8fd3c0fe115c2@nthwave.net>
	<20070103194106.GJ9343@janrain.com>
	<2e31a5dde3b4f30f9f094cd0de27c99d@nthwave.net>
	<20070104183858.GP9343@janrain.com>
Message-ID: <5b698f5a0701041237j373880f9v8dc0c8fcd2b264e5@mail.gmail.com>

/me grins to himself, extending the naming scheme:

1.2-1.1.2-1.3 (openid-library-frontend)

On 1/4/07, Jonathan Daugherty <cygnus at janrain.com> wrote:
> # PRE-REQUISITES
> # --------------
> # * JanRain's Yadis and OpenID 1.2 libraries in Ruby.
> # </quote>
> #
> # and yet there is *no* JanRain OpenID 1.2 library in Ruby.
>
> That's a documentation bug; that is referring to the OpenID *protocol*
> version 1.2, not the library version.  It can be confusing.  (But
> generally speaking, any version number following the word "OpenID"
> refers to the version of OpenID supported by the implementation in
> question.)
>
> --
>   Jonathan Daugherty
>   JanRain, Inc.
>
> _______________________________________________
> Dev mailing list
> Dev at lists.openidenabled.com
> http://lists.openidenabled.com/mailman/listinfo/dev
>


-- 
- Norman Rasmussen
 - Email: norman at rasmussen.co.za
 - Home page: http://norman.rasmussen.co.za/


From j at justinburger.com  Sat Jan  6 14:32:32 2007
From: j at justinburger.com (Justin Burger)
Date: Sat, 6 Jan 2007 14:32:32 -0800
Subject: PHP Adding Multi Optional Fields.
Message-ID: <777D1187-0123-41C3-A651-7504C4EC257C@justinburger.com>

Hi,
I am attempting to get the PHP example consumer working with more  
than one optional field; I was assuming (according to the example)  
that I use the
"addExtensionArg" function for each field, but when I attempt that,  
it only seems to pick up the last field passed thru that function..  
for example:

$auth_request->addExtensionArg('sreg', 'optional', 'email');
$auth_request->addExtensionArg('sreg', 'optional', 'country');

only prompts for country...

Am I defining this incorrectly? That is the only thing I changed in  
the example code.


Thanks in advance,
Justin.



From cygnus at janrain.com  Sat Jan  6 18:31:11 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Sat, 6 Jan 2007 18:31:11 -0800
Subject: PHP Adding Multi Optional Fields.
In-Reply-To: <777D1187-0123-41C3-A651-7504C4EC257C@justinburger.com>
References: <777D1187-0123-41C3-A651-7504C4EC257C@justinburger.com>
Message-ID: <20070107023110.GC24213@janrain.com>

# $auth_request->addExtensionArg('sreg', 'optional', 'email');
# $auth_request->addExtensionArg('sreg', 'optional', 'country');

The reason this doesn't work is because 'optional' is one arg whose
value should be a comma-separated list of field names, as per the Sreg
spec.  The correct code is:

  $auth_request->addExtensionArg('sreg', 'optional', 'email,country');

-- 
  Jonathan Daugherty
  JanRain, Inc.


From dd at davedash.com  Tue Jan  9 20:03:31 2007
From: dd at davedash.com (Dave Dash)
Date: Tue, 9 Jan 2007 22:03:31 -0600
Subject: No XML parser found w/ PHP 5.2.0
Message-ID: <694fa1450701092003h5dfa74cag551b22a7e3631927@mail.gmail.com>

I keep getting this error:

Fatal error: No XML parser was found in
/usr/local/php5/lib/php/Services/Yadis/XML.php on line 352

I'm confused as to what I need to install in order for it to work.  AFAIK
PHP5 comes with DOM ... actually I'm pretty sure of it since phpinfo() says
so.

Any clues as to what I can do to make this work?

Thanks

-d

-- 
Dave Dash
612.670.0621
Discover your favorite restaurant: reviewsby.us
gtalk: dave.dash
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.openidenabled.com/pipermail/dev/attachments/20070109/0ec2cd04/attachment-0001.html 

From cygnus at janrain.com  Wed Jan 10 09:29:09 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Wed, 10 Jan 2007 09:29:09 -0800
Subject: No XML parser found w/ PHP 5.2.0
In-Reply-To: <694fa1450701092003h5dfa74cag551b22a7e3631927@mail.gmail.com>
References: <694fa1450701092003h5dfa74cag551b22a7e3631927@mail.gmail.com>
Message-ID: <20070110172909.GG5268@janrain.com>

# Fatal error: No XML parser was found in
# /usr/local/php5/lib/php/Services/Yadis/XML.php on line 352
# 
# I'm confused as to what I need to install in order for it to work.
# AFAIK PHP5 comes with DOM ... actually I'm pretty sure of it since
# phpinfo() says so.

Which version of the OpenID library are you using?

-- 
  Jonathan Daugherty
  JanRain, Inc.


From dd at davedash.com  Wed Jan 10 09:34:34 2007
From: dd at davedash.com (Dave Dash)
Date: Wed, 10 Jan 2007 11:34:34 -0600
Subject: No XML parser found w/ PHP 5.2.0
In-Reply-To: <20070110172909.GG5268@janrain.com>
References: <694fa1450701092003h5dfa74cag551b22a7e3631927@mail.gmail.com>
	<20070110172909.GG5268@janrain.com>
Message-ID: <694fa1450701100934y6568a858la2afd25efd65b28f@mail.gmail.com>

1.2.1

On 1/10/07, Jonathan Daugherty <cygnus at janrain.com> wrote:
>
> # Fatal error: No XML parser was found in
> # /usr/local/php5/lib/php/Services/Yadis/XML.php on line 352
> #
> # I'm confused as to what I need to install in order for it to work.
> # AFAIK PHP5 comes with DOM ... actually I'm pretty sure of it since
> # phpinfo() says so.
>
> Which version of the OpenID library are you using?
>
> --
>   Jonathan Daugherty
>   JanRain, Inc.
>
> _______________________________________________
> Dev mailing list
> Dev at lists.openidenabled.com
> http://lists.openidenabled.com/mailman/listinfo/dev
>



-- 
Dave Dash
612.670.0621
Discover your favorite restaurant: reviewsby.us
gtalk: dave.dash
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.openidenabled.com/pipermail/dev/attachments/20070110/2322a862/attachment.htm 

From cygnus at janrain.com  Wed Jan 10 09:48:54 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Wed, 10 Jan 2007 09:48:54 -0800
Subject: No XML parser found w/ PHP 5.2.0
In-Reply-To: <694fa1450701100934y6568a858la2afd25efd65b28f@mail.gmail.com>
References: <694fa1450701092003h5dfa74cag551b22a7e3631927@mail.gmail.com>
	<20070110172909.GG5268@janrain.com>
	<694fa1450701100934y6568a858la2afd25efd65b28f@mail.gmail.com>
Message-ID: <20070110174853.GI5268@janrain.com>

So you have something like this in phpinfo() (or "php -i")?

  dom

  DOM/XML => enabled
  DOM/XML API Version => 20031129

-- 
  Jonathan Daugherty
  JanRain, Inc.


From dd at davedash.com  Wed Jan 10 10:07:14 2007
From: dd at davedash.com (Dave Dash)
Date: Wed, 10 Jan 2007 12:07:14 -0600
Subject: No XML parser found w/ PHP 5.2.0
In-Reply-To: <20070110174853.GI5268@janrain.com>
References: <694fa1450701092003h5dfa74cag551b22a7e3631927@mail.gmail.com>
	<20070110172909.GG5268@janrain.com>
	<694fa1450701100934y6568a858la2afd25efd65b28f@mail.gmail.com>
	<20070110174853.GI5268@janrain.com>
Message-ID: <694fa1450701101007y12cc6d41r2e656b4ffb2f1cbd@mail.gmail.com>

I have the following:dom DOM/XML enabled DOM/XML API Version 20031129 libxml
Version 2.6.16 HTML Support enabled XPath Support enabled XPointer
Support enabled
Schema Support enabled RelaxNG Support enabled


On 1/10/07, Jonathan Daugherty <cygnus at janrain.com> wrote:
>
> So you have something like this in phpinfo() (or "php -i")?
>
>   dom
>
>   DOM/XML => enabled
>   DOM/XML API Version => 20031129
>
> --
>   Jonathan Daugherty
>   JanRain, Inc.
>
> _______________________________________________
> Dev mailing list
> Dev at lists.openidenabled.com
> http://lists.openidenabled.com/mailman/listinfo/dev
>



-- 
Dave Dash
612.670.0621
Discover your favorite restaurant: reviewsby.us
gtalk: dave.dash
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.openidenabled.com/pipermail/dev/attachments/20070110/13b70ef1/attachment.html 

From hellfrozeover at gmail.com  Mon Jan 15 16:48:44 2007
From: hellfrozeover at gmail.com (Tom H)
Date: Tue, 16 Jan 2007 00:48:44 +0000
Subject: win32 non pear path bug
Message-ID: <970066760701151648n44cf95ddqe5b44bfde73a9815@mail.gmail.com>

The following error occurs under win32 when
examples/consumer/try_auth.php is run using Auth_OpenID directly
(copying Auth folder into the consumer folder, as suggested in the
readme):

Warning: main(Auth/OpenID.php): failed to open stream: No such file or
directory in \path\toAuth\OpenID\Consumer.php on line 167

This is because the code at the top of examples/consumer/common.php,
which sets the include path hard codes ":" as the path seperator, but
win32 uses ";" as the path seperator.

The PHP constant PATH_SEPARATOR can be used instead:

--- common_old.php	2007-01-16 00:34:26.213269400 +0000
+++ common.php	2007-01-16 00:34:32.947644400 +0000
@@ -2,7 +2,7 @@

 $path_extra = dirname(dirname(dirname(__FILE__)));
 $path = ini_get('include_path');
-$path = $path_extra . ":" . $path;
+$path = $path_extra . PATH_SEPARATOR . $path;
 ini_set('include_path', $path);

 /**

thanks,

Tommeh


From cygnus at janrain.com  Mon Jan 15 16:52:05 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Mon, 15 Jan 2007 16:52:05 -0800
Subject: win32 non pear path bug
In-Reply-To: <970066760701151648n44cf95ddqe5b44bfde73a9815@mail.gmail.com>
References: <970066760701151648n44cf95ddqe5b44bfde73a9815@mail.gmail.com>
Message-ID: <20070116005204.GZ5268@janrain.com>

# This is because the code at the top of examples/consumer/common.php,
# which sets the include path hard codes ":" as the path seperator,
# but win32 uses ";" as the path seperator.
# 
# The PHP constant PATH_SEPARATOR can be used instead:

Thanks!  Applied.

-- 
  Jonathan Daugherty
  JanRain, Inc.
  irc.freenode.net: cygnus in #openid
  cygnus.myopenid.com


From swizard at gmail.com  Tue Jan 16 20:01:50 2007
From: swizard at gmail.com (sangho shin)
Date: Wed, 17 Jan 2007 13:01:50 +0900
Subject: immediate deny bug in ruby-openid
Message-ID: <451b240f0701162001mb4b527ay1c5d9ba5a37fc627@mail.gmail.com>

hi.
I think current implementation of ruby-openid lib has a bug in
immediate deny answer

if request is immediate , service provider should call resp =
req.answer(false, server_url)
in this case, current server.rb implementation set user_setup_url
'checkid_immediate'.

I think it should be set 'checkid_setup' instead of 'checkid_immediate
so that user or consumer can redirect to user_setup_url


i just copied source code of server.rb and Server.php. it seems to act
differently


from server.rb

      def answer(allow, server_url=nil)
        if allow or @immediate
          mode = 'id_res'
        else
          mode = 'cancel'
        end

        response = OpenIDResponse.new(self)

        if allow
          response.add_fields(nil, {
                                'mode' => mode,
                                'identity' => @identity,
                                'return_to' => @return_to
                              })
        else
          response.add_field(nil, 'mode', mode, false)
          response.signed.clear
          if @immediate
            unless server_url
              raise ArgumentError, "setup_url is required for
allow=false in immediate mode"
            end
            # make a request just like this one, but immediate mode
            setup_request = self.class.new('checkid_immediate',
                                           @identity,
                                           @return_to,
                                           @trust_root)
            setup_url = setup_request.encode_to_url(server_url)
            response.add_field(nil, 'user_setup_url', setup_url, false)
          end

        end

        return response
      end


same part of php lib implementation

    function answer($allow, $server_url = null)
    {
        if ($allow || $this->immediate) {
            $mode = 'id_res';
        } else {
            $mode = 'cancel';
        }

        $response = new Auth_OpenID_CheckIDResponse($this, $mode);

        if ($allow) {
            $response->fields['identity'] = $this->identity;
            $response->fields['return_to'] = $this->return_to;
            if (!$this->trustRootValid()) {
                return new Auth_OpenID_UntrustedReturnURL($this->return_to,
                                                          $this->trust_root);
            }
        } else {
            $response->signed = array();
            if ($this->immediate) {
                if (!$server_url) {
                    return new Auth_OpenID_ServerError(null,
                                 'setup_url is required for $allow=false \
                                  in immediate mode.');
                }

                $setup_request =& new Auth_OpenID_CheckIDRequest(
                                                $this->identity,
                                                $this->return_to,
                                                $this->trust_root,
                                                false,  ----------->
this part will set checkid_setup
                                                $this->assoc_handle);

                $setup_url = $setup_request->encodeToURL($server_url);

                $response->fields['user_setup_url'] = $setup_url;
            }
        }

        return $response;
    }


From dev-list-openidenabled at thequod.de  Wed Jan 17 13:21:27 2007
From: dev-list-openidenabled at thequod.de (dAniel hAhler)
Date: Wed, 17 Jan 2007 22:21:27 +0100
Subject: PHP OpenID lib: use of globals
In-Reply-To: <20070116004339.GY5268@janrain.com>
References: <200701160130.29545.b2evo@thequod.de>
	<20070116004339.GY5268@janrain.com>
Message-ID: <200701172221.33660.dev-list-openidenabled@thequod.de>

On Tuesday 16 January 2007 01:43, Jonathan wrote:

> # I require the classes out of a class method (for performance reasons
> # only if necessary) and therefor a "global $foo" statement in your
> # class methods does not work. I would have to use "global $foo" in
> # the method where I'm requiring your classes, too.
>
> This is a known problem with the current release, and the fix is
> merely to include our library files at the file scope, rather than
> doing it inside a method.

I know, but I want to require the files only if the Plugin gets used.

I could have worked around this, of course, but I think a lib like this should 
not depend on such "restrictions".

I've attached a patch against 1.2.1.

I've found another todo, but I'll write to the dev list for this.

Unfortunately I've discovered the test framework too late. Now I have some 
lines > 80 chars and it does not seem to work with PHPUnit 3.x.

I get:
$ ./PHP-openid-1.2.1-modified/admin/runtests
Running test tabs
Running test longlines
Found lines > 80 characters in:

Auth/OpenID/URINorm.php (line 112 exceeds length 80)
Auth/OpenID/Server.php (lines 176, 485, 742, 763, 1184, and others exceed 
length 80)
Auth/OpenID/DiffieHellman.php (line 35 exceeds length 80)
Auth/OpenID/BigMath.php (line 423 exceeds length 80)
Running test nobadbraces
Running test nobadcase
Running test opentag
Running test docblocks
Running test php

Warning: require_once(PHPUnit.php): failed to open stream: No such file or 
directory 
in /home/daniel/dev/b2evo.HEAD/blogs/plugins/openid_plugin/PHP-openid-1.2.1-modified/Tests/TestDriver.php 
on line 16

Call Stack:
    0.0023     124776   1. {main}
() /home/daniel/dev/b2evo.HEAD/blogs/plugins/openid_plugin/PHP-openid-1.2.1-modified/admin/texttest.php:0
    0.0044     172136   2. 
require_once('/home/daniel/dev/b2evo.HEAD/blogs/plugins/openid_plugin/PHP-openid-1.2.1-modified/Tests/TestDriver.php') /home/daniel/dev/b2evo.HEAD/blogs/plugins/openid_plugin/PHP-openid-1.2.1-modified/admin/texttest.php:3


Fatal error: require_once(): Failed opening required 'PHPUnit.php' 
(include_path='.:/usr/local/lib/php:/usr/share/php') 
in /home/daniel/dev/b2evo.HEAD/blogs/plugins/openid_plugin/PHP-openid-1.2.1-modified/Tests/TestDriver.php 
on line 16

Call Stack:
    0.0023     124776   1. {main}
() /home/daniel/dev/b2evo.HEAD/blogs/plugins/openid_plugin/PHP-openid-1.2.1-modified/admin/texttest.php:0
    0.0044     172136   2. 
require_once('/home/daniel/dev/b2evo.HEAD/blogs/plugins/openid_plugin/PHP-openid-1.2.1-modified/Tests/TestDriver.php') /home/daniel/dev/b2evo.HEAD/blogs/plugins/openid_plugin/PHP-openid-1.2.1-modified/admin/texttest.php:3

Running test import
Failures in:  longlines php

I hope the patch gets applied (and cleaned up) though.

Thank you.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PHP-openid-remove-global-statements-in-lib-files.patch
Type: text/x-diff
Size: 27724 bytes
Desc: not available
Url : http://lists.openidenabled.com/pipermail/dev/attachments/20070117/b9af951e/attachment-0001.bin 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://lists.openidenabled.com/pipermail/dev/attachments/20070117/b9af951e/attachment-0001.pgp 

From brian at janrain.com  Wed Jan 17 13:29:31 2007
From: brian at janrain.com (Brian Ellin)
Date: Wed, 17 Jan 2007 13:29:31 -0800
Subject: immediate deny bug in ruby-openid
In-Reply-To: <451b240f0701162001mb4b527ay1c5d9ba5a37fc627@mail.gmail.com>
References: <451b240f0701162001mb4b527ay1c5d9ba5a37fc627@mail.gmail.com>
Message-ID: <9674d5350701171329n7a971f92p22ff47cd767b123@mail.gmail.com>

Sangho,

You are correct.  I'll get this fixed for the next release.

Also, are you writing a server using the ruby lib, or did you discover
this behaviour testing against an existing server?

Thanks,
Brian Ellin
brian at janrain.com



On 1/16/07, sangho shin <swizard at gmail.com> wrote:
> hi.
> I think current implementation of ruby-openid lib has a bug in
> immediate deny answer
>
> if request is immediate , service provider should call resp =
> req.answer(false, server_url)
> in this case, current server.rb implementation set user_setup_url
> 'checkid_immediate'.
>
> I think it should be set 'checkid_setup' instead of 'checkid_immediate
> so that user or consumer can redirect to user_setup_url
>
>
> i just copied source code of server.rb and Server.php. it seems to act
> differently
>
>
> from server.rb
>
>       def answer(allow, server_url=nil)
>         if allow or @immediate
>           mode = 'id_res'
>         else
>           mode = 'cancel'
>         end
>
>         response = OpenIDResponse.new(self)
>
>         if allow
>           response.add_fields(nil, {
>                                 'mode' => mode,
>                                 'identity' => @identity,
>                                 'return_to' => @return_to
>                               })
>         else
>           response.add_field(nil, 'mode', mode, false)
>           response.signed.clear
>           if @immediate
>             unless server_url
>               raise ArgumentError, "setup_url is required for
> allow=false in immediate mode"
>             end
>             # make a request just like this one, but immediate mode
>             setup_request = self.class.new('checkid_immediate',
>                                            @identity,
>                                            @return_to,
>                                            @trust_root)
>             setup_url = setup_request.encode_to_url(server_url)
>             response.add_field(nil, 'user_setup_url', setup_url, false)
>           end
>
>         end
>
>         return response
>       end
>
>
> same part of php lib implementation
>
>     function answer($allow, $server_url = null)
>     {
>         if ($allow || $this->immediate) {
>             $mode = 'id_res';
>         } else {
>             $mode = 'cancel';
>         }
>
>         $response = new Auth_OpenID_CheckIDResponse($this, $mode);
>
>         if ($allow) {
>             $response->fields['identity'] = $this->identity;
>             $response->fields['return_to'] = $this->return_to;
>             if (!$this->trustRootValid()) {
>                 return new Auth_OpenID_UntrustedReturnURL($this->return_to,
>                                                           $this->trust_root);
>             }
>         } else {
>             $response->signed = array();
>             if ($this->immediate) {
>                 if (!$server_url) {
>                     return new Auth_OpenID_ServerError(null,
>                                  'setup_url is required for $allow=false \
>                                   in immediate mode.');
>                 }
>
>                 $setup_request =& new Auth_OpenID_CheckIDRequest(
>                                                 $this->identity,
>                                                 $this->return_to,
>                                                 $this->trust_root,
>                                                 false,  ----------->
> this part will set checkid_setup
>                                                 $this->assoc_handle);
>
>                 $setup_url = $setup_request->encodeToURL($server_url);
>
>                 $response->fields['user_setup_url'] = $setup_url;
>             }
>         }
>
>         return $response;
>     }
>
> _______________________________________________
> Dev mailing list
> Dev at lists.openidenabled.com
> http://lists.openidenabled.com/mailman/listinfo/dev
>


From dev-list-openidenabled at thequod.de  Wed Jan 17 14:02:44 2007
From: dev-list-openidenabled at thequod.de (dAniel hAhler)
Date: Wed, 17 Jan 2007 23:02:44 +0100
Subject: [PHP-openid] PCRE backtrack error in
	Auth_OpenID_Parse::parseLinkAttrs() / todo (patch)
Message-ID: <200701172302.45199.dev-list-openidenabled@thequod.de>

Hi,

I find that the Auth_OpenID_Parse::parseLinkAttrs() method is quite 
ineffective.

It failed here (silently), because of a PREG_BACKTRACK_LIMIT_ERROR in 
preg_match() (introduced with PHP 5.2).

The attached patch adds a TODO with an idea of a better/more efficient 
implementation and a dirty workaround, by disabling this limit.

It seems like the parsing fails if the HTML to get parsed is > 100KB or 
something similar. The default backtrack limit is 100000.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PHP-openid-pcre-backtrack-dirty-workaround.patch
Type: text/x-diff
Size: 1910 bytes
Desc: not available
Url : http://lists.openidenabled.com/pipermail/dev/attachments/20070117/14e9fc7d/attachment.bin 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://lists.openidenabled.com/pipermail/dev/attachments/20070117/14e9fc7d/attachment.pgp 

From swizard at gmail.com  Wed Jan 17 23:51:00 2007
From: swizard at gmail.com (sangho shin)
Date: Thu, 18 Jan 2007 16:51:00 +0900
Subject: immediate deny bug in ruby-openid
In-Reply-To: <9674d5350701171329n7a971f92p22ff47cd767b123@mail.gmail.com>
References: <451b240f0701162001mb4b527ay1c5d9ba5a37fc627@mail.gmail.com>
	<9674d5350701171329n7a971f92p22ff47cd767b123@mail.gmail.com>
Message-ID: <451b240f0701172351m59f9b253k7d79bb5727bebe8e@mail.gmail.com>

I just made http://www.myid.net openid service provider using ruby lib.
It's for korean :)

I didn't know about the bug, until I test with word-press openid plugin
which use immediate mode and try to redirect user_setup_url.


On 1/18/07, Brian Ellin <brian at janrain.com> wrote:
>
> Sangho,
>
> You are correct.  I'll get this fixed for the next release.
>
> Also, are you writing a server using the ruby lib, or did you discover
> this behaviour testing against an existing server?
>
> Thanks,
> Brian Ellin
> brian at janrain.com
>
>
>
> On 1/16/07, sangho shin <swizard at gmail.com> wrote:
> > hi.
> > I think current implementation of ruby-openid lib has a bug in
> > immediate deny answer
> >
> > if request is immediate , service provider should call resp =
> > req.answer(false, server_url)
> > in this case, current server.rb implementation set user_setup_url
> > 'checkid_immediate'.
> >
> > I think it should be set 'checkid_setup' instead of 'checkid_immediate
> > so that user or consumer can redirect to user_setup_url
> >
> >
> > i just copied source code of server.rb and Server.php. it seems to act
> > differently
> >
> >
> > from server.rb
> >
> >       def answer(allow, server_url=nil)
> >         if allow or @immediate
> >           mode = 'id_res'
> >         else
> >           mode = 'cancel'
> >         end
> >
> >         response = OpenIDResponse.new(self)
> >
> >         if allow
> >           response.add_fields(nil, {
> >                                 'mode' => mode,
> >                                 'identity' => @identity,
> >                                 'return_to' => @return_to
> >                               })
> >         else
> >           response.add_field(nil, 'mode', mode, false)
> >           response.signed.clear
> >           if @immediate
> >             unless server_url
> >               raise ArgumentError, "setup_url is required for
> > allow=false in immediate mode"
> >             end
> >             # make a request just like this one, but immediate mode
> >             setup_request = self.class.new('checkid_immediate',
> >                                            @identity,
> >                                            @return_to,
> >                                            @trust_root)
> >             setup_url = setup_request.encode_to_url(server_url)
> >             response.add_field(nil, 'user_setup_url', setup_url, false)
> >           end
> >
> >         end
> >
> >         return response
> >       end
> >
> >
> > same part of php lib implementation
> >
> >     function answer($allow, $server_url = null)
> >     {
> >         if ($allow || $this->immediate) {
> >             $mode = 'id_res';
> >         } else {
> >             $mode = 'cancel';
> >         }
> >
> >         $response = new Auth_OpenID_CheckIDResponse($this, $mode);
> >
> >         if ($allow) {
> >             $response->fields['identity'] = $this->identity;
> >             $response->fields['return_to'] = $this->return_to;
> >             if (!$this->trustRootValid()) {
> >                 return new
> Auth_OpenID_UntrustedReturnURL($this->return_to,
> >
> $this->trust_root);
> >             }
> >         } else {
> >             $response->signed = array();
> >             if ($this->immediate) {
> >                 if (!$server_url) {
> >                     return new Auth_OpenID_ServerError(null,
> >                                  'setup_url is required for $allow=false
> \
> >                                   in immediate mode.');
> >                 }
> >
> >                 $setup_request =& new Auth_OpenID_CheckIDRequest(
> >                                                 $this->identity,
> >                                                 $this->return_to,
> >                                                 $this->trust_root,
> >                                                 false,  ----------->
> > this part will set checkid_setup
> >                                                 $this->assoc_handle);
> >
> >                 $setup_url = $setup_request->encodeToURL($server_url);
> >
> >                 $response->fields['user_setup_url'] = $setup_url;
> >             }
> >         }
> >
> >         return $response;
> >     }
> >
> > _______________________________________________
> > Dev mailing list
> > Dev at lists.openidenabled.com
> > http://lists.openidenabled.com/mailman/listinfo/dev
> >
>
> _______________________________________________
> Dev mailing list
> Dev at lists.openidenabled.com
> http://lists.openidenabled.com/mailman/listinfo/dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.openidenabled.com/pipermail/dev/attachments/20070118/8267b0e7/attachment.htm 

From kevin at radagast.biz  Thu Jan 18 08:18:13 2007
From: kevin at radagast.biz (Kevin Jardine)
Date: Thu, 18 Jan 2007 17:18:13 +0100
Subject: Missing $_SESSION variable
Message-ID: <45AF9DC5.1090506@radagast.biz>

I have a Drupal OpenID client which works fine in a PHP 4.3 environment.

I recently moved the code to a PHP 5.1.6 environment and it stopped 
working. In particular, the _openid_consumer_last_token variable that 
was in the $_SESSION array in handle_login was no longer there in 
handle_return and the process fails with a "No session state found" error.

Has anyone else seen this problem? Is there a PHP configuration option I 
need to tweak to get $_SESSION variables to work properly?

I am using Drupal 4.7.3 in both cases if that helps.

Cheers,
Kevin


-- 
Kevin Jardine
Radagast Solutions

Internet campaign advice and magic
http://radagast.biz

YIM: kevinjardine
Skype: kevinjardine
Eml: kevin at radagast.biz
Tel: +31 (0)6 25581608


From cygnus at janrain.com  Thu Jan 18 08:32:17 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Thu, 18 Jan 2007 08:32:17 -0800
Subject: Missing $_SESSION variable
In-Reply-To: <45AF9DC5.1090506@radagast.biz>
References: <45AF9DC5.1090506@radagast.biz>
Message-ID: <20070118163217.GM5268@janrain.com>

# I recently moved the code to a PHP 5.1.6 environment and it stopped
# working. In particular, the _openid_consumer_last_token variable
# that was in the $_SESSION array in handle_login was no longer there
# in handle_return and the process fails with a "No session state
# found" error.

Have you verified that your browser is sending the same cookie back to
drupal on the return redirect?  (Usually PHPSESSID.)  (Which
browser(s) is/are being tested with this behavior?)

Thanks,

-- 
  Jonathan Daugherty
  JanRain, Inc.
  irc.freenode.net: cygnus in #openid
  cygnus.myopenid.com


From kevin at radagast.biz  Thu Jan 18 08:46:05 2007
From: kevin at radagast.biz (Kevin Jardine)
Date: Thu, 18 Jan 2007 17:46:05 +0100
Subject: Missing $_SESSION variable
In-Reply-To: <20070118163217.GM5268@janrain.com>
References: <45AF9DC5.1090506@radagast.biz> <20070118163217.GM5268@janrain.com>
Message-ID: <45AFA44D.3040007@radagast.biz>

I've used the same browser in all cases - Firefox 2.0.0.1

What I've done is check the $_SESSION variable.

The value is there in handle_login and not there in handle_return.

I do notice one difference between the two installations is that 
session.bug_compat_42 is set to "1" on the installation that works and 
"0" on the installation that does not work.

Could that make a difference?

Cheers,
Kevin

Jonathan Daugherty wrote:
> # I recently moved the code to a PHP 5.1.6 environment and it stopped
> # working. In particular, the _openid_consumer_last_token variable
> # that was in the $_SESSION array in handle_login was no longer there
> # in handle_return and the process fails with a "No session state
> # found" error.
> 
> Have you verified that your browser is sending the same cookie back to
> drupal on the return redirect?  (Usually PHPSESSID.)  (Which
> browser(s) is/are being tested with this behavior?)
> 
> Thanks,
> 


-- 
Kevin Jardine
Radagast Solutions

Internet campaign advice and magic
http://radagast.biz

YIM: kevinjardine
Skype: kevinjardine
Eml: kevin at radagast.biz
Tel: +31 (0)6 25581608


From kevin at radagast.biz  Thu Jan 18 09:03:09 2007
From: kevin at radagast.biz (Kevin Jardine)
Date: Thu, 18 Jan 2007 18:03:09 +0100
Subject: Missing $_SESSION variable
In-Reply-To: <45AFA44D.3040007@radagast.biz>
References: <45AF9DC5.1090506@radagast.biz> <20070118163217.GM5268@janrain.com>
	<45AFA44D.3040007@radagast.biz>
Message-ID: <45AFA84D.3070402@radagast.biz>

I checked the COOKIE value.

It contains

[PHPSESSID] => qivpm9isnum41rdi2l0gti09p4

in both cases.

Cheers,
Kevin

Kevin Jardine wrote:
> I've used the same browser in all cases - Firefox 2.0.0.1
> 
> What I've done is check the $_SESSION variable.
> 
> The value is there in handle_login and not there in handle_return.
> 
> I do notice one difference between the two installations is that 
> session.bug_compat_42 is set to "1" on the installation that works and 
> "0" on the installation that does not work.
> 
> Could that make a difference?
> 
> Cheers,
> Kevin
> 
> Jonathan Daugherty wrote:
>> # I recently moved the code to a PHP 5.1.6 environment and it stopped
>> # working. In particular, the _openid_consumer_last_token variable
>> # that was in the $_SESSION array in handle_login was no longer there
>> # in handle_return and the process fails with a "No session state
>> # found" error.
>>
>> Have you verified that your browser is sending the same cookie back to
>> drupal on the return redirect?  (Usually PHPSESSID.)  (Which
>> browser(s) is/are being tested with this behavior?)
>>
>> Thanks,
>>
> 
> 


-- 
Kevin Jardine
Radagast Solutions

Internet campaign advice and magic
http://radagast.biz

YIM: kevinjardine
Skype: kevinjardine
Eml: kevin at radagast.biz
Tel: +31 (0)6 25581608


From kevin at radagast.biz  Thu Jan 18 09:19:14 2007
From: kevin at radagast.biz (Kevin Jardine)
Date: Thu, 18 Jan 2007 18:19:14 +0100
Subject: Missing $_SESSION variable
In-Reply-To: <45AFA84D.3070402@radagast.biz>
References: <45AF9DC5.1090506@radagast.biz>
	<20070118163217.GM5268@janrain.com>	<45AFA44D.3040007@radagast.biz>
	<45AFA84D.3070402@radagast.biz>
Message-ID: <45AFAC12.4030309@radagast.biz>

When I look in the Drupal sessions table I see:

a sid of qivpm9isnum41rdi2l0gti09p4

and a session value of

_yadis_services__openid_consumer_|s:543:"O:22:"Services_Yadis_Manager":6:{s:12:"starting_url";s:32:"http://kevin.elgg.org/demo/news/";s:9:"yadis_url";s:32:"http://kevin.elgg.org/demo/news/";s:8:"services";a:0:{}s:11:"session_key";s:33:"_yadis_services__openid_consumer_";s:8:"_current";O:27:"Auth_OpenID_ServiceEndpoint":5:{s:12:"identity_url";s:32:"http://kevin.elgg.org/demo/news/";s:10:"server_url";s:55:"http://kevin.elgg.org/demo/mod/openid_server/server.php";s:9:"type_uris";a:1:{i:0;s:28:"http://openid.net/signon/1.0";}s:8:"delegate";N;s:10:"used_yadis";b:0;}s:5:"stale";b:0;}";

If that helps.

Kevin Jardine wrote:
> I checked the COOKIE value.
> 
> It contains
> 
> [PHPSESSID] => qivpm9isnum41rdi2l0gti09p4
> 
> in both cases.
> 
> Cheers,
> Kevin
> 
> Kevin Jardine wrote:
>> I've used the same browser in all cases - Firefox 2.0.0.1
>>
>> What I've done is check the $_SESSION variable.
>>
>> The value is there in handle_login and not there in handle_return.
>>
>> I do notice one difference between the two installations is that 
>> session.bug_compat_42 is set to "1" on the installation that works and 
>> "0" on the installation that does not work.
>>
>> Could that make a difference?
>>
>> Cheers,
>> Kevin
>>
>> Jonathan Daugherty wrote:
>>> # I recently moved the code to a PHP 5.1.6 environment and it stopped
>>> # working. In particular, the _openid_consumer_last_token variable
>>> # that was in the $_SESSION array in handle_login was no longer there
>>> # in handle_return and the process fails with a "No session state
>>> # found" error.
>>>
>>> Have you verified that your browser is sending the same cookie back to
>>> drupal on the return redirect?  (Usually PHPSESSID.)  (Which
>>> browser(s) is/are being tested with this behavior?)
>>>
>>> Thanks,
>>>
>>
> 
> 


-- 
Kevin Jardine
Radagast Solutions

Internet campaign advice and magic
http://radagast.biz

YIM: kevinjardine
Skype: kevinjardine
Eml: kevin at radagast.biz
Tel: +31 (0)6 25581608


From kevin at radagast.biz  Thu Jan 18 14:03:41 2007
From: kevin at radagast.biz (Kevin Jardine)
Date: Thu, 18 Jan 2007 23:03:41 +0100
Subject: Missing $_SESSION variable
In-Reply-To: <45AFAC12.4030309@radagast.biz>
References: <45AF9DC5.1090506@radagast.biz>	<20070118163217.GM5268@janrain.com>	<45AFA44D.3040007@radagast.biz>	<45AFA84D.3070402@radagast.biz>
	<45AFAC12.4030309@radagast.biz>
Message-ID: <45AFEEBD.3000206@radagast.biz>

I set session.bug_compat_42 to "1" and then discovered that array_merge 
no longer works as expected with NULL values under PHP 5. I went through 
my code and made sure that the parameters used with array_merge are 
never null values and always arrays.

That seems to have fixed the problem.

I'm not sure if the JanRain code assumes that session.bug_compat_42 is 
set to "1". Before it was set to "1", if appeared that the $_SESSION 
values were not being saved properly. Might be worth checking.

Cheers,
Kevin

Kevin Jardine wrote:
> When I look in the Drupal sessions table I see:
> 
> a sid of qivpm9isnum41rdi2l0gti09p4
> 
> and a session value of
> 
> _yadis_services__openid_consumer_|s:543:"O:22:"Services_Yadis_Manager":6:{s:12:"starting_url";s:32:"http://kevin.elgg.org/demo/news/";s:9:"yadis_url";s:32:"http://kevin.elgg.org/demo/news/";s:8:"services";a:0:{}s:11:"session_key";s:33:"_yadis_services__openid_consumer_";s:8:"_current";O:27:"Auth_OpenID_ServiceEndpoint":5:{s:12:"identity_url";s:32:"http://kevin.elgg.org/demo/news/";s:10:"server_url";s:55:"http://kevin.elgg.org/demo/mod/openid_server/server.php";s:9:"type_uris";a:1:{i:0;s:28:"http://openid.net/signon/1.0";}s:8:"delegate";N;s:10:"used_yadis";b:0;}s:5:"stale";b:0;}";
> 
> If that helps.
> 
> Kevin Jardine wrote:
>> I checked the COOKIE value.
>>
>> It contains
>>
>> [PHPSESSID] => qivpm9isnum41rdi2l0gti09p4
>>
>> in both cases.
>>
>> Cheers,
>> Kevin
>>
>> Kevin Jardine wrote:
>>> I've used the same browser in all cases - Firefox 2.0.0.1
>>>
>>> What I've done is check the $_SESSION variable.
>>>
>>> The value is there in handle_login and not there in handle_return.
>>>
>>> I do notice one difference between the two installations is that 
>>> session.bug_compat_42 is set to "1" on the installation that works and 
>>> "0" on the installation that does not work.
>>>
>>> Could that make a difference?
>>>
>>> Cheers,
>>> Kevin
>>>
>>> Jonathan Daugherty wrote:
>>>> # I recently moved the code to a PHP 5.1.6 environment and it stopped
>>>> # working. In particular, the _openid_consumer_last_token variable
>>>> # that was in the $_SESSION array in handle_login was no longer there
>>>> # in handle_return and the process fails with a "No session state
>>>> # found" error.
>>>>
>>>> Have you verified that your browser is sending the same cookie back to
>>>> drupal on the return redirect?  (Usually PHPSESSID.)  (Which
>>>> browser(s) is/are being tested with this behavior?)
>>>>
>>>> Thanks,
>>>>
>>
> 
> 


-- 
Kevin Jardine
Radagast Solutions

Internet campaign advice and magic
http://radagast.biz

YIM: kevinjardine
Skype: kevinjardine
Eml: kevin at radagast.biz
Tel: +31 (0)6 25581608


From daniel-hofstetter at gmx.ch  Fri Jan 19 02:32:25 2007
From: daniel-hofstetter at gmx.ch (Daniel Hofstetter)
Date: Fri, 19 Jan 2007 11:32:25 +0100
Subject: Bugfix for minor bug in Services_Yadis_PHPSession
Message-ID: <45B09E39.3020406@gmx.ch>

Hi all,

If I use the PHP OpenID library (and with it Yadis), I get the following
error when I run my tests:

Unexpected PHP error [array_key_exists(): The second argument should be
either an array or an object] severity [E_WARNING] in
[/home/dho/projects/cake_1.2.x.x/vendors/Services/Yadis/Manager.php line 40]

This error occurs because $_SESSION is empty. So I added a check for
emptiness to the "get" function:

function get($name, $default=null)
{
    if (!empty($_SESSION) && array_key_exists($name, $_SESSION)) {
        return $_SESSION[$name];
    } else {
        return $default;
    }
}

-- 
Daniel Hofstetter
http://cakebaker.42dh.com


From dev-list-openidenabled at thequod.de  Wed Jan 24 15:23:06 2007
From: dev-list-openidenabled at thequod.de (dAniel hAhler)
Date: Thu, 25 Jan 2007 00:23:06 +0100
Subject: [PHP OpenID]: restore AUTOCOMMIT in SQLStore
Message-ID: <bcadbdf50701241523g15742902qe7adddd869a45f43@mail.gmail.com>

Hi,

in the SQLStore constructor, the following gets done:
        // Be sure to run the database queries with auto-commit mode
        // turned OFF, because we want every function to run in a
        // transaction, implicitly.  As a rule, methods named with a
        // leading underscore will NOT control transaction behavior.
        // Callers of these methods will worry about transactions.
        $this->connection->autoCommit(false);

Unfortunately, this is likely to conflict with passing in an already
established (and globally used) $connection: You would have to
manually turn AUTOCOMMIT on again - probably after
$consumer->complete() and $consumer->begin() etc.

Wouldn't it make more sense to _not_ set AUTOCOMMIT=0 and use BEGIN()
instead explicitly?!

I hope this makes sense?!


From cygnus at janrain.com  Wed Jan 24 15:36:02 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Wed, 24 Jan 2007 15:36:02 -0800
Subject: [PHP OpenID]: restore AUTOCOMMIT in SQLStore
In-Reply-To: <bcadbdf50701241523g15742902qe7adddd869a45f43@mail.gmail.com>
References: <bcadbdf50701241523g15742902qe7adddd869a45f43@mail.gmail.com>
Message-ID: <20070124233601.GL15812@janrain.com>

# Unfortunately, this is likely to conflict with passing in an already
# established (and globally used) $connection: You would have to
# manually turn AUTOCOMMIT on again - probably after
# $consumer->complete() and $consumer->begin() etc.

It's true that we wrote that code assuming we'd get our very own db
connection.  The safest thing to do would be to store the original
autocommit state and reset it after we're done, although I don't see a
way in PEAR DB to get the current autocommit state.  (If you know of
it, please submit a patch.  I'm not in the mood to print_r($conn) to
find out where it lives, if it lives at all.)

# Wouldn't it make more sense to _not_ set AUTOCOMMIT=0 and use
# BEGIN() instead explicitly?!

Well, there is no "begin()" method of the DB class, and the overall
design of DB would imply that issuing a BEGIN query manually is not a
good idea as the transaction state is managed for you internally.

In order to know that commit() and rollback() calls actually mean
anything, it's *necessary* to turn auto-commit mode off.

-- 
  Jonathan Daugherty
  JanRain, Inc.
  irc.freenode.net: cygnus in #openid
  cygnus.myopenid.com


From cygnus at janrain.com  Wed Jan 24 15:37:33 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Wed, 24 Jan 2007 15:37:33 -0800
Subject: Bugfix for minor bug in Services_Yadis_PHPSession
In-Reply-To: <45B09E39.3020406@gmx.ch>
References: <45B09E39.3020406@gmx.ch>
Message-ID: <20070124233732.GM15812@janrain.com>

# Unexpected PHP error [array_key_exists(): The second argument should be
# either an array or an object] severity [E_WARNING] in
# [/home/dho/projects/cake_1.2.x.x/vendors/Services/Yadis/Manager.php line 40]
# 
# This error occurs because $_SESSION is empty. So I added a check for
# emptiness to the "get" function:

It seems to me that this is because session_start() has not been
called.  That code is otherwise valid, because it assumes $_SESSION
exists and is at the very least == array(), in which case no empty()
check is necessary.  What do you think?

-- 
  Jonathan Daugherty
  JanRain, Inc.
  irc.freenode.net: cygnus in #openid
  cygnus.myopenid.com


From cygnus at janrain.com  Wed Jan 24 15:46:38 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Wed, 24 Jan 2007 15:46:38 -0800
Subject: Missing $_SESSION variable
In-Reply-To: <45AFEEBD.3000206@radagast.biz>
References: <45AF9DC5.1090506@radagast.biz> <20070118163217.GM5268@janrain.com>
	<45AFA44D.3040007@radagast.biz> <45AFA84D.3070402@radagast.biz>
	<45AFAC12.4030309@radagast.biz> <45AFEEBD.3000206@radagast.biz>
Message-ID: <20070124234638.GN15812@janrain.com>

# I set session.bug_compat_42 to "1" and then discovered that
# array_merge no longer works as expected with NULL values under PHP
# 5. I went through my code and made sure that the parameters used
# with array_merge are never null values and always arrays.
# 
# That seems to have fixed the problem.

Hi Kevin,

Thanks for letting me know about this.

Php.net has this rather cryptic bit to say about bug_compat_42:

  PHP versions 4.2.3 and lower have an undocumented feature/bug that
  allows you to initialize a session variable in the global scope,
  albeit register_globals is disabled. PHP 4.3.0 and later will warn
  you, if this feature is used, and if session.bug_compat_warn is also
  enabled. This feature/bug can be disabled by disabling this
  directive.

I don't know what "initialize a session variable in the global scope"
even means.  Do you?  (It's probably something from PHP's dark past
that I *don't* want to know is possible.)  I'm not sure how this
affects our libraries specifically.  Our code actually makes no
assumption one way or the other.  (Rather, it assumes sane behavior.)

I know that in my PHP configuration (both apache and cli),
bug_compat_42 is set to 1.  However, I'm not sure it has ever had any
bearing on what I expect my code to do.  (So, I would love some
clarification on the nature of this, if you can provide it.)

Thanks again,

-- 
  Jonathan Daugherty
  JanRain, Inc.
  irc.freenode.net: cygnus in #openid
  cygnus.myopenid.com


From dev-list-openidenabled at thequod.de  Wed Jan 24 16:00:49 2007
From: dev-list-openidenabled at thequod.de (dAniel hAhler)
Date: Thu, 25 Jan 2007 01:00:49 +0100
Subject: [PHP OpenID]: restore AUTOCOMMIT in SQLStore
In-Reply-To: <20070124233601.GL15812@janrain.com>
References: <bcadbdf50701241523g15742902qe7adddd869a45f43@mail.gmail.com>
	<20070124233601.GL15812@janrain.com>
Message-ID: <bcadbdf50701241600g139ce07ao1dbef70ecdd51e92@mail.gmail.com>

2007/1/25, Jonathan Daugherty <cygnus at janrain.com>:
> # Unfortunately, this is likely to conflict with passing in an already
> # established (and globally used) $connection: You would have to
> # manually turn AUTOCOMMIT on again - probably after
> # $consumer->complete() and $consumer->begin() etc.
>
> It's true that we wrote that code assuming we'd get our very own db
> connection.  The safest thing to do would be to store the original
> autocommit state and reset it after we're done, although I don't see a
> way in PEAR DB to get the current autocommit state.  (If you know of
> it, please submit a patch.  I'm not in the mood to print_r($conn) to
> find out where it lives, if it lives at all.)

According to http://pear.php.net/package/DB/download/1.7.0RC1 there's
a property autoCommit / autocommit, which would hold this value.

>From looking at PHP-OpenID uses (e.g. different plugins), most seem to
just not implement the autoCommit() method and therefor it will not
get changed at all. That's what I've done now, too.

>
> # Wouldn't it make more sense to _not_ set AUTOCOMMIT=0 and use
> # BEGIN() instead explicitly?!
>
> Well, there is no "begin()" method of the DB class, and the overall
> design of DB would imply that issuing a BEGIN query manually is not a
> good idea as the transaction state is managed for you internally.
>
> In order to know that commit() and rollback() calls actually mean
> anything, it's *necessary* to turn auto-commit mode off.

I see. I don't know the PEAR::DB class really. We're using an ezSQL
derived class in our project.


From cygnus at janrain.com  Wed Jan 24 16:03:35 2007
From: cygnus at janrain.com (Jonathan Daugherty)
Date: Wed, 24 Jan 2007 16:03:35 -0800
Subject: [PHP OpenID]: restore AUTOCOMMIT in SQLStore
In-Reply-To: <bcadbdf50701241600g139ce07ao1dbef70ecdd51e92@mail.gmail.com>
References: <bcadbdf50701241523g15742902qe7adddd869a45f43@mail.gmail.com>
	<20070124233601.GL15812@janrain.com>
	<bcadbdf50701241600g139ce07ao1dbef70ecdd51e92@mail.gmail.com>
Message-ID: <20070125000335.GO15812@janrain.com>

# According to http://pear.php.net/package/DB/download/1.7.0RC1
# there's a property autoCommit / autocommit, which would hold this
# value.

Do you think I can assume that 1.7.0RC1 is going to be widely
deployed? :)  (Thanks for the info, though.)

# From looking at PHP-OpenID uses (e.g. different plugins), most seem
# to just not implement the autoCommit() method and therefor it will
# not get changed at all. That's what I've done now, too.

That seems fine to me.

-- 
  Jonathan Daugherty
  JanRain, Inc.
  irc.freenode.net: cygnus in #openid
  cygnus.myopenid.com


From dev-list-openidenabled at thequod.de  Wed Jan 24 17:05:26 2007
From: dev-list-openidenabled at thequod.de (dAniel hAhler)
Date: Thu, 25 Jan 2007 02:05:26 +0100
Subject: [PHP OpenID]: restore AUTOCOMMIT in SQLStore
In-Reply-To: <20070125000335.GO15812@janrain.com>
References: <bcadbdf50701241523g15742902qe7adddd869a45f43@mail.gmail.com>
	<20070124233601.GL15812@janrain.com>
	<bcadbdf50701241600g139ce07ao1dbef70ecdd51e92@mail.gmail.com>
	<20070125000335.GO15812@janrain.com>
Message-ID: <bcadbdf50701241705ub3e799atd46b176a2dcea3a5@mail.gmail.com>

2007/1/25, Jonathan Daugherty <cygnus at janrain.com>:
> # According to http://pear.php.net/package/DB/download/1.7.0RC1
> # there's a property autoCommit / autocommit, which would hold this
> # value.
>
> Do you think I can assume that 1.7.0RC1 is going to be widely
> deployed? :)  (Thanks for the info, though.)

As far as I understood it's named autoCommit before 1.7.0RC1 and
autocommit afterwards (and you would have to test for both). But maybe
it got just introduced in some 1.7-beta though.


From kevin at radagast.biz  Wed Jan 24 23:58:50 2007
From: kevin at radagast.biz (Kevin Jardine)
Date: Thu, 25 Jan 2007 08:58:50 +0100
Subject: Missing $_SESSION variable
In-Reply-To: <20070124234638.GN15812@janrain.com>
References: <45AF9DC5.1090506@radagast.biz>
	<20070118163217.GM5268@janrain.com>	<45AFA44D.3040007@radagast.biz>
	<45AFA84D.3070402@radagast.biz>	<45AFAC12.4030309@radagast.biz>
	<45AFEEBD.3000206@radagast.biz>
	<20070124234638.GN15812@janrain.com>
Message-ID: <45B8633A.1020605@radagast.biz>

Hi Jonathan,

I have not had the chance to compare the behaviour with this flag turned 
off or on.

However, I did discover that a bug in my code related to a PHP 5 
incompatibility was causing the code to run after a perfectly correct 
del('_openid_consumer_last_token') function call in the JanRain library.

I believe that this was the reason that this variable was missing from 
$_SESSION.

So I think that this problem was related to a problem with my code 
rather than the JanRain library.

My apologies if I've wasted your time.

Cheers,
Kevin

Jonathan Daugherty wrote:
> # I set session.bug_compat_42 to "1" and then discovered that
> # array_merge no longer works as expected with NULL values under PHP
> # 5. I went through my code and made sure that the parameters used
> # with array_merge are never null values and always arrays.
> # 
> # That seems to have fixed the problem.
> 
> Hi Kevin,
> 
> Thanks for letting me know about this.
> 
> Php.net has this rather cryptic bit to say about bug_compat_42:
> 
>   PHP versions 4.2.3 and lower have an undocumented feature/bug that
>   allows you to initialize a session variable in the global scope,
>   albeit register_globals is disabled. PHP 4.3.0 and later will warn
>   you, if this feature is used, and if session.bug_compat_warn is also
>   enabled. This feature/bug can be disabled by disabling this
>   directive.
> 
> I don't know what "initialize a session variable in the global scope"
> even means.  Do you?  (It's probably something from PHP's dark past
> that I *don't* want to know is possible.)  I'm not sure how this
> affects our libraries specifically.  Our code actually makes no
> assumption one way or the other.  (Rather, it assumes sane behavior.)
> 
> I know that in my PHP configuration (both apache and cli),
> bug_compat_42 is set to 1.  However, I'm not sure it has ever had any
> bearing on what I expect my code to do.  (So, I would love some
> clarification on the nature of this, if you can provide it.)
> 
> Thanks again,
> 


-- 
Kevin Jardine
Radagast Solutions

Internet campaign advice and magic
http://radagast.biz

YIM: kevinjardine
Skype: kevinjardine
Eml: kevin at radagast.biz
Tel: +31 (0)6 25581608


From dev-list-openidenabled at thequod.de  Thu Jan 25 15:15:04 2007
From: dev-list-openidenabled at thequod.de (dAniel hAhler)
Date: Fri, 26 Jan 2007 00:15:04 +0100
Subject: OpenID-simplereg: Split openid.sreg.fullname into .forename and
	.surname
Message-ID: <bcadbdf50701251515s2d632f9ekc20b54dc55e4e24@mail.gmail.com>

Hello,

IMHO it would be better to have separate fields for "First name" and
"Last name" (forename and surename).

You can always build fullname out of this, but not the other way
around reliably.

What do you think?


From kevin at janrain.com  Thu Jan 25 15:48:00 2007
From: kevin at janrain.com (Kevin Turner)
Date: Thu, 25 Jan 2007 15:48:00 -0800
Subject: OpenID-simplereg: Split openid.sreg.fullname into .forename
	and .surname
In-Reply-To: <bcadbdf50701251515s2d632f9ekc20b54dc55e4e24@mail.gmail.com>
References: <bcadbdf50701251515s2d632f9ekc20b54dc55e4e24@mail.gmail.com>
Message-ID: <1169768880.5733.407.camel@localhost>

On Fri, 2007-01-26 at 00:15 +0100, dAniel hAhler wrote:
> Hello,
> 
> IMHO it would be better to have separate fields for "First name" and
> "Last name" (forename and surename).

Names get pretty complicated pretty fast.  Sure, it's just first and
last, you say.  But what about middle names?  Multiple middle names?
Titles?  Suffixes?  Structure and presentation of names also varies by
culture.  That's why we decided the *simple* registration extension
would just have one fullname field, and anything with more complex
requirements could be served by something not called "simple."




From kevin at radagast.biz  Thu Jan 25 23:05:56 2007
From: kevin at radagast.biz (Kevin Jardine)
Date: Fri, 26 Jan 2007 08:05:56 +0100
Subject: OpenID-simplereg: Split openid.sreg.fullname into .forename	and
	.surname
In-Reply-To: <1169768880.5733.407.camel@localhost>
References: <bcadbdf50701251515s2d632f9ekc20b54dc55e4e24@mail.gmail.com>
	<1169768880.5733.407.camel@localhost>
Message-ID: <45B9A854.4070300@radagast.biz>

As a concrete example, in the Netherlands, many people have 
Tussenvoegsels (eg. "van", "de") which are technically part of the last 
name, but are not counted as such for sorting. Eg. "Van Dyke" is found 
under "D" in Dutch phone books.

In Hong Kong, many people adopt English given names, and the order of 
the name depends upon the first name. If the given name is an English 
name, it typically comes first, but if it is Chinese, it typically comes 
last. There are exceptions, however. Some Chinese citizens prefer to 
reverse their Chinese names in an international context, and some do not.

In some countries, eg. Mongolia and parts of Indonesia, people 
traditionally do not have last names.

In Iceland, there is something a bit like a last name, but it varies 
from family member to family member based on gender.

And so on.

The "firstname lastname" model is actually restricted only to a few 
cultures.

Cheers,
Kevin

Kevin Turner wrote:
> On Fri, 2007-01-26 at 00:15 +0100, dAniel hAhler wrote:
>> Hello,
>>
>> IMHO it would be better to have separate fields for "First name" and
>> "Last name" (forename and surename).
> 
> Names get pretty complicated pretty fast.  Sure, it's just first and
> last, you say.  But what about middle names?  Multiple middle names?
> Titles?  Suffixes?  Structure and presentation of names also varies by
> culture.  That's why we decided the *simple* registration extension
> would just have one fullname field, and anything with more complex
> requirements could be served by something not called "simple."
> 
> 
> 
> _______________________________________________
> Dev mailing list
> Dev at lists.openidenabled.com
> http://lists.openidenabled.com/mailman/listinfo/dev
> 
> 
> 


-- 
Kevin Jardine
Radagast Solutions

Internet campaign advice and magic
http://radagast.biz

YIM: kevinjardine
Skype: kevinjardine
Eml: kevin at radagast.biz
Tel: +31 (0)6 25581608


From kevin at radagast.biz  Thu Jan 25 23:53:31 2007
From: kevin at radagast.biz (Kevin Jardine)
Date: Fri, 26 Jan 2007 08:53:31 +0100
Subject: OpenID-simplereg: Split openid.sreg.fullname into .forename	and
	.surname
In-Reply-To: <45B9A854.4070300@radagast.biz>
References: <bcadbdf50701251515s2d632f9ekc20b54dc55e4e24@mail.gmail.com>	<1169768880.5733.407.camel@localhost>
	<45B9A854.4070300@radagast.biz>
Message-ID: <45B9B37B.2020507@radagast.biz>

I forgot to mention that the name order is not only an issue in Asia. In 
Hungary, the family name is also written before the given name.

See:

http://en.wikipedia.org/wiki/Hungarian_name

Cheers,
Kevin

Kevin Jardine wrote:
> As a concrete example, in the Netherlands, many people have 
> Tussenvoegsels (eg. "van", "de") which are technically part of the last 
> name, but are not counted as such for sorting. Eg. "Van Dyke" is found 
> under "D" in Dutch phone books.
> 
> In Hong Kong, many people adopt English given names, and the order of 
> the name depends upon the first name. If the given name is an English 
> name, it typically comes first, but if it is Chinese, it typically comes 
> last. There are exceptions, however. Some Chinese citizens prefer to 
> reverse their Chinese names in an international context, and some do not.
> 
> In some countries, eg. Mongolia and parts of Indonesia, people 
> traditionally do not have last names.
> 
> In Iceland, there is something a bit like a last name, but it varies 
> from family member to family member based on gender.
> 
> And so on.
> 
> The "firstname lastname" model is actually restricted only to a few 
> cultures.
> 
> Cheers,
> Kevin
> 
> Kevin Turner wrote:
>> On Fri, 2007-01-26 at 00:15 +0100, dAniel hAhler wrote:
>>> Hello,
>>>
>>> IMHO it would be better to have separate fields for "First name" and
>>> "Last name" (forename and surename).
>> Names get pretty complicated pretty fast.  Sure, it's just first and
>> last, you say.  But what about middle names?  Multiple middle names?
>> Titles?  Suffixes?  Structure and presentation of names also varies by
>> culture.  That's why we decided the *simple* registration extension
>> would just have one fullname field, and anything with more complex
>> requirements could be served by something not called "simple."
>>
>>
>>
>> _______________________________________________
>> Dev mailing list
>> Dev at lists.openidenabled.com
>> http://lists.openidenabled.com/mailman/listinfo/dev
>>
>>
>>
> 
> 


-- 
Kevin Jardine
Radagast Solutions

Internet campaign advice and magic
http://radagast.biz

YIM: kevinjardine
Skype: kevinjardine
Eml: kevin at radagast.biz
Tel: +31 (0)6 25581608


From l4igem2 at list.ru  Fri Jan 26 04:46:26 2007
From: l4igem2 at list.ru (Sergey Shepelev)
Date: Fri, 26 Jan 2007 15:46:26 +0300
Subject: index.php/serve over FastCGI problem
Message-ID: <1268164493.20070126154626@list.ru>

Hello.

 I'm using nginx webserver and PHP/fastcgi to run OpenID server.
 I encountered problem when user requests an URI like
 page.php/path?param=value. It doesn't get served by PHP. Possibly it
 is a httpd-php connection problem but either way... is this thing
 really a must? Can I change index.php/serve to something like
 index.php?serve=1 in all files? (With changing index.php request
 handlers 'if' statement as well). Won't this break compatibility with
 OpenID clients?

-- 
? ?????????,
 ?????? ?. ???????.
mailto:l4igem2 at list.ru
ICQ: 145-542-767



From simon at simonwillison.net  Sat Jan 27 08:16:46 2007
From: simon at simonwillison.net (Simon Willison)
Date: Sat, 27 Jan 2007 16:16:46 +0000
Subject: Python library: getRetryURL equivalent in v1.2.0
Message-ID: <a3b4f3b60701270816w12c66053w257c63172b1fc99d@mail.gmail.com>

Hi all,

Version 1.0.5 of the Python OpenID library included a useful set of
functionality for retrying OpenID authentication (for example, if the
user wasn't originally signed in to their identity provider). In
particular, there was a method called getRetryURL()

This method no longer exists in version 1.2.0 of the library. What is
the expected way of duplicating this functionality?

As an aside, the browsable documentation on openidenabled.com is out
of date - it's still the docs for the previous version of the library.

Cheers,

Simon