Dev List Summary

Hey All, It is pretty awesome to see PHP 4.1.0 roll out the door - shame that I missed it. I ended up being mostly offline with technical troubles. :( Anyhoo, here is the Dev list summary for this week. This is basically just a text version of what I worked up Zend. Feedback and corrections are welcome as always. I will work on the CVS list summary tomorrow. ------ BUG/TLK: Over the last few weeks there have been a series of discussions about how PHP de-allocates memory. While there have been no resolutions yet, you can follow the discussion at: http://marc.theaimsgroup.com/?t=100651165900001&r=1&w=2 http://bugs.php.net/bug.php?id=13437 REQ: A user has requested that existing persistent database connections be closed when the maximum number of persistent connections have been opened. PHP Dev regular Yasuo Ohgaki is looking into the request. To follow the progress of this request, visit http://bugs.php.net/bug.php?id=13463 BUG: A user has noticed that PHP's transparent session id propogation feature rewrites XHTML-compliant forms improperly. See http://bugs.php.net/bug.php?id=13472 FIX: The fix implemented for bug 6755 prevented PHP from compiling on the AIX platform. The fix has been fixed, allowing PHP to properly compile on AIX for the new PHP 4.1.0 release. FIX: Dan Kalowsky fixed a bug that prevented odbc_error() from returning the proper error code in certain situations. The fix should appear in PHP 4.0.2 FIX: Thies Arntzen corrected a bug in imap_rfc822_parse_adrlist() (Yep, that is the real name of the function :) where the function was modifying the first argument passed to it. Thanks to Philip Steeman for discovering and reporting the bug. NEW: Yasuo Ohgaki wrote ob_get_level - a small function to returns the current level of output buffering nesting. Derick Rethans cleaned and committed the function, and then added documentation for it. The function should appear in PHP 4.0.2. Visit http://www.php.net/ob_get_level for mor information. TLK: There was a short discussion on splitting the PHP Dev list into a bunch of smaller, more specialized lists. As has happened with previous discussions on the topic, it was decided that the list should stay whole - despite the high volume of traffic. REQ: Thomas Frössman has suggested an improvement on how PHP generates random numbers for session ids numbers on Windows platforms. As of yet, the suggestion seems to have gone unnoticed. To follow any possible discussion, see http://marc.theaimsgroup.com/?t=100749391700002&r=1&w=2 TLK: There were several discussions about template systems and PHP. The discussions got a bit nasty at times - who knew that templates could be such a touchy subject. ;) NEW: Earlier this year, a user requested a pg_cancel() function that would allow them to stop long running Postgres database queries. Yasuo Okgaki has started on a basic implementation of the user's request. He is currently waiting for feedback from another developer before proceeding further. See http://bugs.php.net/bug.php?id=9273 for more details FIX: Zeev Suraski improved the handling of error situations for unbuffered MySQL queries in PHP. TLK: A discussion on the behavior of exit() was revived when one of the PHP Dev regulars re-encountered an odd behavior of the construct. The discussion centers around exit's behavior of sending integer return values to stdout in addition to setting them as an exit status in the shell. For more information on this discussion (and the discussions that preceded it), visit: http://marc.theaimsgroup.com/?t=99048980400003&r=1&w=2 http://marc.theaimsgroup.com/?t=100170293000002&w=4&r=1 http://marc.theaimsgroup.com/?l=php-dev&m=99539268507735&w=4 NEW: Benjamin Yates has provided a patch for the socket extension that improves how the extension deals with and reports errors. Thies Arntzen is reviewing the patch and, barring any problems, plans to integrate it. REQ: A user has requested that support for SNMP version 3 be integrated into PHP 4. No one has responded as of yet - to follow any further developments, visit http://marc.theaimsgroup.com/?w=4&r=1&s=snmpv3+for+php4+and+later+&q=t TLK: Rui Hirokawa has made an interesting proposal for improving and extending PHP's handling of GET/POST/Cookie (GPC) data. He suggests that user's be able to define handlers that filter, check or otherwise manipulate data from GPC sources. To follow response to Rui's proposal, visit http://marc.theaimsgroup.com/?t=100789136000059&r=1&w=4 --zak