Derick, On Sat, 2010-11-13 at 16:31 +0000, Derick Rethans wrote: > Changed paths: > U php/php-src/branches/PHP_5_3/ext/date/php_date.c > U php/php-src/branches/PHP_5_3/ext/date/php_date.h
> -PHPAPI int php_idate(char format, time_t ts, int localtime) > +PHPAPI int php_idate(char format, time_t ts, int localtime TSRMLS_DC) This breaks the ABI in a point release. > -static zval * date_instantiate(zend_class_entry *pce, zval *object > TSRMLS_DC) > +PHPAPI zval *php_date_instantiate(zend_class_entry *pce, zval *object > TSRMLS_DC) > -static int date_initialize(php_date_obj *dateobj, /*const*/ char > *time_str, int time_str_len, char *format, zval *timezone_object, int > ctor TSRMLS_DC) > +PHPAPI int php_date_initialize(php_date_obj *dateobj, /*const*/ char > *time_str, int time_str_len, char *format, zval *timezone_object, int > ctor TSRMLS_DC) Adding new APIs in point releases has been done before, but should be done with care. > /* Backwards compability wrapper */ > PHPAPI signed long php_parse_date(char *string, signed long *now); > PHPAPI void php_mktime(INTERNAL_FUNCTION_PARAMETERS, int gmt); > -PHPAPI int php_idate(char format, time_t ts, int localtime); > +PHPAPI int php_idate(char format, time_t ts, int localtime > TSRMLS_DC); > #if HAVE_STRFTIME > #define _php_strftime php_strftime > PHPAPI void php_strftime(INTERNAL_FUNCTION_PARAMETERS, int gm); Ah, I remember the discussion. How can it be a "Backwards compability wrapper" if it isn't backwards compatible in its signature? Or does the comment apply only to php_parse_date()? Can we use TSRMLS_FETCH() in 5_3 and move this away from the comment in trunk? johannes -- PHP CVS Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php