At 22:21 18-8-2002, Edin Kadribasic wrote:
> > >As long as there is momentum on the development process and on the QA > > >process when needed, > > >I don't think release momentum matters that much. > > > > Right. Since the first part of the sentence does not stand in reality, >the > > direct result is that momentum matters, a lot :) Lack of momentum is the > > main reason PHP releases are taking several weeks now. It's not as if > > people are testing it thoroughly for weeks and weeks... > >Exactly right. During the last release cycle, Derick made elaborate test >cases, etc. but the result was not very promising. Very few people actually >took part in the QA process and as a result 4.2.0 was released with some >serious bugs in it. And those who did take part were confronted with "Oh yeah, that test is not good" (I don't mean the special cases that Derick compiled, as they are more aimed at compilation). So again: Is the testkit QA-proof? Are the tests that fail, failing due to a bug or a buggy test script? See below for php-STABLE, on AIX. Especially the array stuff, where many fixes have gone and the first three fail and everything relating to aggregation fails. Since we know the OO bug test is bogus, can we remove that? Further more - how about adopting the regression tests that Sebastion Nohn has compiled? And since we're providing bcmath as a module, I'm surprised there are no tests for that. >I believe that 4.2.3 should be released, and that it should be very fast. If >Derick has no time for the release, and Zeev volunteers to do it, I really >see no point in delaying it further. Especially since 4.3.0 release doesn't >seem to be very near. +1 Another argument for it: Having a stable 4.2 release invites people to play with the new features in 4.3. $ grep failed ./test.out Session Object Serialization (001.phpt) ... failed Session Object Deserialization (003.phpt) ... failed session_set_save_handler test (004.phpt) ... failed Custom save handler, multiple session_start()s, complex data structu ... failed References between variables in sessions (006.phpt) ... failed aggregating everything (aggregate.phpt) ... failed aggregating all methods (aggregate_methods.phpt) ... failed aggregating methods specified in the list (aggregate_methods_by_list ... failed aggregating methods matching regular expression (aggregate_methods_b ... failed aggregating all default properties (aggregate_properties.phpt) ... failed aggregating default properties specified in the list (aggregate_prop ... failed aggregating default properties matching regular expression (aggregat ... failed retrieving aggregation info (aggregation_info.phpt) ... failed deaggreating (deaggregate.phpt) ... failed Test array_merge and array_walk (001.phpt) ... failed Test arsort, asort, krsort, ksort, rsort, and sort (002.phpt) ... failed Test usort, uksort and uasort (003.phpt) ... failed Tests for floor en ceil (floorceil.phpt) ... failed Various pow() tests (pow.phpt) ... failed strtotime() function (002.phpt) ... failed DB configurable error handler test (db_error2.phpt) ... failed XML Parser: parse from file (002.phpt) ... failed XML Parser: parse from file resource (003.phpt) ... failed PEAR_Config (pear_config.phpt) ... failed PEAR_Error: default error handling (pear_error3.phpt) ... failed Simple POST Method test (002.phpt) ... failed GET and POST Method combined (003.phpt) ... failed Two variables in POST data (004.phpt) ... failed Three variables in POST data (005.phpt) ... failed Testing $argc and $argv handling (011.phpt) ... failed Output buffering test (006.phpt) ... failed OO Bug Test (Bug #7515) (029.phpt) ... failed Tests failed: 32 ( 24%) Met vriendelijke groeten / With kind regards, Webmaster IDG.nl Melvyn Sopacua -- PHP Development Mailing List <http://www.php.net/> To unsubscribe, visit: http://www.php.net/unsub.php
