Weekly todo activity, Sep 19 to Sep 26

Weekly todo activity, Sep 19 to Sep 26

Task #24: Plugins (bring back interp hooks or events)
Sat Sep 25   Jonathan Swartz modified summary
  http://www.masonhq.com/docs/todo/access.html?id=24

Task #474: More complete CGIHandler emulation of $r
Sat Sep 25   Jonathan Swartz changed version_goal from 1.2x to 1.3x
             Jonathan Swartz writes: Move any further development into 1.3x
  http://www.masonhq.com/docs/todo/access.html?id=474

Bug #524: Add $m->internal_redirect (was: $m->abort and $m->redirect do not propagate out of subrequest)
Sat Sep 25   Jonathan Swartz modified summary
  http://www.masonhq.com/docs/todo/access.html?id=524

Bug #537: <% #foo %> causes a syntax error at component compilation
Tue Sep 21   Jonathan Swartz changed version_goal from Future to 1.27
             Jonathan Swartz writes: This syntax seems to work in 1.26. Not
               sure when it was fixed. Just need to add tests (for both
               single and multi line case) and we can close this bug. 
Sat Sep 25   Jonathan Swartz writes: I'm wrong. The bug with <% #foo %> still exists. 
             Jonathan Swartz changed version_goal from 1.27 to 1.3
  http://www.masonhq.com/docs/todo/access.html?id=537

Bug #539: Using 'next' or 'last' without a loop can corrupt the component stack
Sat Sep 25   Jonathan Swartz changed version_goal from 1.2x to 1.3
  http://www.masonhq.com/docs/todo/access.html?id=539

New bug #553: $m is not available in preamble if in_package is not set to the default value

Sun Sep 19   Dave Rolsky created bug
             Dave Rolsky set version_found to 1.10
             Dave Rolsky set severity to Minor
             Dave Rolsky set version_goal to 1.27
             Dave Rolsky set owner to autarch
  http://www.masonhq.com/docs/todo/access.html?id=553

New task #554: Simplify and document output buffer / filter / abort behavior
               The interactions between clear_buffer, flush_buffer, scomp,
               abort, and filters have been poorly documented and somewhat
               haphazardly implemented for a while. In 1.1 the buffer
               object was introduced to fix some of these problems, but it
               ended up causing other bugs, introducing new confusing and
               rarely used features (e.g. the ability to flush intermediate
               buffers), and adding unnecessary overhead to every print. 

               I propose to 

               * Change $m->flush_buffer to only flush the top-level
               buffer, having no effect on intermediate (e.g. scomp or
               filter) buffers. This reflects the method's original intent
               and most common use, and it allows us to clean up buffer
               implementation. 

               * Keep $m->clear_buffer always clearing all buffers. This
               means that intermediate buffers are cleared even if a
               $m->flush_buffer is done previously. 

               * Document comprehensively the interactions between
               clear_buffer, flush_buffer, scomp, abort, and filters, and
               make sure all these cases are backed up with unit tests.
Sat Sep 25   Jonathan Swartz created task
             Jonathan Swartz set difficulty to 2
             Jonathan Swartz set version_goal to 1.3
             Jonathan Swartz set owner to jswartz
  http://www.masonhq.com/docs/todo/access.html?id=554

New task #555: Optimize component calls
               With sites calling up to several hundred components on a
               page, it is critical to weed out any unnecessary overhead
               occurring on each component call. That includes both
               HTML::Mason::Request::comp itself and the generated object
               file.
Sat Sep 25   Jonathan Swartz created task
             Jonathan Swartz set difficulty to 2
             Jonathan Swartz set version_goal to 1.3
             Jonathan Swartz set owner to jswartz
  http://www.masonhq.com/docs/todo/access.html?id=555

New task #556: Allow </&> tag to contain literal start path
               If a <&| &> tag contains a literal string FOO (first
               character one of [\w/_.]), allow either </&> or </& FOO> as
               an end tag. If the tag contains an expression, only </&> can
               be used.
Sat Sep 25   Jonathan Swartz created task
             Jonathan Swartz set difficulty to 2
             Jonathan Swartz writes: See
               http://marc.theaimsgroup.com/?t=109544641200002&r=1&w=2 for
               one of many design discussions.
  http://www.masonhq.com/docs/todo/access.html?id=556




-------------------------------------------------------
This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
Project Admins to receive an Apple iPod Mini FREE for your judgement on
who ports your project to Linux PPC the best. Sponsored by IBM.
Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php