Welcome to “This Week in the Zend Framework”. Each week we stop for a moment and take a look at the developments in the framework. We examine the development process through the threads on the mailing list. We analyze the intricacies and nuances of the important posts. In short, we rehash the nuttiness that surrounds the development of a highly complex piece of code. So join us this week as we stroll along memory lane.

This week started off slow but picked up speed and by the end of the week we had a full-blown discussion going on. Actually there were 3 major threads that almost everyone not actually busy coding took part in. (and a few of our intrepid code warriors took time out of their busy schedules to chime in.)

  • First up, Ralph Schindler kicked us off with Zend_InputFilter_DateFormat or something…. This kicked off the first of several +1 fests on making PHP 5.1.x a requirement for the Zend Framework. Currently it stands a 5.0 but given the responses, I personally believe that may change. (Then again, I thought Blu-Ray was a lock, what do I know?)
  • Andi Gutmans continued the discussion on proposal Zend_Timestamp, Zend_Date and Zend_Time requesting clarification on the proposal. Again, this one turned quickly to the topic of 5.1 because most of the functions in the proposal seem to be covered in the new Date extension available in 5.1.
  • Inash Zubair had a short monologue regarding Preview of improvised Zend_Http_Client. He published proposed code changes (and then a patch to his code changes to fix the bugs) to, um, can you guess which class? (I thought you could.)
  • Drew Bertola (/me waves hi to drewb and all of IO) posted his own monologue about a problem he was having. Problem (bug?) with View/helper. Eventually Arnaud Limbourg interrupted him to tell him that this problem was a known issue but not before Drew posted it to the bug tracking system.
  • Andi Gutmans announced that the Framework team had finally had time to finish reviewing all the proposals that had been made so far. His thread Proposal Summary was the driving force behind a lot of the mail on the list this week. From the comments they made on some of the proposals, it’s obvious that a lot of thought and discussion went into them. Whether your proposal was accepted or not, you’ve got to respect the fact that sometimes hard choices have to be made.
  • Andries Seutens opened a record breaking thread Zend_Tree rejected. The 32 message thread started out just announcing to the list that the framework team had not agreed that the component was core-worthy. It then went on to discuss what should be done with proposals that are rejected by the Framework team. Several ideas were tossed around and as it started to veer off-course, it spawned the thread Zend Component Source Forge Repository that continued on with parts of the discussion. The original discussion was to see if there was enough community support for developing the component. The second thread stayed on-message by discussion various methods for setting up a repository for items that are deemed “good” by the community but not officially sanctioned by Zend. Surprisingly, it doesn’t seem as simple as throw up a website and let people upload. In my never-ending quest to pretend I’m a journalist, I don’t often express personal opinions on the subjects I cover. Since I can’t figure a monetization angle on this one, I won’t here either.
  • Rob Allen posed several questions and thoughts about his ZF_Conf proposal that was marked PENDING by the framework team in the thread ZF_Config Proposal Discussion.
  • Jeffrey Thompson started up the thread ZF Forum when he posed the question, hey, instead of/in addition to the mailing list, why not a forum. The short answer….um, no. (if everybody used this new web thing, where would our mail admins work?) Long answer, well, I’ve given you the link to the thread, do I have to do everything? Regardless of your feelings on this one, expect to see it again. It’s one of the mailing list perennials. Bookmark this one along with the “Mailing List Standards” discussion so next year we can just point people to this one and move on to more important discussions like rewriting Trac in PHP.
  • Speaking of perennials, Lucene turned up again spawning 3 different threads. Zend Search Lucene is a continuation from last weeks discussion. This is a higher level discussion over why there isn’t a PHP extension for Lucene. Zend Search Memory Usage surprisingly talks about…memory management. Apparently, if you are trying to use Lucene to index a bucket-load of databases records, you need 2 bucket-loads of memory. Finally, Zend Search updating index again raised questions about creating an index vs. updating one. If you are going to use Lucene, this is an important concept to grasp.
  • Finally, a bit of obscure news that you may have missed. WE HAVE A BUG TRACKER!. The un-official announcement thread Bugtracker started when Steven Van Poeck found and announced it. Thanks Steven for being on the ball. Now everybody get out there and report those bugs!

Hey hey! That’s it for this week. Visit us next week for more exciting action, same Bat-Time, same Bat-URL.

=C=