Heidily Ho campers! It’s that time again. Yep, the old inbox is full so I know it’s time to do a framework mailing list roundup. So roll yer jeans up to yer knees and wade on in, you’ll get used to the smell after a while.

First, it occurs to me (as I’m sure it has to you) that the framework roundup needs a theme-song. (No, not Code Monkey I’m thinking more of Raw Hide but with a techno beat.)

Ok, let’s do this thing.

  • First out of the chute, a proposal. Proposal: Zend_Config posted by Rob Allen. This one lit up the night sky folks. An interesting proposal and a lot of good discussion about it. It seems that everybody has an opinion about exactly how configuration should be handled, Rob was kind enough to formalize his. If your too lazy to go dig it up and read it, you’re probably a manager; here’s the overview.

Zend_Config is a very simple configuration file reader. It provides an easy means to read configuration files
and access the data within them as a set of key->value pairs. It will support at least one nested level of data.
Initially providing support for ini files, it should be easy to extend for other formats such as YAML.

  • Rui Barreiros unintentionally opened a can of worms by posting his detailed thoughts on Auth/ACL/Perms, Framework issues etc. His intention was to spur on the development of the Auth classes and at first there was a little discussion about this. The thread was quickly hijacked by a discussion on how the mailing list is run. Normally I wouldn’t bother mentioning this except it is just so darned fun to watch the flame-wars erupt. The Star Bellied Sneeches insisting that there is actually something called a standard for how mailing lists are run and those without stars insisting that because they like it the other way, it should be so. Seriously, the only Internet standard for mailing lists is “The person paying the bandwidth bill gets to decide.” Rui, I’m sorry, your thread got stole.
  • Ralf Eggert, who can always be counted on for a well thought out email, started a thread on Form generation without configuration. A very interesting “non-proposal” which in the end, turns out to be exactly like something someone else has already written. It’s still a great idea and while I realize Solar won’t be integrated into the Zend Framework, I am hoping they can both play nice together. (meaning we can at least share ideas if not code) (/me waves hi at Paul!)
  • Richard Thomas (after all the proper forms were filed) submitted an official proposal for Zend_Perms Proposal. I’m not sure if I’m just the only one who saw this or what but it generated absolutely no discussion. Have we beat the ACL horse to death already? Richard, I thought it was an good proposal. (but then who am I anyway?!?) Here’s the overview:

Provides a multi-level binary access layer allowing user permission access to code.

  • Prasanna Ramakrishnan (look Steven Van Poeck, I’m copying and pasting names!) found an issue with Zend_Search. We haven’t seen that module mentioned in a couple of weeks. The thread ZSearch Problem pointed it out. Luckily for us, Alexander Veremyev (Sweet generic deity reference, is no one named Bob Smith anymore?) the principal programmer on the Zend_Search team has already found it and fixed it. Alexander summed it up with:

I should also mention, that Zend_Search_Lucene doesn’t manage segments automatically now. Lucene index files are not suitable for adding documents to the segments. It’s a feature.

  • Chauncey Thorn started an interesting thread on Zend_Filter::isCcnum. Not only is it interesting to read but there are some useful links in there for anyone who writes programs that work with credit cards. Thanks Chauncey for the thread, the code and a name that doesn’t tie my fingers in knots.
  • Zend_Filter_Input got more ink this week when Manuel Pichler proposed code to help eliminate the E_NOTICE warning that popup when used. Rob Allen and Nico Edtinger both contributed code and ideas to this thread as well.
  • Andi Gutmans, in the thread Zend Framework surpassed the .NET framework pointed us all to his blog where he talked about the fact that if you search MSN for the term framework, Zend Framework is the #1 link. Everybody had a good laugh at the evil empire’s search engine and then we all got back to work.
  • As if the informal proposals and the formal proposals weren’t enough, now we have a new category to track the Pre-proposal. And the first one spotted in the wild was this week’s Pre-Proposal: Zend_Pdf_Html. This message sparked quite a discussion on the usefulness of a HTML-to-PDF converter. Seems the “we don’t need no stinkin’ converter” crowd came out on top but not before an interesting discussion broke out. This thread had a minor relapse with Zend_Pdf_Html big idea. Still doesn’t look good for the proposal but Alexander Veremyev summed it up with:

I’ll take a look on html2pdf, but I am a little bit skeptical in expected results…

  • processing the values of checkboxes & radio buttons by Michael Yevdokimov was an interesting thread. It seems Michael wanted to find a way to process parameters of the same name through a GET. Long (but interesting) story short, can’t get there from here, use POST.
  • Ralf Eggert (making his second appearance in today’s roundup) started a very interesting thread on Routing process for multi language websites. Almost everybody got in on on the action for this thread. If you work with multi-lingual web sites, this is a thread you are going to want to read and follow. Nico proposed an interesting solution to the problem that bears further investigation. I won’t spoil the ending for you except to say it probably won’t work in IIS. Nuff said.
  • Chauncey Thorn (wow, really working the 80/20 rule here today) tossed out a patch for Zend_Filter::isZip(). It got no discussion whatsoever but since I don’t get to mention Canada in my roundups much, I thought I’d throw it in.
  • Ufuk Kocolu pointed out an issue with Router.php. The thread router sub-directory problem points out using the framework from a sub-directory causes the router to always execute IndexController::noRouteAction(). Amit Badkas proposed code to fix the issue and Andi stated that this was an issue that the framework team was aware of and would be fixed in the next version of the router.

Well, I see by the old bogometer that I’ve exceeded this weeks limit. So I’ll wrap it up by saying once again, thanks you to all the contributors in the framework community. If it weren’t for you, I’d be sitting here playing Warcraft.

Until next week!

=C=