worklog

I am no longer the victim of mind dribbling porn purveyors, I have patched my wordpress to the latest…

At least it was only bad for a couple of weeks.

http://canvaspaint.org/ A copy of MS paint using JS and CSS and the canvas tag. Don’t bother trying it with IE, you’ll need a modern browser like firefox, opera or safari

Listening to a webcast of an interview with Alexander Limi about accessibility concerns in Plone and how it was built into Plone from the beginning. It turns out that by building Plone to standards they inadvertently made Plone mostly accessible. After they saw how some disabled users were able to use Plone without having previously thought about it, the development team made a conscious decision to persue and improve accessibility as a core development value in Plone. As a nice aside, he slams the AJAX craze.

There is a summary at webstandards.org http://www.webstandards.org/2006/09/02/accessibility-webcast-on-plone/. For the full webcast, checkout The National Center on Access and Education’s full webcast: Accessibility and the Open Source Content Management Movement.

Ok, we have too many listservs and noone asks if I’d actually like to be on one. Isn’t that what RSS and feeds are about? Tell me where the information will be and I’ll go get it when I want it.

Sometimes good things are broken…

Had to fix the contents of template-functions-post.php, when iterating through pages it was outputting an extra <li tag. wrapped around an <ul> some stuff </ul> without a parent <ul element. This was in the function wp_list_pages call.

That was the only fault of wordpress. The template however… wow, it was kind of broken. I didn’t notice until I noticed that I had two of the cat backgrounds showing up and couldn’t figure out why - turned out to be multiple div closings when there weren’t divs and then the misuse of the id attribute - naming multiple things for the same id. Moral of the story? 1) Validate your pages, 2) Understand looping - it seems that sometimes people don’t know when or where to stop

I can’t remember if I blogged about this earlier, but a friend passed me a link to this series on using open source software to do a corporate website. It is a pretty good read:

Using open source software to design, develop, and deploy a collaborative Web site

One CMS that they leave out is Joomla which I guess would be more or less equivalent to Mambo for purposes of the article since they have the same parentage. Joomla has undergone some pretty serious reorganization from Mambo and their road map is interesting, but in my experience with Joomla, I have to agree with their basic critique of Mambo:

The easy installation seemed to get us to a point where almost all the function we needed was available and ready to be themed. However, as with many CMSs, the templating is limited to a tag system that leaves you at the mercy of the quality of the markup that is substituted for the tags. This is fine if the markup is valid, semantically structured, and adequately sprinkled with CSS ID and class attributes to aid styling. If it isn’t, then you can find yourself delving into the guts of the application to figure out how to correct the generated output.

I found Joomla pretty easy to get started with, but then building custom components was problematic and the documentation didn’t cover why you would do something one way or another or how to elegantly tie components into modules into categories…

Via webstandards the details on IE7’s fixes to microsoft’s implementation of CSS are out at the IE blog. Of particular interest to me is this: ”

<?xml> prolog no longer causes quirks mode

Cool, does not having it cause quirks mode for IE7? Because right now, since IE6 is idiotically implemented such that an <?xml> prolog does cause quirks mode, help.unc.edu strips the prolog to make sure that IE6 goes into standards mode - quirks mode causes wierd border issues and spacing. I’m glad that now I’ll have to sniff browsers once again…

Today help.unc.edu had some problems. Some of the documents were causing Cocoon to spin out into space and never return. The cause? Well, the DOCTYPE declaration on the top of some of the xhtml fragments that we pre-publish. The W3 site was down (www.w3.org) and Cocoon was trying to get the XHTML transitional dtd from that site.