Karma's Nucleus DevBlog

In FireFox, you can create special keyword bookmarks. I created one for the Nucleus bug tracker (on the SourceForge.NET project):

• Url: http://sourceforge.net/tracker/ index.php?func=detail&aid=%s&group_id=66479&atid=514643 (remove the space in the URL. I added it to allow word-wrapping)
• Keyword: bug

If I now enter "bug 967580" in the navigation bar, I end up... at bug 967580

Yesterday, I fixed a problem with skin import/export when CDATA sections were involved. A CDATA section is a way to include just about any "character data" in an XML file, without having to worry about the well-formedness of the XML. It's used like this: <![CDATA[...contents...]]>

Since the only way a parser can tell if such a section is finished, nested CDATA sections are a big no-no. The sequence ]]> can not occur inside a CDATA section.

Nucleus exports its skins/templates as such a section (it's the safest way, since we don't know if its valid XML), so problems arise when the skin/template itself is using a CDATA section. The skinbackup.xml file is no longer well-formed. Although only RSS/Atom/... skins are affected, and import still worked fine, the problem had to be taken care of.

A message from hcgtv: "I'm switching DSL services, I'll be back online in a couple of days."

This post tries to explain how the parsing of skins/templates works in Nucleus.

Let's start with a peculiarty. Everyone is using skinvars as <%skinvar%>, but did you know that these will also work: <%skinvar<%, %>skinvar<% and %>skinvar%>? They do. This is because of the way the parser works.

In my very first post on this blog, I promised to create a design of my own for this site. So far, I haven't done much. More than a week ago, I did some experimenting, but I don't think I'll go that way.

My requirements for a design are pretty strict

• complete separation of content and style (which means that 95% of the design will be CSS-work)
• semantically correct XHTML 1.1 (lists, no tables unless there's tabular data, no <br /> to create visual linebreaks)
• accessibility-guidelines in mind (meaningful page-titles, contents on top, ...)

I've switched to PHP 5 RC3 / Apache 2 on my local test-environment. A first discovered (and fixed) bug: [ 972115 ] PHP5: install.php fails to discover install path. I haven't seen any other problems yet. Radek has had some Blog:CMS users running into problems, however. I'll keep running PHP5 for now, and see what other problems arise.

I've claimed bug 967580 - Class variables not explicitly defined. In the next few days, I'll be taking care of one or more classes each day.

From the NUDN site: "IRC meeting time is set for Wednesday 16/6/2004 at 1800 UTC" (=20:00 for the belgian and dutch developers)

• Agenda
• Feature list to be discussed. Need to make sure I've read through it :)
• irc.freenode.org #nucleuscms (feel free to join)

The new Nucleus logo always makes me think of the logo of a gingerbread (peperkoek in dutch, pain d'épices in french) company called Vondelmolen. It also has an "L" in the middle of the name, and has a windmill in the logo. I always start my day with a slice of their gingerbread, so I get to look at the logo often enough.

I'm pretty sure Moraes (the Nucleus logo designer) didn't know about this logo, so plagiarism is out of the question :)

Update 2004-06-13: The commit from yesterday contained an error which caused nested ifs not to work correctly. Now fixed.

Today, I committed a small improvement to the PARSER.php class into CVS. It greatly improves performance on sites with complex if-else structures.

In a comment on my previous post on the Atom/RSS feed link, hcgtv pointed out that when clicking the Atom link, he was asked to download it. Actually, this is not a bug, it's a feature: the atom.php file sends it contents as application/atom+xml (it's listed in the Atom Syndication format spec, but I don't see any requirements about this). I would expect feed readers to subscribe to this MIME type and allow you to subscribe to the feed when opening it.

In general, Atom and RSS are not meant to be displayed in a browser as raw XML. In comes the feed:// URI scheme. With this scheme, clicked feed links will open in your feed reader. FeedDemon v1.10 (which I use) supports this scheme. I don't know about others, but I expect it to become common use soon.

There are actually two approaches to avoid the XML showing up in the browser: one is to use the feed:// URI scheme. The other is to apply some style to the feeds like Blogger and Blog:CMS do (example). Or both :)

New in the right side column: links to feeds (Atom/RSS) and to the other sites that are part of NUDN.

One suggestion that was recently sent to me, was to add Google AdSense advertisements on the forum and/or nucleuscms.org.

I'm not opposed to the idea. I wouldn't add them to the nucleuscms.org page though: only to the support forum. The only possible downside could be that the ads might link to other tools :)

Other thoughts on this are welcome.

You might have noticed that the "NucleusCMS developer network" logo had some gray garbage pixels immediately next to the "S" from Nucleus. I've taken these out (I guess they were part of the CMS part) and saved the file as dev3.gif

Looking forward to create a design for this site, I worked through the chapter "Styling a Weblog" in the More Eric Meyer on CSS book.

It's quite an interesting book. The book is project based, and contains some interesting chapters. I'm looking forward to do the projects on CSS driven dropdown menus, attractive tabs (sliding doors technique) and a Zen Garden design

Be warned, howerer: This book won't teach you about CSS. It assumes you already know the theory and puts it into practice. By reading some chapters (I worked through 3 out of 10 now) I already learned some interesting new things.

I have a Ctrl-S disorder. I keep hitting that key-combination, even if I haven't changed anything in a file.

I'm back again with an english dev-blog, as part of the Nucleus Developer Network. In English. The main website for NUDN can be found at http://dev.nucleuscms.org/ and collects thoughts/news from all developers.

In setting up this blog I found a first problem: the login cookie conflicts with the login cookie I have for nucleuscms.org (the main website). Resulting in the annoying 'login for each page you visit' problem. Whenever I visit an admin page, the cookie from .nucleuscms.org is sent instead of the one from karma.nucleuscms.org. I now solved it temporarily by manually editing the database so the loginkey for both sites is equal. I'll think of a better solution later.

I've also imported the skin from hcgtv (it worked smoothly!) for now. I'll try and come up with a design of my own soon, though.