Mathieu Fenniak's Weblog

Not a whole lot technical done today. I put a new navigation tool in the top-right corner of my growlmurrdurr based web pages. However, since it’s larger than the previous one and doesn’t have a black border around it, it sometimes fades into things and eclipses stuff on the web pages… that’s no good.

It’s got some spiffy but simple JavaScript behind it to make the tab-like property work… if it’s big and multirowed to you, then the JavaScript doesn’t work on your web browser. Please let me know. But it works fine for me with Mozilla… except for the problems mentioned already.

I think maybe a more dramatic difference between colours of the tab-like things might look better. Hm… I’m not a web designer, that’s for sure. Just a software designer.

The Growlmurrdurr Picture System is now feature complete and seems to be operating niftily. This is a replacement for the ‘curator’ python script I’ve been using on a number of websites for automated thumbnail-ed image galleries. This system has a number of advantages over the other system:

• Automatic instant updates. When new pictures are added into a directory/gallery, they appear on the web page immediately without needing a periodic task running to update the HTML pages.
• Thumbnails, medium-sized images, and support files are stored out of the way. The /pics/ directory on the web server, for example, just contains the raw source images in their largest available size. Thumbnails and viewable size pictures are automatically created in hidden directories.
• XML files are used to store data on each picture. The title, the description, the dimensions. A web interface exists for editting the title and description of images without creating a multitude of ‘.desc’ files, like the curator script used.
• Comments! It’s a Growlmurrdurr based system, uses XML to store data… it’s just natural that the exact same code for a comment system for weblogs could be reused on my pictures.
• Still maintains nice URLs. No nasty GET URLs full of parameters to scripts and stuff. A URL for a file is generally /directory/index.cgi/directory/picture.jpg, which is pretty darn clean I think.

Although now feature complete, I still have some extra things I want to add into the system, as well as a massive amount of documentation and cleanup. I’m going to resist the urge to clean it up so much that the same sources run the weblog and the picture system though, because that would in reality make the system less flexible. As it is, the picture system is just a massively modified copy of the weblog sources.

Anyways, see the picture system at work on my general pictures and geocaching pictures pages. After some cleaning, I hope to move Lacy’s pics to this system too, and Colin’s site. I’ll have to see what they think of it, and address any concerns they have.