DistroWatch.com: Sabily

This is example why Linux blows Windows (and MAC OS) off the planet – the proliferation of distributions of linux is growing, not diminishing.  What Linux enable is users being able to take core code and build their own OS.  This particular one is aimed at Muslims – there are other aimed at schools, hackers … this is a true OS, one that isnt so inflexible and doesnt dictate which version you have. The most friendly one is fedora – installs generally go like a dream, dealing with obscure hardware components with ease.  Linux has come a long way.  And after Vista fiasco, surely its time for you to consider an OS that doesnt let you down so often, and gives you what you need, rather than loads of pointless crap you dont.

DistroWatch.com: Sabily

Semantic-butchering corporations

During recent consultancy assignment, I have seen a dangerous misconception of a semantic web technology with taxonomy. Tags are a set of words that can be associated, in single or multiple assignments, to content. Although you can specify a set of tags before you even start generating content, generally tags ressult from evolution of a website, rather than be designed. It is not mandatory to organise tags in a hierarchial structure – that is more for ease for the human brain, which finds it easy to process procedural information. But in order to follow the semantic/microformat principles (data stored for ease of human and machine), it is better general practice to organise tags in groups and sub groups. These can be changed/repositioned at any time, with no impact on exisiting data. Unless ….

OK, so we have our nice taxonomy, and feel confident we are following new web development methodologies. Then we have the site application. One ridiculous peice of content management I have seen recently, is when content tagged with multiple tags, ends up published numerous times. Why? Because the way the site has been designed at folder level, was based on old-fashioned hierarchy. Each tag had it’s own folder, and as a tag could also be duplicated in many sub-folder names, the CMS created duplicates in all the same-named folders.

This is a problem when a CMS simply nod towards semantic web principles. The danger being, there is a part-implementation, and the classic mistake of trying to fit a square peg (outdated CMS) into round hole (semantic web). So we now have a massive inefficiency on the website itself – the web application portal on the www. All he careful tagging work done at content level is now effectively useless. Add a breadcrumb navigation, and be assured it will look nonsensical to the user. They will instinctively recognice the classic hierarchial menu, then get confused why when they click on “Fashion” -> “News”, then click on an article link to end up in “Retail” -> “News”. Modern CMS’s do (to varying degrees) seperate out the data from the design, but this is no good if they do not also understand why that is important.

There are very few corporate level CMS’s that are future-proofed, as they rely on constant support and upgrade contracts. But maybe now is the time to realise that maybe that is not what is needed – with a flexible, content could be provided in all sorts of way. Aggregation is increasing more and more, rss/opml have become standards for feeding news/articles, and other microformats have evolved to cater for other types of information. These can be managed very simply internally, as the UI’s for data entry are simple and structured. This is just the start – once these standards become more and more adopted, the semantic web will come ever closer. If you want to be seen, you have to do the work to get the user’s attention.