Google Translate

Wednesday, 16 April 2008 - web server error

Hmm, oops. right column error

Sometimes, things just screw up for no reason. Like programming code. Though the programmer writes a few lines, he depends on thousands (millions?) of lines of code in supporting libraries. Inevitably, things will break. And then end users face unexpected errors.

But don't worry, National Library. I've seen Yahoo! break down before. And Facebook had its source code revealed once upon a time. So don't feel so bad. Like the saying goes, "S**t happens."


Saturday, 5 April 2008

In lieu of an official report, we'll blame a couple of complacent guards

It's interesting how the blame for Mas Selamat's escape has been laid squarely on a few detention centre personnel, even though no findings have been revealed officially. They've been labeled as "complacent" and "negligent" and are at the short end of a very big stick for an unfairly long time.

It's also interesting how the government keeps saying repeatedly that it'll reveal everything (that's not classified) in a comprehensive report into the escape. Then, in the other breath, it points fingers at the guards. Who were -- I repeat in case you didn't hear it the first time -- "complacent" and, therefore, clearly at fault.

In fact, it's more interesting that the longer the government delays releasing its report, the more often it is that unofficial but official-sounding statements keep being mentioned. Very soon, there won't be a need for the report, except for bureaucratic, red-tape reasons, because everyone would already have accepted those statements as "facts".

Finally, it's interesting that judgement has been delivered before findings and due process have been completed. Especially from a legal perspective.


Ministry of Education relaunches its website

Ministry of Education logo
Recently, the Ministry of Education's website was revamped and I was informed about it by Lucian, the webmaster (whom I know through WebSG). He had actually told me about the beta site, but I held back on my review till after the actual site was relaunched.

Being a guy dedicated to Web standards, it's not surprising to find that the revamped MOE website is clean and polished, both on the front-end and behind the scenes. Due to standards conformance, it loads quickly and smoothly in modern browsers (I tested with Firefox on a Mac). It also includes some Web 2.0 flashiness, like a scroll-slideshow smack centre on the homepage.

What I'm most impressed, from a programming point-of-view, is the "Organisational Structure". All of the contacts make use of VCard, a microformat. Some browsers would be able to detect this and automatically create a contact card for you. This contact card could then be used with, for example, Mac OS X's Address Book. No need for paper name cards!

Another feature, though not readily known unless you read the page's source, is that the hierarchical chart is not a graphic, but relies on stylesheets and VCards! Minimal graphics involved, and still usable if you look at it with a text-based web browser (e.g. those used by blind people). Apparently, Lucian designed this himself. In the words of Keanu Reeves, "Whoa!"

But, of course, there are bound to be things that don't please everyone. In my case, I found three things, which were present already in the beta version, that I found wanting.

Some may say that the visual design is too spartan, and I tend to agree. Wise choices of colour and good use of graphics would definitely help in drawing the eye to key areas, especially on the content-rich homepage. The way information is presented now makes me feel like I'm reading a novel.

I also thought that the main menu was not very user-friendly. The way it works now is: you click on an item in the menu bar, and a submenu with an accompanying image would slide out. Click another item and another submenu-plus-image would slide out. Unfortunately, due to the use of the image, the submenu is always positioned to the left, not directly below the clicked menu item, as one would assume.

According to user interface gurus, the distance needed for a person to move his pointing device (e.g. mouse) to perform an action should always be minimised. (The joke is that the shortest distance between the mouse and clickable area is when the mouse is directly over that area.) Forcing the visitor to move the mouse to the left to click on a submenu item, unfortunately, breaks the menu's ease of use.

(There's also a minor "error" with the menu, which I can repeat sporadically. When clicking on a menu item (to expand the submenu) on a freshly loaded page, there'll be a flicker where the full submenu is shown, before its sliding effect takes over.)

Finally, I think that the information architecture may need to be tweaked. Case in point: on the homepage, I can go straight to a page called "Desired Outcomes of Education". According to that page's bread crumb trail, it falls under the "Education" section. However, looking at the main menu, there's no "Education" menu. In addition, I can't find the page in any of the submenus!

This probably sounds like a trivial issue. But in the long-term, it need not be. What happens when the homepage's content is refreshed and the "Desired Outcomes" link is removed? If a person returns to the site looking for it, how will he find it? Well, he could definitely use the search box. But I think it would be faster for a person who is used to the navigation system of the site to find the page by clicking through the menus. (And hopefully it doesn't take more than three clicks for him to get to the page!)

Viewed another way, the way that content is organised in the back end (i.e. which section a particular content belongs in) and how it's presented on the front end (i.e. which menu that same content falls under) seems to be disjointed. The front end was obviously designed with visitors' consideration in mind, whereas taxonomy or cataloguing was of bigger importance in the back end.

As a result, the website comes across looking like it suffers from a split personality. I'm sure there are reasons (technical, design or otherwise) for the differences. But to a lay person, this has the potential to confuse him on how to find some particular content. It doesn't help that the sitemap follows the content structure, not navigation system.

In the end, I think most visitors will just rely on the search box, bypassing the fanciful menu system. Whether that should be considered good or bad, I'll leave it to Lucian and his team.