Test, test and test some more

Oh I was livid with myself yesterday. I'd worked on a hellish site that wasn't our own and that had been inherited. A total mess really, both front and, as I'm reliably informed by my developer colleague, back end. I was charged with re-designing areas of the site whereas I would dearly have loved to have re-designed and built the whole thing, as would my developer colleague.

What has caused me to cringe inwardly and will do for days to come, was that I'd made a couple of 'minor' adjustments to some pages which I'd foolishly assumed would look the same in Firefox as they appeared in IE. I therefore can't have tested these changes in Firefox before the site went live. Embarassingly it was a safari user who noticed a problem. I'd simply set a border in a stylesheet to zero - an act you would think would be pretty inocuous, but it had extended a table column so that it no longer appeared the same width as its neighbours. I burn with shame. It was immediately fixed (although I have still to work out exactly why that act caused the table to break in that way), but nonetheless an embarassment. Grrr. A lesson learned, I guess. However pressured and rushed you might be to get a site up live and however distracted by parallel projects, get into a routine of test test test again (repeats over and over to self).

Website construction from a third party's design

A recent web design project of mine has been to build the front end to a site that will contain some dynamic, customer-managed content. Now that's fine. However, the customer used a design agency to design the look and feel of the site. This always fills me with a level of intrepidation as more often than not, the designer charged with providing the design has little web knowledge. I was provided with photoshop files representing the relevant pages. A very nice-looking design, but an absolute nightmare to build. That's not strictly true. Easy enough to build if you would be happy to splice and dice images and throw those together in a table, but to build a site with any integrity and especially robust enough to cope with dynamic content, a nightmare.

Oh its all very nice to create a design with a fixed height with fixed text that will never change in size or amount. Then you can do all manner of appealing things with gradient backgrounds, curved boxes, overlapping images. Oh it was enough to make me weep. Now given a few weeks of experimentation, I might have been able to do something that almost resembled the design, but I don't have a few weeks. Despite all the cunning css curved boxes/borders tricks, curved corners are still a headache to build. They still clutter up code and still rely on unecessary images, even if they do look quite appealing. I often question whether anyone notices enough to warrant the extra time, effort and code required. But designers are obsessed with them (I'm allowed to mock designers because I am one).

So I sat and whinged a bit, squinted and huffed, thought 'what the hell am I gonna do with this?' several times over, and then had a bash. I'd pretty quickly worked out what I thought would not be possible and decided my best course of action was to show the client where the design would fall down. I created a page, having had to omit some of the design niceties even at this stage, and then changed the font size via the browser to show how increasing the size would break the page. How all the neatly aligned boxes would change height, how an image would no longer align with the foot of the page, how suddenly you would get ugly white space at the top of the adjacent column and how the menu would kick over two lines. I also created a version that broke when more text was added. I then made another page, simplified and adapted to be 'web friendly' that I felt still sensitively captured the essence of the original design and showed that to the client. To be honest, I don't think they noticed that half of the borders didn't have curved corners or that some of the background gradients had been removed. I then showed how it would adapt to increased font size and increased content. The client was perfectly satisfied with this.

Incidentally, one of the page designs showed two columns of continuous copy. Hmm that would have been straightforward for the customer to update.

Now I've built sites from print designers's designs many times over the years and I am only too aware of their fundamental lack of understanding for what a website is and what needs to be considered when designing for the web. And I don't have a problem with this. Why should they? If day to day they design for print, they wouldn't know how a website should work, how its constructed and how it varies across browsers and how it needs to adapt to changing content. A customer will automatically approach their design company when they consider a website and that's fair enough - they are looking to extend their branding and their identity to the web. And designers are often happy to listen to advice and allow me to adapt their designs as I see fit. Being a designer, I like to think I do this with an awareness for what they are trying to achieve. Its often a refreshing experience and a challenge as it allows you to get outside your own design mindset and work on something that's differently approached. Sometimes trying to realise someone else's vision forces you to discover new techniques and breaks you out of a web designer's mentality. However there some fundamental web design rules that need to be at least acknowledged if not followed.

What annoyed me in this instance was that on visiting the agency in question's site, they actually sell web design as a service. Was a bit baffled therefore as to why they weren't building the site and why they'd provided me with what I can only describe as a print designer's web page design.

Ooooh macbook - oooohhhh...

A friend just bought a new macbook and I have to confess to being rather seduced by Apple's packaging and marketing material. It does look pretty. As does the cd sized manual and the tiny remote control. I started with Macs (well actually I started with a Dragon 32 but that's for another post!) as a designer and moved onto PCs as I journeyed into the area of web design. I did own a G3 on which I dabbled with print design, but I've kinda got out of macs. I still get Mac User regularly because, well, I guess I kind of like the idea of being a trendy designer mac user, but do I think they are superior? I'll reserve judgement. Its all horses for courses, but they sure do look priddy!

Working on a chinese website

I've recently been charged with creating a stylesheet for a Chinese website. I've had some vague experience with Japanese, but that was several years ago - in the world of the Internet, an age. So the first challenge was to get the chinese characters to display. I had to save the file to a specific unicode format to enable the browser to cope with chinese characters.

In addition to that, I needed to include the meta tag, http-equiv=content-type content="text/html; charset=gb2312". This informs the browser which character set to use in order to render the text for the site.

So all in all something new for me. Of course it made me consider a whole range of cultural differences that could affect web design - the significance of colours, reading direction, conformity to user expectations etc. For this particular project, my input didn't move beyond creating a style that resembled a corporate site and so was pretty cosmetic, but it has piqued my curiosity for non-western website user conventions...