Developing for tomorrows web Wednesday June 25, 2008

Posted in: Web development

1. Paul Cripps

   Jun 25, 11:59 AM

   It’s a really good point; I feel we should be building for tomorrows browser’s as it helps push things forward. That is the very nature of technology.
   
   If we always step into the past, making sure it works on every browser we’ll end up limiting ourselves and producing cheaper websites. If we take our designer thoughts off for a second I bet most of the time the “degraded experience” wouldn’t be noticed by your average user. As long as the site is legible and works as intended…
   
   I love “bleeding edge browsers” BTW!

2. Drew

   Jun 25, 11:59 AM

   I think the reality is that a website is only ever viewed with today’s browsers, but that the definition of what today’s browsers are is constantly shifting. IE6 is one of today’s browsers, because that’s what a bunch of users still browse with today.
   
   The real question is what does it mean to develop for only the newest browsers? Does that just mean not going the extra mile to cater to the bugs of recent less capable browsers like IE6? On the web the lowest common denominator is always text-only, so where does that fit in?
   
   I think Yahoo’s approach of Graded Browser Support is still a winner.

3. Robin

   Jun 25, 12:11 PM

   I was recently doing some updates for an in-house site and was adding some background images to the CSS. I wanted to just use PNGs, with lovely alpha transparency, but at the back of my mind I knew I would have to either add a workaround for IE6, or add styles to the IE6 stylesheet providing GIF alternatives. Checking out the Google Analytics for the site I saw that 70%-odd of all the site’s users were using IE, and of those, 40%-odd were still using IE6. So my point is, as much as developing for crummy browsers like IE6 is a pain in the backside, unless we want to alienate a fair amount of users, it’s still a necessity for a lot of sites

4. Dan Oliver

   Jun 25, 12:23 PM

   It’s something I’ve certainly been thinking more about recently and I’ve seen quite a few times in recent months where designers have produced websites where the user experience varies from browser to browser.
   
   With this is mind, I’ve certainly placed much less importance on Internet Explorer 6 in comparison to 6 months or so ago. A classic example is something as simple as transparency on PNGs. As your well aware IE6 just doesn’t want to know when it comes to transparent PNGs. Months ago this would have really bothered me, but a couple times now we have all sat down at work and decided IE6 users can be presented with a GIF of lesser quality. It sacrifices the user experience slightly, but it’s not significant enough to lose sleep over. As long as the really important things are more or less covered and the site looks something like we intended then we’re happy.
   
   If it was something that I felt seriously hindered the user’s experience I wouldn’t consider it, but nine times out of ten it isn’t. The chances are if someone is still using IE6 or an equally poor browser then they won’t be the kind of person that would pay so much attention to these smaller details or take the time to really appreciate attention to detail (there may be the odd exception, but in general…).
   
   As far as developing for the future goes, I’m using CSS3 more and more for small details just to enhance the design slightly for the capable browsers. In places I’ve also specified a none standard font, just in case someone out there might have it installed on their system. And if they do, they are going to get a jolly nice looking webpage. These smaller details are just nice to haves really. It’s frustrating we can’t have it across all browsers, but we can’t.
   
   So I could cut all this rambling short and just say I can live with offering older browsers a degraded experience providing it’s not significantly degraded. If the design is near enough how it should look, I can live with it it.

5. Nathan Pitman

   Jun 25, 01:18 PM

   @Drew: Thanks for drawing my attention to the Yahoo! ‘Graded Browser Support’ initiative. Obviously it’s difficult for a small web agency to cost effectively test against all those scenarios, even with tools like Litmus but it certainly makes me think that as a company we should have our own ‘Graded Browser Support’ table.
   
   @Dan: Yes I think this is exactly what I was trying to get at, though you perhaps summed it up better. Interesting too to hear that your current experience is also very similar, opting to live with a reduced experience for IE6 users with regards to PNGs in particular.

6. Raanan Avidor

   Jun 26, 04:39 AM

   It is not an or question. It is not building for today or tomorrow. We must build for today and tomorrow. Building for IE6 and FireFox 3. We also have to explain this to our clients. The job is hard and this is why we get the big money :)

7. Matt Robin

   Jun 27, 08:23 AM

   My current view seems to be a mix of (the same sort of view that Drew expressed, above) and that man Andy Clarke – embracing future browsers willingly and being less accommodating for those older browsers (IE6 – I’m looking at you!) to some extent.
   
   Actually, my support for older browser versions was more generous, but recently – with new versions (IE7 and soon IE8, and FF3) rolling-out at a quicker rate than before, I think it is reasonable to expect that most users are upgrading more often than they did in the past.
   
   Also, and this might just be wishful thinking on my part (haha!) – the range of differences between versions seems to be getting less pronounced with each new one, so making a new site to work with IE7 and IE8…mmm, bad example….how about, err, FF2 and FF3 – should be relatively minimal compared to trying to make a design work exactly the same with IE6 (and it’s poor handling of PNGs). Web design support for IE6 will soon get as thin as it did for IE5 – maybe even by the end of the year! In December, we’ll be saying “Who the hell still uses IE6?!” – at the moment, I think there are still enough people using though, sadly.