Should I design for Firefox?

How many are using it?

[Red_Eye]

I have been thinking about changing my site to pure CSS. So have been looking at my site stats in more detail to see which browsers vistors to my site use. I found the results quite surprising. The top 2 were IE6 (no surprise) and Netscape 7, which I thought was all but dead with the appearnce of Firefox. Less than 0.2% of visitors to my site use firefox. I realise that this is only the case for my site. But I was wondering what others have found and if it changes depending on site topic.

I like and use Firefox on a regular basis, and I was automatically going to make sure that my site displayed correctly in firefox, but my site stats suggest that I should also look more closly at other bowsers first (like netscape 7)

[Matt Probert]

You should design for web browsers. Period.

CSS is about suggesting appearance, it doesn't dictate. You should remember that and not think along lines of "displaying properly" or "inproperly".

Matt

[Red_Eye]

Matt,
Thanks for your reply could you expand on it a little. If I design a site to look a certain way then I want it to be displayed that way in all browsers. I am not talking about pixel perfect design, I'm not into that. But I know that not all browsers are equal when it comes to CSS positioning.

You should design for web browsers. Period.

Also do you mean "shouldn't"?

[Trace]

If you follow the standards, your pages should appear correctly in all browsers, eliminating your problem.

You might end up tweaking a couple of little things for IE, but they will be pretty minor.

[drhowarddrfine]

Just to emphasize what Trace said, you don't design for Firefox, you use FF for design. FF is much closer to standards compliance than IE is. IE is about five years behind and getting worse. So if you use FF to check your design/code then you know you at least got that right. Then look at it in IE. If it's messed up, then you can adjust your code somehow to make it work in IE. At least you know you did your part right and you won't develop any bad practices. So it doesn't matter how many people use FF vs IE. It's a matter of making your code right.

[encyclo]

Well you're in luck: Netscape 7 and Firefox share the same core rendering engine (Gecko), so if the site works well in Firefox it should work equally well in Netscape 7. NN7 is based on a slightly earlier version of Gecko so if you have a significant number of users with it you should download it for testing, but you're unlikely to have any major problems.

Just following standards doesn't mean that you don't need to test thoroughly in your users' favorite browsers.

[Red_Eye]

Thanks for the information that makes things much clearer.

[JAB Creations]

Almost. While Netscape and Firefox have the same rendering engine they are VERY different depending on versions. Bloatware version 8 has an early nightly build of the Gecko 1.8 engine before they even labeled it 1.7x+ in the nightly builds.

Netscape 6 and 7 are basically Gecko 1.0X each. There are a lot of early Gecko bugs and a ton that still linger all the way to the current 1.7X series. The 1.8 Gecko platform is superior to the 1.7 series. I would suggest downloading a Firefox Nightly Build, use a middle aged version of the Mozilla Suite (anywhere from 1.3-1.6), use Netscape 6 or 7 (in my case both) to test various versions of Gecko. While unless your code is indepth you should not notice anything to drastically different frmo version to version.

Keep in mind Gecko renders a page according to THE stanards while MS attemps to MAKE standards. While your page may look good in MSIE this would be like putting on a spectacular show and at the end getting fired because you either did it at the wrong place AND you did the wrong performence. MSIE will keep you from understanding this.

The only way to code is for a Gecko platform first. Then work your way to Opera. After and only after Gecko and Opera should you then test a page out with MSIE. Usually I baby step test sites at each new item added but you will have your own preferences. Keep in mind Safari (mac) and Konquerer (linux) are based on the KHTML engine though it has some mind bending bugs that are clear and present though have yet to be fixed.

[MatthewHSE]

You should design for web browsers. Period.

Also do you mean "shouldn't"?

No, I'm pretty sure he said what he meant. (Correct me if I'm wrong, Matt.) You should design in such a way that your pages will at least be readable in all web browsers, regardless of make, model, date, or user base. (That said, my site breaks terribly and illegibly in NN6.2; I don't know why for sure but it's a float clearing problem of some sort.)

If Firefox were the type of browser that required mega-hours to make a page that worked in it, (Matthew shrugs in the direction of IE) then I'd say your stats would justify ignoring the Firefox users. However, seeing as how Firefox is extremely easy to develop for, I would think that the time outlay necessary to make your pages work in it would be well worth it.

For one thing, if your site really breaks badly in Firefox, your low Firefox numbers could be something like a self-fulfilled prophecy.

All this, of course, assumes that your pages DO break in Firefox. You haven't said yet if they do or don't, but it should be a quick check. If the site works for recent versions of Netscape, it's a pretty safe prediction that it will work for Firefox, too.

[Red_Eye]

MatthewHSE,

Thanks for that I now understand the approach I should take. Currently my site doesn't break in Firefox, but at the moment I am mostly using tables for layout and css for style. Which is quite easy to get working in all browsers. However I am intersted in changing the layout and was thinking of changing to all css. But following this discussion I will be designing for firefox, as it seems to be a browser that supports the standards more closly than IE. So I was thinking of taking the approach that Trace suggested.

[tjhorne]

I have been in you're shoes and what I had the best success with was previewing all of my work in Firefox. Then there would always be some minor differences between IE and FF that I would have to work around.

And most of the time after doing that, I would review my work in Opera and about 99% of the time it was acceptable.

Let me know if you have any questions.