18 replies to this topic

[Adrian]

Every visitor counts.

I find it a bit iffy how some people talk about backwards compatibility, and recommending still supporting browsers such as NS4 as a way of making every visitor count.

Going off on a bit of a tangent, what about people with various impairments?
By sticking with tables, and supporting NS4, you're generally making it a lot harder to look after those visitors. If every visitor counts, then that means them as well. CSS based designs can be much more accessible (not necessarily, but they can be).

So how about the TENS OF MILLIONS of users worldwide who don't use NS4, but are given a rough ride because a designer would prefer to support the <1% of NS4 users, compared to the realtively higher percentage of impaired users.

Linking back to the topic about CSS and SEO. SE Bots are a little like visitors with visual impairments. Bots can't 'see' the page, and well structured code should generally be easier to read and sort out than stuff that is all over the place.

CSS based designs generally enable more semantic markup(again, depends on the code produced), which in theory should be easier for them to read and parse. Though I've seen no real proof of that, it does make sense, and helps in all sorts of other ways.

If you're going to claim every user counts, mean it, don't just look at the stats in your logs, think about the people using those browsers as well.

[qwerty]

I almost completely agree, Adrian. I'd certainly give higher priority to people with disabilities than those with archaic browsers, since the latter group is most likely smaller, and their situation is for the most part one they've chosen. While some of the people on NS4x are working in environments where they're not permitted to upgrade their browser, many of them could if they chose to.

But I don't think screen readers and spiders are really analogous. Spiders aren't concerned with the structure or the flow of a document. They're just grabbing words, so even if each word is in a separate table cell, it probably doesn't make much difference to them. I know, that's an exaggerration. It makes some difference, as the proximity of one word to another matters, but you get the point.

And if we're to believe what we've been reading about the next step forward, MSN is going to be segmenting pages based on layout -- that is, how they look in a browser, rather than how the code is set up, so even then they won't be looking at the content the same way a screen reader does.

[Adrian]

There are differences between Spiders and Screen Readers, certainly, but they are both going through the code in effectively the same way. If a table is properly setup, then yeah, a screen reader can get through it without too much confusion, and as a spider doesn't really need to understand it, it doesn't matter so much if it actually is confusing.

Still, if a site is inaccessible to people such as those using screen readers, it may well be that spiders are having some issues as well. I'd generally expect accessible code to be less problematic than inaccessible code.

Regardless of the similarities/differences, bottom line, don't use "every visitor counts" to justify supporting older browsers, if you don't also count the impaired visitors.

[Randy]

Regardless of the similarities/differences, bottom line, don't use "every visitor counts" to justify supporting older browsers, if you don't also count the impaired visitors.

Bravo on that one Adrian.

In fact, it's the exact argument I used to get an entire university system to upgrade everybody's browsers quite a few years ago.

They wanted NN4 (with no JS even) compatiable for their internal staff.

They also wanted disability compatiable because they are required to meet those standards by law by accepting Federal funds.

I told them to make a choice because doing both was not possible with the things they had in mind.

Of course it only took them 2 years and a 160 page reference paper to convince them that I was right.