72 replies to this topic

[Craig B]

IMO, using a hybrid design is perfectly acceptable until the older browsers disappear (and good riddance!). There is no reason to lock out millions of users just to prove that a site can be coded in XHTML (strict for instance). Sure, some of the pages may not be 'forward compatible' but I would ask - How many web pages dating 5+ years are still around today? Is the trade-off worth locking out paying customers today?

-Craig

[Ron Carnell]

And you both also see no problem with using tables for layout? Thats a hack in itself.

Not by my definition.

Using tables for layout is semantically questionable, but it's not a hack because, for all practical purposes, every single browser since 3.x renders them identically.

At the core of every hack is an implied (or forced) if-then clause, essentially turning a markup language into a programming language. The use of tables presents a different set of problems, which I agree are not trivial, but it's still a sequential instruction set. Not a hack, in my book.

Most importantly, I can be fairly certain that Opera 19.1 will still render my tables as expected.

You both say you don't like to use hacks and yet you both continue to support NS4.*

Yep. I set my font properties in the body style, and if NS doesn't understand that (and we all know it doesn't), I know the site will still be very usable with a default font. For most sites (not all), I refuse to use any CSS-2 positioning that will make the site unusable in a non-compliant browser. I'm willing to accept degradation of my design. I won't accept corruption of it.

Again, I have no problems with people who decide otherwise. We all have different goals and most have different audiences. I just ignore the crusaders who insist their goals should be mine.

Works for me.

[Ruud]

Hm, I'm out of my league here: I'm not doing any site that sells a product. Still... On my sites about 2% of visitors use Netscape. On a large number of visitors that means a reasonable number of people. Say you have 10 thousand unique visitors per month it makes for 200 Netscapers. Ah money! ... Well, no, the way that I read it not every visitors is a customer, right? So if 10% of those visitors become customers you've worked on code for 20 people.... Dunno if that is considered a good ROI or not.

Apart from that - to each his own I guess :-)

Ruud

[mcanerin]

You know what, though? I'd like to see a breakdown on which browser users are buyers, and of what.

For example, if it turns out that 90% of NS4 users are buyers, then the tune in this thread would change pretty fast, I think. Likewise, if it could be shown that people who use NS4 are cavemen who can't connect securely to anything and don't buy anything, that might influence things, as well.

For a while now I've had the feeling that MSN searchers tend to buy differently from Google searchers, just due to the type of people who like portals (or don't know how to switch default pages) compared to experienced searchers. I imagine the same concept could be extended to Browser users.

Would an Opera or Mozilla user like different things and buy in different ways than an IE user? Are IE 5 users more likely to be corporate than IE6 users? Is someone on a Mac more likely to be impressed by a flash site than someone on a linux box?

I don't think this kind of thing is unrelated. When people make choices other than the default or popular choice, they are showing their personality and interests. And you can probably make some assumptions about people who accept the default "out of the box" system settings as well.

I'm not saying that you can get a complete profile on someone by their browser/OS choices, but I think whenever you make a choice you tell people a little about yourself.

I get the feeling that NS4 users are likely to not be buyers of "home" or "personal" stuff (since they are probably at a highly restrictive workplace) but might be researching large tech purchases. You might want to think about that when you include or exclude the browser - you are also including or excluding the choices the user has made - for good and bad.

Ian

[Randy]

I guess I fall a bit into both camps.

For my own sites, I have not bothered to worry about NN4.x support for a looooong time. I used to be a bit mean to them even; Detecting if is was an old NN browser and redirecting them off to another page that only had a link to Netscape's download area and a message that they really needed to join the 20th century and upgrade.

But I'm not as onery anymore. Now I just give my sites a brief once-over to make sure they can at least be navigated in older browsers and move on. I really do get very, very few old Netscape visitors these days. All of them that are seem to be coming from Corporations that need to upgrade their internal systems. So I guess for a B2B site it would be important to build in some NN4.x support.

Now when I used to design for others, I always made the sites NN4.x compliant. But that's been a couple of years ago. I still get calls and email from other designers --especially newer ones-- every now and then trying to work out how to do this or that in a way that will be NN4.x compliant. But I don't design for others anymore so I don't force myself to jump through those hoops.

I do however sometimes use tables as containers just to keep from having to do browser detection. It's a cheap trick maybe, but I don't ever see tables being unsupported by the browsers. At least not until after I've retired.

[bobsledbob]

If anyone wants to see a decent use of CSS 2, take a look at www.mozilla.org. This is a layout done without tables which degrades decently for older browsers such as NN4. I particularly like the use of the media="screen" and media="print" in the link tag for the stylesheet, something that I'm looking forward to using in my sites.

Personally, I'm in the camp of using html 4.01 transitional with CSS Level 1. This is the only way to avoid the "browser war hacks," as Ron Carnell put it, and keep the design elements in one common stylesheet.

[Kev]

Tables as hacks- W3C state that tables shouldn't be used for layout purposes. Thus layout tables by definition are hacks. Not only are layout tables hacks, they introduce barriers to people whose primary means of navigation is not sighted. Try linearizing all your layout tables and see what your site looks like. Thats how someone with a visual disability 'sees' it.

XHTML locking people out- simply not the case. A site coded to XHTML standards will be open to many more visitors than a site coded to cater to a specific browser. Because a site is coded to standards it means that not only are *all* modern browsers (IE5.5 +, Opera 6+, Mozilla 1+, Firebird, Safari, Konqueror, Galleon, Lynx, Netscape 7 to name but a few) are capable of seeing your site as you intended with a few (and I mean one, possibly two) one or two line hacks, it also makes your site much more accessible to disabled people. I don't know about the figures in the US, but over 8million people in the UK are classed as disabled and having the necessary purchasing power to spend online- thats a fairly significant audience.

On the other hand, if I code a site to NN4.* standards then I can only be sure that users of this browser are seeing it as I intend. To make my content available to other non-NS4.* users I have to make extensive hacks- javascript browser detection, javascript version detection, working CSS- probably a whole new site. I also immediately lock out all my disabled customers.

To me its obvious- I can code a site that adheres to standards and attract a large audience with one single site, or I can attract a small audience with a few seperate sites with the same content.

Standards based design doesn't lock people out, it includes them- coding for one specific browser on the other hand definitely excludes them.

I've mentioned accessibility a few times. It can't be ignored any more. In the US you have the 508 legislation which caters (I believe) for governmental sites. This will almost certainly be extended to business sites. Here in the UK we have the DDA (Disability Discrimination Act) which already does do this.

The essence of both bits of law are the same- you have to make your sites accessible. Several groups have already been taken to court and successfully prosecuted including the Olympic Comittee.

[Adrian]

LOL, Ron, I hope you're chuckling at the concept of someone suggesting I'm not a standards advocate, I know we've discussed the various merits and issue of support for NS4 before.

bwgroup, perhaps you'd care to look at my own website (the Camaban one in my sig) and see what its like. Also have a go trying it in NS4 and perhaps have a skim through the rant about web standards and semantics.

I'm a big standards advocate, I would recommend people go for standards whenever possible, but being realistic, that is not always possible. In some markets NS still seems to hang on at a decent enough percentage to warrant addressing.

I'm very aware of the idea of forwards compatibility, I've ranted about it elsewhere before as well and read what Jeffrey Zeldman has said about it, and yes, where possible I believe people should go for it, but again, it comes back to target markets!

You are not going to have a successful site if you don't take account of the people who are visiting. In MOST cases now, you're able to look at IE5/5.5+ which gives you a lot more room for manouvre.

If you ignore NS4, you're not aiming at 100% of the web at all, how can you be if it breaks in NS4?

As you've mentioned accessibility, its worth commenting on. Yes its law, yes you must make reasonable attempts to make you're site accessible to impaired users. If you're target audience is going to include impaired visitors, then yes, you have to take that into account aswell. If a greater proportion of your visitors are impaired in some fashion than using NS4, then place more priority on making the place accessible.

Once again, knowing your market is a major foundation of a successful site.

[Kev]

bwgroup, perhaps you'd care to look at my own website (the Camaban one in my sig) and see what its like.  Also have a go trying it in NS4 and perhaps have a skim through the rant about web standards and semantics.

I'm not sure what you intended it to look like so its hard to offer valid opinions but IMO, the grey 'designed by adrian lee' text overlaps from the blue sidebar to the main content area by about 4-5 pixels.

The section beginning "The site code is an example of the kind of thing I do..." overlaps the right edge of the rest of the site significantly.

On the RSS feeds page, the Mezzoblue link sits under the other 3 on the left as does the sitepoint link on the right. Again the right hand content overlaps the implied border of the rest of the page (drawing an imaginary line down from 'skip navigation' passed the horizontal rule)

On the 'who am I' page the same errors with the right hand content repeat themselves (you also have basic HTML rendering issues "first 'BIG' Project. ")

This is Mozilla 1.6 on Win2K BTW. I don't have NS4 installed on any machine I own so I'm unable to oblige you there.

In IE6 the site look even worse. All the problems above are present as well as broken padding issues on the body tag, preternatural shortening of the blue sidebar and falied rollovers on the link boxes.

If you ignore NS4, you're not aiming at 100% of the web at all, how can you be if it breaks in NS4?

True. I'll settle for 99.9999% of the web then

Once again, knowing your market is a major foundation of a successful site.

You've fallen into the trap of thinking of disabled people as a market, they're not. Some belong to one market, some to another, some to yet others still. If you're building a commerce site you don't create barriers to entry where its easy not to do so.

[Adrian]

OK, so you've managed to pick out mostly 'problems' which aren't problems, just the design. I take it you didn't care to look at the code, which is the main point of this discussion. Tableless layout and things like the nav bar is made up of a styled, unordered list. That kind of thing. I may have made one or 2 changes rceently that mean it doesn't quite vaildate fully, but it does generally.

2 of the 3 additional problems you've picked up on in IE6 are IE6 issues. IE does not support :hover style on anything other than <a> tags, hence the rollover issue on a <div>, plus theres the already commented on problem with height of columns in IE not working like it does in most other modern browsers.

You also missed that in most browsers, the nav bar is fixed, in IE it scrolls due to its lack of support for position:fixed.

None of those problems break the site, people using IE6 can see and use it perfectly well, it just looks slightly different and doesn't have a couple of the etxra bells and whistles.

I'm happy to live with that, I don't care if the site looks slightly different in different browsers, as long as it still works and looks generally right. Some of its like a little extra bonus for people using more advanced browsers.

For reference, the site completely break in NS4, its completely unusable because everything overlaps. I could probably using the @import trick to give it just some of the styles to make it usable, but I don't want to. Its my personal site, its supposed to provide some content and be used as a bit of a testing ground for my own learning.

As far as impaired users are concerned, you do still need to think about how many people are actually going to be visiting with the various kinds of problems. The DDA defines that you should make reasonable effort to make a web site accessible. How can you define what is reasonable if you have no idea what the usage amoung impaired users is?

If you're someone like the RNIB you know you're going to need to make a fair amount of effort towards visually impaired users. But what about physically impaired? or people with mental problems? Different features and techniques for different impairements. Whats reasonable for the people trying to access your site?

It brings up one of you're earlier points, in order to be accessible, one of the points is to use more up-to-date technology. In this sense that suggests going XHTML instead of plain HTML.

Its fine idealising about doing it all fully to the right standards, but in the real life scenario of some businesses, a compromise has to be drawn.

[Kev]

[quote]OK, so you've managed to pick out mostly 'problems' which aren't problems, just the design.  I take it you didn't care to look at the code, which is the main point of this discussion.  Tableless layout and things like the nav bar is made up of a styled, unordered list.  That kind of thing.  I may have made one or 2 changes rceently that mean it doesn't quite vaildate fully, but it does generally.[/quote]

Well, to me, the design of a site is equally as important as the rest of it. I like my designs to be exactly the same in all browsers. Designers generally come up with a design to carry the brand forward- I've worked on sites for Local Education Authorities, Disney, Ideal Standard, Nat West and Jarvis over the years and I can't imagine any of them bring particularly happy with a site that doesn't look and act the same across as many browsers as possible.

I did indeed look at the code and whilst its standards compliant in most places it needs adjusting to take account of browsewrs quirks- like the <a> only rollover- easy to fix.

[quote]You also missed that in most browsers, the nav bar is fixed, in IE it scrolls due to its lack of support for position:fixed.[/quote]

I have my resolution frighteningly high on the page I opened up using IE I dind't need to scroll hence didn't get the issue you describe.

[quote]None of those problems break the site, people using IE6 can see and use it perfectly well, it just looks slightly different and doesn't have a couple of the etxra bells and whistles.[/quote]

Its usable yes, but is anyone going to want to stick around on a site that looks broken in the worlds most popular browser? I don't think so.

[quote]Some of its like a little extra bonus for people using more advanced browsers.[/quote]

And yet with a little extra work (and I do mean a little) all your users could access these advanced bonus features.

[quote]How can you define what is reasonable if you have no idea what the usage amoung impaired users is?[/quote]

You research, do user testing and find out. For a personal site thats obviously overkill so coding to AA (Priority 2 standards) is my bottom line.

[quote]Different features and techniques for different impairements.  Whats reasonable for the people trying to access your site?[/quote]

Very good point and one the DDA fails to adequatley adress IMO. Flash for instance is lambasted unmercifully as inaccessible but for a user with a cognitive based disorder, a picture based system could be ideal.

[quote]Its fine idealising about doing it all fully to the right standards, but in the real life scenario of some businesses, a compromise has to be drawn.[/quote]

I don't agree. There should never be a design compromise. Coding to standards increases your freedom, not hampers it.

[Craig B]

I work on a site that still records over 100,000 visits per month of Netscape 4.7x users in the server logs. I guess I used a confusing phrase when I mentioned that strict XHMTL 'locks users out' in my previous post. What I meant was that user experience (look & feel, copy, navigation, etc) cannot (or should not) be negotiated with the client just to satisfy a developer's instinct to use cutting edge mark up.

On my personal site (iappeal.ca), I do use CSS layout exclusively as I have the freedom to do so. At the same time, I do not agree with some of the W3 recommendations. For example:

target="_blank" - this is retarded. Why should we have to use a javascript hack to open a new window (to validate)?

div height attribute (for columns) - why can't we set a div to 100%? I have to use javascript to find the longest column and then make the others the same height.

What I am getting at is that it is a little too early to adopt the complete W3 recommendations unless you have the freedom to do so.

-Craig

Edited by Craig B, 06 February 2004 - 01:11 PM.

[niceguyeddie]

target="_blank" - this is retarded. Why should we have to use a javascript hack to open a new window (to validate)?

Why should a designer have control over the fact that I want to open a new window or tab just to visit a link? At least you having to use javascript gives me the option to turn it off and browse in a way that is usable to me.

[Kev]

What I meant was that user experience (look & feel, copy, navigation, etc) cannot (or should not) be negotiated with the client just to satisfy a developer's instinct to use cutting edge mark up.

Not 100% sure what you mean there. If you mean that using compliant code (and compliancy is hardly cutting edge- its been around since the web!) means you have to compromise your design then you're incorrect.

target="_blank"  - this is retarded.  Why should we have to use a javascript hack to open a new window (to validate)?

Well, as niceguyeddie states- you shouldn't have control over how a visitors browsers core functionality works. And anyway- it is possible: www.juicystudio.com/tutorial/xhtml/module.asp

div height attribute (for columns) -  why can't we set a div to 100%?  I have to use javascript to find the longest column and then make the others the same height.

You can: www.alistapart.com/articles/fauxcolumns/

it is a little too early to adopt the complete W3 recommendations unless you have the freedom to do so.

You have the freedom to right now. You may have got the impresion from my non-supporting of NS4 that its impossible to support that browser with a standards based set of code. It isn't. My arguement is that catering for that browser is hampering the adoption of stanards and that the clients you'd gain by making a site open to everyone instead of NS4 users would vastly outweigh the clients you'd lose.

[niceguyeddie]

You have the freedom to right now. You may have got the impresion from my non-supporting of NS4 that its impossible to support that browser with a standards based set of code. It isn't. My arguement is that catering for that browser is hampering the adoption of stanards and that the clients you'd gain by making a site open to everyone instead of NS4 users would vastly outweigh the clients you'd lose.

I agree completely here. The longer that we find and use work arounds for buggy browsers rather than actually work towards standards, the longer the process is cycular. At least right now I can almost assure myself that I will look the same on Moz and other recent browsers by hitting xhtml strict. It's less development time for me in the end, and greater satisfaction for the user at the same time.