However, as Zeldman notes:

No hammer fits all nails, and no solution, however elegant, will work for every situation. But if we’re open minded, Andy’s proposal may work in more situations than we at first suspect.

Remember NS4?

This technique is very similar to what we used to do for Netscape 4 users when use of that browser started to tail off. We would serve our CSS using @import, knowing that Netscape would ignore that command and either get no CSS at all or a simpler stylesheet served with <link>.The Universal stylesheet for IE6 is a slight variation on the same technique.

As in the NS4 days, if and when to go this route depends on the situation. Of course you’re not going to do something like that if 30% of your visitors are using that browser. But what if it’s 6%? Or 4%? You wouldn’t choose this approach if branding an image were an important part of your site, but if the content is the focus (as Andy suggests), then maybe this is an appropriate option to consider.

Is IE6 worse than NS4?

Does anyone remember how much agony we went through in the last days of Netscape 4?  I’m starting to think that IE6 is dragging along longer than Netscape 4 did, and is harder to get rid of. With Netscape 4 there seemed to be fewer stubborn and pervasive obstacles to upgrading.

They at least need to know

I strongly believe that users of older browsers, where possible, should at least be informed that they are using an outdated browser. This situation is never going to change if users don’t even know that there’s a problem (since, you know, Microsoft is doing such a great job of explaining that). Even if their only recourse is to go to their IT department and complain.

This, again, is a matter of appropriateness. Of course I’m not going to put a message on the UW home page telling IE6 users to upgrade. But on the web development site? That would be appropriate.

Then I read this:

HTML is easy — so easy that even a first-time user can do it. That’s because you don’t have to learn HTML to use it.

Leading word processing applications, such as Microsoft Word and AppleWorks 6, actually generate HTML webpages for you with just a few clicks of a mouse.

Noooooooo….

* link only available while my work computer is on, which is only when I’m at work, or you can try this one.

I brought this up over a year ago and this time, others are talking about it. Jeff Croft says that “Your markup validator, whether it’s the one on the W3C site or one built into your favorite coding tool, is not a measuring stick for greatness.”

It’s like a spell checker … but not

This is an oft made comparison: the validator is like a spell checker. You wouldn’t turn in a school assignment or publish a print document without spell checking would you? Right. You’re also not going to put a sticker on it to say that you passed the spell checker.

And, of course, we all know that the spell checker can’t catch your grammar mistakes or misuse of words blah blah blah, you know what I mean here.

On the other hand, it might not make that much of a difference if you’ve got a few validation errors. In fact, most of the time it won’t. This page might have some validation errors, I actually don’t know if it does or not, I haven’t checked lately. It would be pretty dumb to spend your time going around checking validation on your sites all the time. Although I must admit, i do have the HTML validator extension installed in Firefox. When I use Firefox (which is only for testing…).

It’s not that important to me

Really, it’s not. It may make me slightly uncomfortable, and I may try to fix things if its convenient. But most of the time I’ve got better things to do. It’s not the end of the world if a page isn’t valid.

Suggesting validation to others

In the comments on Croft’s article people are talking about suggesting validation to others in forums. We do tend to do this sometimes at TWF but it’s often because the code is so bad that it’s really difficult to pick out what the problem might be. We may also suggest that people run the validator if they’re having a lot of display issues. We would never just say “run a validator” and leave it at that though. We would explain that if the code is correct it can often help to straighten out display problems etc.

Web Standards != Validation

The validator also doesn’t cover all of the many facets of standards compliance. Web standards are much more than validation alone. To me, it’s most important to follow standards in this sense (using proper semantics, separating presentation from content, minimizing extraneous code etc.). I’m sure you’ve seen the examples of really bad code that passes the validator.

And you?

How often do you validate? Do you obsess about it? Does it take precedence over other aspects of web development?

Edit: My apologies if anyone attempted to comment on this post and wasn’t able to. I upgraded my wordpress and it was doing some weird things.

Here’s something I don’t say often: Suck It Up. If you work on more than a few projects, they can’t all smell like today’s fresh linens. It doesn’t mean you’re a bad programmer. It simply means that you’re prioritizing.

And I think that’s something that goes for any type of project. There’s rarely enough time or money to make everything perfect, whether it be a programming project at work or the paint job in the dining room. You have to find the point of balance and decide when it’s worth it to put in the extra time and/or money. Does this really matter in the big picture?

This can be a difficult thing to get over and, as a perfectionist, I know I often spend time on things that don’t matter that much.

Opinions I like

• Bobbing Heads and the IE8 Meta Tag (Burning Bird)
• They shoot browsers (Jeremy Keith for ALA)
• Doctype switching for IE 8 (Roger Johansson)
• Has Internet Explorer Just Shot Itself in the Foot? (Andy Budd)

And the TWF discussion thread.

Don’t piss off the clients!

I understand Microsoft’s point on this. I’ve been there, in a modest way. You don’t want to do anything that makes your clients angry, especially if they’re going to make a lot of noise about it. Apparently this happened with IE 7 and the manageria at MS don’t want to go through that again. I understand. It sucks to take a beating like that.

But, on the other hand, MS has been taking a lot of pains to kiss up to developers lately. What about that? Well, in this case they went straight to the most respected web developers site and got them to say some nice things about it.

What exactly is going to fall apart so badly?

They don’t even have an alpha version out yet. IE 7 fixed most of the worst CSS bugs. And even if something was a little different it certainly wouldn’t cause sites to become completely unusable. Scripting I honestly don’t know that much about. When I first read about this issue I was under the impression that it was just CSS that was going to be impacted but Zeldman says otherwise.

Jeremy Keith expands more on this in his ALA article. I agree that it’s strange to talk about doing all this before anyone has an idea of how severe the damage would be.

And even so, if it’s just CSS or just scripting or either, wouldn’t it be better to target that separately? So, to have an option to say that this site uses the IE 7 scripting mode?

I also strongly agree that if they’re going to be the meta tag it should work in the opposite way. That is, the meta tag could be added if you want your site to stay in IE 7 mode and not the other way around.

At TWF Gary brought up the problem with active X controls which I haven’t seen talked about too much elsewhere. This may be where the real problem lies. So why do we need to break CSS to allow these old active X controls to keep working?

I’m still a bit confused about this because even the MS announcement and 3 out of 4 of the ALA articles are mostly talking about CSS.

Quotes and notes

Jeremy Keith says:

This is gobsmackingly audacious.

Right on!

Eric Meyer says:

Developers have been forced to conform to past browsers’ behaviors while making educated guesses about what future browsers would do.

But we do know what future browsers will do. It’s in the spec. CSS 1 is CSS 1 whether it’s 2006 or 2015. I can’t recall ever having to make a change to a site when a new version of Opera or Firefox came out.

Meyer also says that:

Forward-compatible development and its cousin, progressive enhancement, were necessary and proper because they were the only hope we had of sites continuing to work into the future.

The only hope? More like the correct assumption. I develop under HTML 4 and CSS 2 knowing it will work forever. And this is where it gets scary. If browsers assume that developers are going to lock in to a specific version of their browser then they don’t have to try to implement the specs correctly, with forwards compatibility in mind. In fact, they can toy around with whatever they want, knowing that developers are tied to their versions, not to any standard. Before you know it we’re back in the 90′s when different browsers had completely different implementations and their own proprietary tags.

Aaron Gustafson recommends not using the IE8-edge modifier because “even Eric Meyer can’t predict layout or scripting bugs that may be accidentally introduced by a new browser version.” Oh, come on. You mean browser makers are just going to change things that are already implemented correctly? Unlikely. Okay, so I’d be unlikely to use that on a really mainstram website (like my work sites), but I wouldn’t hesitate on others. Although, on really experimental sites I’d be more likely to just leave it out altogether to prove a point (IE sux0rs!).

Coding with “reckless abandon”? What? Have we suddenly forgotten how standards work?

Some alternatives

There are many other things that MS could have done to mitigate this:

• Do another level of doctype switching (Roger Johansson has suggested this and I thought of it as well. Not ideal but certainly better than a meta tag)
• Release IE 7 as a standalone, so anyone dependent on apps that only work in 7 can continue to use them until the apps are upgraded to v. 8
• Work with site developers to help them upgrade their sites to IE 8. I think this would be a great move – Opera does this all the time. I don’t recall seeing anything around the IE 7 release from MS detailing the problems that developers may encounter and how to fix them. Providing development support would possibly be more cost effective than attempting to build a browser that perpetually supports multiple rendering modes. Wouldn’t their clients love it if Microsoft paid some developers to fix their sites for them?
• Release more betas, sooner and more often so developers can test further ahead of time and bugs can be fixed faster. In addition, educate developers about what will be changing in IE 8 as soon as possible so appropriate changes can be made.
• As brought up by Gary, announce an end of life date for IE 7, say, 5 years ahead, that would give developers plenty of time to move old applications to standards.

Does Microsoft want sites to only work in their browser?

Liam and I were talking about this one night and it does make sense. Applications that are tied in to using IE only forever are good for Microsoft. They probably really like stuff like that because it forces a whole lot of people to use IE, which means they’re also using Windows.

So really, they don’t want developers to use standards, they want them to do stuff the IE-only way. This meta tag obscures proper standards in IE — they’re only there if you know how to access them. Which means more people can continue on making stuff that only works in IE. Just a conspiracy theory

Liam writes:

My bet is they’re trying to avoid standards compliance — retaining the locked-in status quo — whilst still appearing to be doing the right thing for PR purposes.

When looked at in this way things start making much more sense: people can carry on writing non-standard code, working in the dark of vendor lockin. It also makes coding with standards confusing for beginners (bonus point for Microsoft!).

If you haven’t already you should read some of Liam’s comments about OOXML. This is something web developers seem to be largely unaware of but it says a lot about how MS really feels about open standards despite their newfound “commitment” to web standards.

If it was just temporary

I wouldn’t have such a huge problem with it if this was a one time thing. Okay, you can lock into IE 7 or IE 8 rendering if you want to. But having to put one in for every site for every browser version for ever and ever? Nuh-uh.

Really, it’s just wrong

Standards are standards are standards. They work now, they work next year, they work in 10 years. Always the same way. You don’t have to say which browser you’re targetting because you’ve already said which version of html/xhtml you’re using. CSS doesn’t require version targetting since it’s always forwards compatible.

The web standards movement started in 2001 (if not earlier). That was 7 years ago. If you’re still not developing to standards then you just haven’t been paying attention.

Purpose of a browser upgrade campaign

The purpose of a browser upgrade campaign is to:

1. tell people they’re using an outdated browser
2. explain why this is a problem (optional but effective)
3. suggest alternatives

Why do we need to do this?

Well, obviously, browsers release new versions and old versions become outdated. Many people do not realize they’re using the outdated version, causing problems for web developers and their own security.

Sure, some of these people are stubbornly clinging to their old favourites. Others have (somewhat) valid reasons for continuing to use the old version. But many users won’t know they’re even using an old browser if nobody tells them. They may ignore the first message, but what about the second, or the third? The fourth?

I figure if people keep seeing these messages they might actually do something about it. I also think that we have a bit of an obligation as web developers to educate people about browsers. We can’t just sit around and complain that 20% of visitors are still using IE 6 without doing anything about it.

The examples

IE from Microsoft

First we have IE’s own v. 7 upgrade campaign (right, click to enlarge). This came up one day when I loaded IE6 to check something – I’m not sure if they only show it once or if it would continue to be displayed every so often.

The first problem with this it doesn’t explicitly tell you that you’re using an old version. It sort of implies that this IE 7 is something new. But does the target audience even know which version they’re using? Probably not. It does explain the benefits of the new version, but doesn’t explicitly tell the viewer that they should download it. Come on, these are people who still haven’t figured that out yet. You need to be more clear.

It also doesn’t explain that there is any problem with continuing with the old version. The page is designed as marketese that would probably cause most users to click away. It’s displayed as a full page rather than any kind of alert, which makes it seem like a marketing campaign rather than an error message.

Facebook

Facebook does a much better job of explaining the browser upgrade problem:

This comes up on your Facebook home page, at the top. It’s styled a little different than the usual Facebook content and is clearly an error message. It accomplishes all three of the points I mentioned above:

1. it tells you there’s a problem
2. it explains that this might cause problems with your use of the site (without going into too much detail)
3. it suggests several alternatives

TWF/A Padded Cell

We’ve put up a similar warning for IE 6 users at TWF and A Padded Cell:

I can’t say I was directly responsible for this (Liam put it in) but it does accomplish the three main objectives. And, as you can see, it is obvious that IE 6 isn’t able to render the site correctly (there’s supposed to be a logo there!). We’ve set this up as an ad and so far it’s been seen 6700 times and gets an 8.48% click through rate which is way higher than our other ads. We get about 20% of our traffic from IE 6 which is higher than the 17% we get from IE 7!

And now for something different: Netscape

I installed Netscape the other day after getting a report of problems with the site I maintain at work. This is what came up when I launched the browser:

Same thing, it explains there’s a problem and suggests alternatives. I think they could be a little more clear about why users shouldn’t continue to use Netscape. For example, there will be no additional security upgrades which may make your computer vulnerable.

I find it really interesting that they’re recommending Flock above Firefox. Liam and I tried Flock once out of curiosity. We found it to be unbelievably cluttered and filled wth stuff most people wouldn’t need. But, then, that’s the Netscape way, isn’t it? Flock is targeted to people who are really into Web 2 sharing sites, which sort of explains the interface, but also makes it even more confusing as to why they’re suggesting it to the Netscape crowd. I somehow doubt that people still stuck on Netscape are really into Web 2 content sharing!

One thing none of these campaigns do is explain that downloading and installing a new browser is nothing to worry about. I think some people might be afraid of installing new software, especially if what they have seems to work for them or think that the process would be too difficult, or they might mess it up somehow.

More examples

Have you seen any other examples of current browser upgrade campaigns? Please post!

You just need to make a few small changes. But the CSS is such a mess! It would be so much better if you could just re-write it. Cut down on the crud and make it all nice and clean and elegeant. Take it all down like Mike Holmes.

But there’s not enough time for that. The changes are due by tomorrow so you have to just make do with what’s there and add in your own necessarily crappy code.

Sigh… this happens to me all the time these days. And I’m sitting here thinking, should I just fix it? How long would it take? There’s a little brick wall here — do I just climb over or do I start again?

Granted, this code isn’t *that* bad. It’s relatively standards-ey (circa 2004-ish). Just lots of layers of disorganized CSS a bunch of IE5 hacks. It still bugs me though.

I was inspired to come up with my list after looking at the font stacks on Joe Clark’s site. Check out his typography stylesheet for some interesting ideas.

I find myself using the list I from my A Padded Cell article all the time now. Just today I added a few more options as a back-up for a heading that’s already replaced by sIFR. From the way I have it’s easy to pick one font from each column, although I find Rutter’s chart to be a lot cleaner and better as a quick reference (and he’s got Adobe fonts included). I thought a lot about the best way to display that information, especially since I wanted to include information about which fonts are near equivalents of each other, as originated here.

In an upcoming update to A Padded Cell I’m using this font stack:

"Lucida Grande","Lucida Sans Unicode","Eras Medium ITC","Mg Open Moderna",Helvetica,Verdana,sans-serif

It’s working well so far but we need to do some more thorough testing before final release. I’ve also used that list to add some Linux alternatives to the font stack in various stylesheets I’ve come across at work.

If you’re finding my list to be useful I’d love to hear about it!

The Web has become essential to our everyday lives. We can’t let one company, Microsoft, control its innovation. This is a fight to protect the Web’s future.

I hope that as developers and Web users you will all show support for this cause. This is not Opera vs. Microsoft, but rather the continued innovation of the Web vs. Microsoft.

The complaint was filed with the European Comission and essentially requests that Microsoft

1. implement web standards properly
2. detach IE from the Windows operating system or provide other browsers by default

• Overview from OperaWatch
• Press release
• Open Letter from Håkon Wium Lie

Div-itis: What it is and How to Avoid it

Sorry for the lack of posting lately! I’ve had some other things on the go but that’s (mostly) over now so I should be back to regular posting again soon.