110. The mighty Meyer

Paul Boag

On Show 110: Eric Meyer on version targeting, the business benefits of usability testing and do I have to have a blog?

Download this show.

Launch our podcast player

News and events | The profit and loss of usability | Eric Meyer on version targeting | Listener emails

News and events

Before we kick off the news section today I should point out that I actually link to many more articles, tutorials and news stories than I ever include on the show. If you go to boagworld and click the links in the header you can receive all of these posts by either email or RSS.

CSS reference library

Raise your hands if you have ever used the W3 School. Okay put your hand down now because you are looking kind of stupid. For those of you who didn’t raise your hands the W3 School is a comprehensive reference sources for almost every web language around. It really is very impressive.

The problem is that it doesn’t provide the nicest user experience. If you are looking for CSS reference information then a possible alternative is the new Sitepoint reference section. At the moment it only includes information on CSS however they are obviously looking to expand.

Its design is a lot more attractive than W3 School and although it is not as comprehensive it will provide you with all the latest information on CSS properties including CSS 3. It also provides you with an indication of which browsers support that property.

Stay on :target

Talking of CSS 3 there is an interesting article by Brian Suda over at Think Vitamin entitled Stay on :target. Target is a new CSS pseudo-class being introduced in CSS 3. Although it has yet to be implemented by Internet Explorer it is supported by many other browsers.

Essentially what target does is allow you to style a target element in much the same way you can style on hover. So when you click on a link that goes to a fragment identifier (what used to be called anchor links) then it styles that fragment.

So for example a good use of this attribute would be to highlight the content that you have just been jumped to (like the fade to yellow technique) or to hide and show content when the user clicks on a tab.

Of course many of you might be asking why I bring this up when it isn’t supported in IE. Well, not only is it an interesting glimpse into what is to come, it can also be used right now when the functionality provided isn’t crucial. For example you wouldn’t want it to hide and show content but it might be okay to highlight linked content with it.

Read the article and you can see the potential yourself.

Search behaviour patterns

Another article worth reading is Search behaviour principles over at Boxes and Arrows. This article looks at how people search and the different types of searchers there are. It also looks at what things affect the way people search and techniques that can be used to accommodate their approaches.

As I have said before on this show, not enough attention is given to search and this article will really get you thinking about how to better accommodate it. Best of all there are still a lot of hints that can be applied here even if you are lumbered with an inflexible search mechanism that cannot be customised or tweaked.

10 principles of effective web design

Talking of search usability brings us nicely on to our final story of the day which is a post from Smashing magazine. We haven’t mentioned them in a while but this week they are back with another top ten list. This time it is the 10 principles of effective web design.

Personally I think it is a misleading title as actually the post is about basic usability techniques. That said, it is a very good list and contains some really solid advice. It also contains lots of examples which really helps you get your head around the issues.

I don’t think the post has a lot to offer seasoned usability testers but if you have always avoided the subject in the past then this is a nice simple starting point.

Back to top

Feature: The profit and loss of usability

We have looked at the subject of usability testing relatively recently on the show and have mentioned it countless times before. However, never have we taken a step back and asked the fundamental question: why is usability testing important? This week we look at how usability testing can be presented to clients and dispel some of the misconceptions about it.

For more detail on what we cover read my post The profit and loss of usability.

Back to top

Expert interview: Eric Meyer on version targeting

Paul: So on last weeks show we discussed Internet Explorer 8 and some of the plans Microsoft have for it and joining me this week is Eric Meyer. Hello Eric, good to have you on the show.

Eric: Cheers, hello.

Paul: Cheers, very British of you.

Eric: I try to localise my content.

Paul: Yes, very good. Most impressed! We’ve got Eric on the show really because this whole news surrounding Microsoft and what they were planning to do was broken on the A List Apart website, in 2 articles, one of which Eric wrote. The reason I think I wanted to get you on the show to talk about it is because I was kind of struck by the honesty of your article that a lot of what you were saying was very much ‘I really don’t want to like this idea but…’ kind of attitude. Which is the kind of attitude that I think goes down well with the people that listen to this show. So I was wondering Eric, could you kick off by giving a brief explanation of what it is that Microsoft are proposing doing, just so that we’ve got a bit of an understanding of it?

Eric: Um, well pretty much what they’re proposing is that Internet Explorer, Internet Explorer 8, and in the future will recognise a single element, a meta tag which can be used to say ‘This is the version of Internet Explorer that I want you to act like’ so that Internet Explorer 11 can act like Internet Explorer 8 did, or Internet Explorer 7.

Paul: Right.

Eric: So the idea basically is well they call it Version Targeting so that you can have your page actually targeted to a certain version of Internet Explorer even if newer versions have come out.

Paul: Ok. So, I mean at face value that sounds very much like browser sniffing, which is obviously something that has kind of become considered bad practice. How does it differ?

Eric: Well my perspective is that it differs because it actually reverses the direction of sniffing. Browser sniffing To date it has been a case of authors trying to figure out what browsers they want to sniff for and what should happen as a result. This was sort of the reverse side of. It would be the browser sniffing the page to say, ‘What do you want me to do?’, at least in the case of Internet Explorer. It should be clear that Microsoft has so far as I am aware anyway not proposed that anyone else has to do this, or that this has become any part of any sort of standard. They’re saying if it is what we want. This is what we’re planning to do on the browser. So they have the same echoes. I really feel like they are not the same thing. There are a lot of people who disagree with me but they are really very different because one of the reasons that browser sniffing by authors has become viewed as a bad practice is because it becomes fragile over time. First of all there’s a case of people saying ‘If this is Internet Explorer do such and so’, and they don’t consider that there might be a version this version of Internet Explorer that might change what I’m trying to work around. Another reason is that people make guesses about what will be in user agent strings which is usually what people browser sniff on and sometimes those guesses turn out to be wrong as when Safari first came out its user agent strings had the phrase White Gecko in parenthesis.

Paul: Yeah.

Eric: A lot of people had done browser sniffing scripts that looked for the word Gecko so all of the scripts thought that Safari was Mozilla, Firefox or whatever and so some of them broke. In a way it is hard to fault the authors because how are they supposed to know a browser from Apple, not based on Gecko, would have the word Gecko on it’s user agent string and yet that happened so this would be much more controlled it seems to me were it would be browser saying ‘What do you want me to act like?’ and that’s the other reason this has varied from browser sniffing usually tries to predict the future, whereas this version targeting that looks like it will happen in Internet Explorer it actually has to predict the past which is inherently easier to do.

Paul: Why has Microsoft decided to take this position? It seems to me like in some ways they’ve created a rod for their own back in the sense that they are producing I don’t know, Internet Explorer 11 and in that has got to be the rendering engine for 10, 9, 8, 7 and all the way back. Is that not just a lot of work on their part? What’s driving them in this direction?

Eric: I think it will be a lot of work on their part. I do not believe that the idea is to have completely separate rendering engines for each of those browsers, but to have effectively if then else statements in the rendering engine or cases or switch statements for a little more program language savvy. The reason that they’re doing this is that they really have a problem, well they have a dilemma, and they need to get out of the dilemma. The dilemma that they face is basically this: There are a lot of pages out there both on the public web and in private intranets that were developed to IE6 or IE7 let’s say and weren’t developed in an open standards way they weren’t developed the way a lot of people in the standards movement develop sites which is to code to the standards and then figure out how to work around the problems that they encounter. If you have an intranet, a massive intranet site, and your company standard is IE7 the default tendency is to develop it to IE7, you use behaviours that are not correct according to the standard, but are the way that IE7 acts. You have all these past mistakes which every browser makes past mistakes. IE7 it can be argued has made more mistakes than others, not an argument I want to get into. They have that situation where there’s this large legacy of pages. So they can’t have those break for a number of reasons, many of them business reasons, actual monetary reasons but at the ground level you could look at it as if the browser changes enough that those pages start to break it becomes a bad browsing experience for users of that browser. At the same time, the Internet Explorer team would like very much to improve their standards support. They did this for IE7; they fixed a lot of things. They went more towards the updating the browser and didn’t worry quite as much about breaking sites, and they got into quite a lot of trouble over it both internally and externally. There were people who took on the task for breaking the web. We’ve been here before as an industry back in 2000. This is how DOCTYPE switching became to be because the first few iterations of IE and Netscape did very badly at CSS support. They tried and more power too them but Internet Explorer got the box model wrong. Netscape got so many things wrong that they had to junk it and start off with a new browser. So DOCTYPE switching was basically invented as a way of being able to say ‘Given these criteria render pages the way they used to be rendered and give them these other criteria go the more standard route’. From the Microsoft prospective this is like a super DOCTYPE switch except in this case instead of hinging it on a DOCTYPE they’re hinging it on information about what the browser version is.

Paul: Yeah. It is interesting when I first heard about this and I first read your article and the other one on A List Apart, my initial reaction was oh no all this feels wrong and indeed which is what in the sense you said that you went through when you were writing the article and then we actually were discussing it on last weeks show with John Oxton and John Hicks, the same kind of thing came out that we felt very uncomfortable with this and we were reflecting on a lot of things that were said including Jeremy Keith who made this point of you’re actually in a position where you’re actively having to tell Internet Explorer 8 how to behave like Internet Explorer 8 and not IE7, which seems a bizarre set of circumstances but since then I’ve been thinking about it a little and I’m struggling to come up with a reason that I really object to this beyond a principle of the thing that I feel like I shouldn’t agree with it but I’m having trouble articulating why I guess. What kind of things are you seeing? You must have seen a lot of objections come up when you are brave enough to step up and say what you did on A List Apart, I’m sure there was quite a backlash. What kind of things are people saying and throwing against this?

Eric: Well I can see quite a few objections.

Paul: *laughs*

Eric: Well one of them is an objection that Jeremy brought forth which I think I am glad that were talking about that which is that the default behaviour is to default to IE7 as I understand. So when IE8 comes out, if it doesn’t see this version target meta tag in a page it will assume that page is to be treated as IE7 would have treated it, which again completely reverses what were used to. What were used to is when a new browser comes out the page will be interpreted according to the latest and greatest unless you’re using a quirks mode DOCTYPE switch but anyway, it’s bizarre, and yet I did go through the same thing when I first saw it I was like ”;What!”; *laughs*

Paul: Yeah.

Eric: Good old WTF basically.

Paul: *laughs*

Eric: And when I first saw this it was at the beginning of January when Aaron Gustafson submitted the first draft of his article, I hadn’t actually heard about this before then, and I started arguing with him about it in the A List Apart editing forum and very quickly came to realise, ”;Why am I objecting to this?”; and I took the time to step back and say ”;What’s the deal here?”; and that’s were my article came from with the two of us going back and for and it was like, could you write an article about that because that’s what a lot of people are going to go through. So there’s the default behaviour, you have to have the meta to have the latest and greatest, yeah, I would like to see that changed, I would like to have Internet Explorer act the way browsers always have because that’s what I’m used too, but at the same time I have to be realistic about the problems that they face, that I was explaining before.

Paul: Yeah, because that wouldn’t solve their fundamental problem which is that people that aren’t aware of working to the latest standards aren’t going to be aware that they have to put meta data in there pages either.

Eric: Right, so people have said that Microsoft, of anyone, are in a position to educate people who are not aware of the need to do that one thing and that’s the point I made when I was talking to a member of the IE team about this. There are other objects that I think mostly have been related to that for example Bruce Lawson posted just in the last few days, the default behaviour means that from an accessibility point of view any page that doesn’t have this meta tag is stuck with IE7’s accessibility support which apparently is not were it should be, so he’s made the point that freezing pages to IE7 in the absence of any other information freezes them in this inaccessible state, I’m not an accessibility expert so I don’t know how best to categories that but, will have accessibility problems and keep them for years and years as opposed to, ya know, as accessibility problems improve in Internet Explorer, getting those benefits and that a very real objection and a very real concern, and I think one that needs to be made to the Microsoft people because accessibility is very important. There are other objections from the javaScript community, it’s actually been interesting in the javaScript community, there’s been a real dichotomy, there have been some javaScript library authors who’s been like ”;Yes! Finally, thank god somebody is talking about versioning!”; right, and there’ve been others who have been like ”;Oh my god, are you kidding, libraries will have to support all these backwards engines”;.

Paul: Yeah.

Eric: Again, I’m not an expert to know what side of that certain dichotomy I would come down on, in my own blog posts I’m saying, wouldn’t you just use object detection and not worry about what the version number is?, but I guess that doesn’t fully address the problem, or at least that’s what I’ve been told. Again it’s something that needs to be figured out, because there have been some fantastic javaScript libraries and fantastic things that can be done with those and if this is a move that will hamper the growth of that field then that’s something else that would also bother me. There’s also the security objects, the idea that every backwards versions that you have your complicating your ability to keep your security holes closed, that’s the kind of thing we’re I have to say ”;Look that’s down to the browser maker to figure out”;.

Paul: I guess that the problem with the whole default behaviour thing is how badly it’s going to break things, how many sites are going to be broken if the default behaviour was the latest browser and how badly are these sites going to be broken, are they going to be unusable? Are we talking about some small changes? I guess there are a lot of websites out there as well that just aren’t supported at all, content has been put online, nobody is checking them, nobody is checking when a new browser comes along, and if when IE8 was released and those websites became completely inaccessible and unusable then obviously there’s a big issue there because your loosing vast amounts of content on the web. I guess a lot of it is unknown entities at the moment isn’t it? It’s guess work.

Eric: Yeah. There’s a lot of guess work and that’s part of the problem, I think if we knew, if there were some kind of statistics, like if the IE team were to come out and said we built a version were the default behaviour was latest and we tested it against these 1000 QA sites and 20% of them broke, 5% were completely unusable, that would be one thing, or if they said we tested and 2 sites were broken, you could still read them but they look a little weird, that would be different and nobody does really know and at least nobody outside of Microsoft and maybe not even them, I don’t know if they’ve done that kind of testing yet. And, I’m sorry, there is one other objection that come from a couple of other browser manufacturers that this is in affect, anti-competitive, whether it’s meant to be or not, but it has that effect because Internet Explorer has such a large market share they have to worry about what it does and so to them it becomes a lot harder to do that, so if they are going to keep up this effort it’s going to become a lot harder to do in this version targeting world after a few releases, of course the question is whether they would need to do that or not and there’s a lot of debate about that, there are people who are like, ”;Just stop, stop making reference to Internet Explorer and just let it strangle itself to death”;. I don’t want to get into all the back and forth that’s happened but that is another objection.

Paul: I guess, is there some danger here as well that once you’ve got this version targeting in place Microsoft could go off on some complete tangent and start introducing all kinds of proprietary functionality into their browser that isn’t supported in the other and we end up in a situation were we’re having browser wars again with different browsers implementing different proprietary tags and stuff like that, or are we truly beyond that do you think?

Eric: Umm, It is certainly a risk and how much of a risk influences were people stand on this. I don’t see that personally as being much of a risk, because yes, the Internet Explorer team could introduce a load of proprietary properties that let an author change the colour of the browser kernel right or replace the backwards and forwards buttons in the browser, whatever, and if nobody else supports that, I think those efforts will largely support them strangling themselves, I mean, colours scroll bars we don’t really hear about anymore, people will still do them, but nobody else really cares.

Paul: Yeah.

Eric: So, I would not claim to be smart enough to see everything that could possibly be done but I think in a lot of ways that might not be a bad thing, because it would let Microsoft experiment with less constraints, they can do initial implementation of things that are in CSS3 and if the behaviour of those things changes, in a later version they can fix what they did, they can change to match the spec, and then still have backwards compatibility for pages that took assumption of that, there are people that would say ”;Well they should never implement anything that hasn’t been completely nailed down”;, but that doesn’t really work, because one of the ways that you get yourself out of the candidate recommendation stages at the W3C is to have implementations and the only way to have implementations is for someone to implement them and if they find out during that phase, and this is what the candidate phase is about, if they find out in that phase that such and such property that has been specified, doesn’t work in the real world for whatever reason, they’re then able to go back and change it, but in the mean time if your Internet Explorer and 11’ty billion pages have been implemented to your particular version of the property.

Paul: So, you wrote this A List Apart article that went out of the 21st of January, and we should probably say that we’re recording this interview on the 29th of January and it’s not going to go out for another week, so things might move on in-between but there’s obviously been a lot of discussion since you released that article, have your views really changed in any way, are you still in favour of this approach or have there been more concerns raised that have made you doubt it?

Eric: Hmm, I’m still in favour of the general idea, the implementation I’m a little less, I guess I would say I’m more agnostic about, but the general idea of backwards compatibility so as not to break sites, whilst still being able to change and improve and fix the browser I’m all in favour off. The default behaviour still bothers me, and I’m still having trouble working out if there is a case of honest to goodness technical reasons or it just feels wrong, there has to be something, how’s it’s gone about is a different question, if the IE team and the people that they have to answer to effectively can be convinced that having the default behaviour switched and then educating people whole want to freeze their sites, on how to freeze there sites using the meta, they can be convinced to do that and that that will fix the problems that they face then I’m all for it. Accessibility concerns do bother me and I think that that’s another argument in favour of switching the default behaviour but at the same time I also have to recognise that switching the default behaviour, as you said earlier, from a certain prospective, completely negates the entire reason to do this, given a certain set of circumstances if they’re going to switch the default behaviour it might almost as well not even do it. Again though that also depends on how much breakage on how many sites and what kind.

Paul: Yeah.

Eric: The other unknown that we’re facing here is we don’t yet know just how different the IE8 behaviours are compared to IE7, we have a clue in that we already know that there’s an internal build of IE8 that can support the ACID2 test, which they weren’t even really very close too, well I guess that depends on who you talk to, but they didn’t support it before so in or to support ACID2 they have to have things like generated content with CSS which we completely don’t have in IE7 they have to have fixed some of there parsing bugs and so on and so forth that hints at a very large change, an even bigger change than that between IE6 and IE7, but we don’t know at this stage, it could be, and this has been one of my objects to ACID2 or one of my discomforts with ACID2 over the years, it could be that they’ve just implemented exactly the things that were needed to ACID2 work and not more, which is the wrong way to go about supporting ACID2 and I don’t know that they’ve done that, but it’s possible. So I may be that it’s not such a huge change, but those things came about as sort of a broader push towards implementing more advanced selectors and fixing bugs and implementing area content and really pushing into the areas of CSS2 that they don’t support and the areas of advanced CSS modules that they don’t support yet, then that could be an enormous change and in that case there’s more chance of them breaking a ton of sites if they don’t have a mechanism like this in place to deal with them.

Paul: So I mean in some senses, after discussing this for however long we’ve been discussing it, there’s very little at this stage that we can really do about this, for the average web designer, Internet Explorer 8 isn’t going to be out in the immediate, one presumes and even when it does we’re talking about relatively minor changes that they’re going to have to make to their site, I guess the biggest immediate concern is maybe how you think about building sites at the moment, do you worry so much about things like progressive enhancement if maybe the way that Internet Explorer works in the future, I guess they’re still good principles.

Eric: Yes, absolutely they’re still good principles and that is how I have developed sites and will continue to develop sites and really I think what it comes down to here is who’s going to have to add the meta tag.

Paul: Yeah.

Eric: Is it going to be people who do forward development, ya know forward compatible development, is it going to be standards aware developers or is it going to be people who aren’t in that sort of craft. It’s not fair in a way, that the people who have been doing things in the right way will have to take that on, but on the other hand it could well if you look at the numbers, there are a lot less of us.

Paul: And in some ways we’re much more equipped to do it.

Eric: Right, to understand why it’s needed and how to do it the right way and when not to do it for that matter.

Paul: Yeah.

Eric: Because when you develop a site for someone and it seems like a one off and they really not going to be keeping it up, you might want to keep the meta tag off because you know that that will default to IE7 and that’s what you want or you might explicitly include it because you want it to default to IE7 and IE8 or ya know on my site if this happens I’ll probably use the edge keyword of IE equals 1024.

Paul: Yeah.

Eric: So I’ll always get the latest and greatest and so that’s for me personally, but for a client I would have a much harder time justifying that because clients don’t care if it’s forward compatible development or not they just care about whether their site works and will continue to work right and so the point has been made, as you did, that in a lot of ways people who are sort of more clued in are better equipped to take this on as opposed to somebody’s grandmother who uses Dreamweaver or whatever to put together a site for her poetry circle and this happens all the time, they don’t even look at the mark-up let alone would understand why this would need to be done why there would need to be some kind of meta tag, first you would need to explain the concept of tags etc, etc, etc.

Paul: I mean a lot of these people don’t even realise that there are multiple versions of a browser, or indeed that there are multiple browsers.

Eric: *laughs* And those are also issues that will need to be faced, somebody is going to have to do this. A lot of these advances that IE8 seems to have ready to go as indicated by passing the ACID2 test may well have to come back out.

Paul: Yeah.

Eric: You can always back changes out of a code base and that really seems to me to be the dilemma and I’ll want to say again, there are people who reject that as being the choice, there are people who feel very strongly that that isn’t in fact a choice, even if you except that’s a choice that they should keep the changes in but be willing to break sites because sites always have to be updated anyway and that’s how browsers have always done it, so there’s a large amount of history that says that can happen but browsers can still advance.

Paul: We will see won’t we, that’s what it boils down to. We will have to wait and see. Okay, Eric, thank you so much for coming on the show and talking that through, it’s an interesting area and I think it’s an area that will have significant impact on how things develop in the future, so it’s good to have your perspective on it. Thanks very much.

Eric: Paul, thanks for having me on, I really appreciate it.

Back to top

Listeners email:

Basic CMS options

As in last week’s show I so badly bastardised the name of our textmate reviewer, I decided it is only fair to allow him another go this week. He asks:

I have a friend who wants me to build a website that they can update. They don’t know how to write HTML or CSS. What is the best way to solve this problem?

The obvious solution is to use a content management system of some type. Indeed this is something we have covered before on the show back in show 24. If you were building a relatively big site that needs constant updating, then I would definitely encourage you to check out that episode. However if your needs are simpler a full blown CMS might be overkill. In such situations you should consider a simple blogging system like WordPress.

One final option maybe to use a tool like Contribute. That way you could build a nice simple static site but your friend would have a WYSIWYG tool to edit it.

The importance of blogs

Our second question is from Lauren:

My question is about starting a blog. Right now, my nights are filled with homework, but once I complete my program, I can truly dedicate my time to building my portfolio. Are blogs really that important now? I don’t know if I can come up with meaningful content to make my blog stand out. Does this mean that I won’t be able to find a good job in the industry without one?

You talk about a portfolio and blog as if they are separate things, but I would suggest they should be combined. As somebody who regularly hires web designers, I have to say I find portfolio websites frustrating. At worst they show me some pretty pictures, at best examples of fully functional websites. However, what they don’t tell me is the thought process behind the designs. Why did you choose the colors you did? Who is the target audience and how did this affect the design?

A blog can also be used as a portfolio but allows more depth into the designs you are presenting. Also, a blog allows you to link to other content on the web and express your views on that content. This shows me as a potential employer that you are well read and interested in what is out there.

As a student looking for your first job I wouldn’t expect your blog to include new CSS techniques or innovative approaches to user testing. However, I would like to see examples of work fully explained and evidence that you are reading other web design sites and have opinions on what you are reading.

Boagworks

Boagworld