Category: Web development

I just stumbled upon a post about Operas PR-manager Eskil Sivertsen’s comments on Nokias S60 browser — where he basically agrees with the review in the Register. Not everyone agrees with him, of course, and think his words were harsh. Everyone is entitled to their own opinion of course, and our tastes vary, but one of the comments made me stop up and think a bit:

Concerning Opera mini – I really don’t like the way the browser changes the layout of full html pages to fit the viewing platform, as it means the designers lose control of how their pages look. I much prefer the S60 browser’s solution to viewing large pages on a small screen.

If I want pages optimized for a small screen I’ll use WAP. I personally believe that the responsibility of a proper page layout/viewing rests with the page designer and not the browser app.

My thoughts are spinning around this question: How important is the original layout, as the designer meant the site to be seen?

One point to consider here is the purpose of the design. Is it meant to make the site look pretty and inviting only? To enhance the readability of the text? To guide where you’re viewing to the most interesting links and pages on the site? Is the design important to the content, that the design itself provides part of the content? A second point is about the content itself: Is it meant to be read? Or just viewed, or what?

I think it will be safe to claim that for most sites, the point is for the content to be read. Maybe commented on and be discussed, but definitely read. In this case, how important is it that the design is preserved, in every case? It would be nice when it’s logical, but are the cases when it’s not that logical to preserve the original design?

Sites are usually designed to be viewed on a large screen, and the designs are based on this situation. Few sites are designed for smaller screens (or other media) even though there’s a lot of talk about accessing the web with mobile phones these days. What shall the browser on these phones do, if there are no stylesheets for them to tell how the design is supposed to be on small screens? After all — doesn’t this mean “sorry, no design for you”?

The phone browsers do handle it differently. Some pretend to have big screens, and zoom in on parts of it to make it possible to read. Other browsers reformat the whole thing and present a long, narrow page. (Are there more than Opera that does this?) Which of these approaces respect the designer’s wishes? Which are correct? And which are best?

The browsers that pretend to have big screens may be said to respect the designer — if the designer meant that all devices should behave like a big screen. However, if the designer meant “I don’t know what’s the best design for this device — present the content as you wish” then any rendering is OK. The correct way to render would be to follow the specified style sheet for small screens, if presents, and render without styles if not. (At least to my understanding — you may disagree.) The best way?

The best way would be to render the page so that the content is accessed effectively, i.e. easy to read, and that it’s quick and easy to navigate. If it’s hard to read what’s on the page, and navigation is complicated, then something’s wrong. It doesn’t harm that the browser is fast either, and have effective use of available memory.

So, which browser is best? You decide.

Banning Opera, part III

In the great debate – or ruckus – about the reasons for banning Opera users from visiting web sites, the point is this: Unregistered Opera browsers using targeted ad banners from Google are (mis)using the content on the web pages without the owners consent. The targeting means that competitors may be advertised, or even identical ads to the ones in the web pages may be shown. But it is the use of their content that seems to make some web masters ire, from what I’ve read in the comments here and in the thread in the Opera forums.

So I’ll concentrate on how the content of the web pages is used.

First up is Opera. How does this browser use the content? Well – it does parse the formatting commands in the document to show it as it is described. Just like a browser is supposed to do, and just like any other browser do. It doesn’t analyse the content to extract a meaning out of it, nor does it find any keywords to describe what it is about. When it has rendered the content as instructed, Opera does nothing more with it. In this way, it behaves just like any other browser.

But it does something more, even if it’s finished with the web page and doesn’t do more with that. It does send the URI to Google, saying “Hi, this is where I am now. What do you want to show in your own space in the browser interface?”

Google is the one indexing the pages, the one extracting keywords from the pages to discover what it is about. That’s why so many have answered that Google is the one to blame, and the one to go after. Some agree that yes, Google is doing these webmasters wrong, and they should find a solution. Others aren’t quite that convinced. Let’s take it in more detail.

First of all, Google has just as much right to traverse the Internet as any other, search engines and humans. And web masters have their rights to block search engines from all or parts of their sites. Google uses its rights to traverse and index the pages on the net, except the pages it’s asked not to index.

Secondly, what is Google actually doing? It does just what you’re doing right now: It reads pages, find the meaning in them, and can use what it has learned later on. Actually you, being a sentient being, does a much better job out of finding the meaning in the texts you read. Thus, you can do a much better job out of using what you’ve learned from my (or any other) pages than Google can.

Let’s say you do. You use what you’ve read on my pages for your own good, maybe even earn money on it. Are you allowed to do this, is it legal? Well, as long as you’re not just copying my stuff or haven’t signed a non disclosure agreement with me to get access to my pages, sure – it’s perfectly legal. Nor are there any ethical problems. The pages is available for anyone to read and possibly learn from.

If you had used a computer program to distill the information from my pages in some way, before you read the result, would the situation have changed? No, not really. Even if the process was completely computerised; reading, distilling the information, act on what’s learned – it still would be just as legal. That’s a description of how Google works, too. It reads my pages, it distill the information to find relevant keywords to describe the content, and act on what it has learned by showing relevant ads. (At least try to – for me personally it has been quite a lot of blog-related ads, unrelated to the content… But as I said, you are doing a much better job of understanding the meaning than Google. ;-))

So – is Google using my content to show ads? Well – it is using my content somewhere in the process. To use the content, you must be able to understand it, too. Thus, Google is using my content when it is distilling keywords from it. Just as you use my content to learn what it is about when reading it. When you use what you’ve learned you’re not using my content anymore – but can the same be said about Google when it use the keywords to show ads?

My initial response is no, it use what it has learned. But if I would say yes, how would my content be used? There’s no trace of it in the ads, nor can I find it any other place. True – it is in Google’s cache – but it’s also in the cache on your computer. That doesn’t mean you’re using it. So again – my second response is also no. Google use my content to learn what it is about – a use I’ve allowed – but not in the process of showing ads. And – do Google combine information from different pages, even different sites, to serve as relevant ads as possible? In that case Google use more than the information on the single pages to decide relevancy.

The combination of Opera and Google Ads – does that change anything? Well, let’s see.

It’s clear that ads in Opera itself may be in competition to ads on the web pages. While competition isn’t always wanted, it’s not a problem either. What some see as a problem, is the way the ads are picked: “Our content is used, without our consent, to serve targeted ads directly in competition to our own ads.” Those who have signed up for Google Adsense may even discover that there are identical ads in both Opera and the web pages.

Now, we know that no content is reproduced in any way when Google shows the ads. As previously argued, Google reads the pages and extracts information to learn what the content is about – which is how Google use the pages – and then use what it has learned to show ads. That usage has been allowed. Webmaster who have signed up with Google for using Adsense on their pages (such as I have) have even allowed this explicitly.

Of course, even if Google is doing everything legally, it doesn’t mean that no one is frustrated over not having control over what Google does with what it has learned. Therefore we can hear people trying to claim more rights than they have (something that’s not too unusual elsewhere either, unfortunately.) But again, neither Google nor Opera is doing something wrong.

What’s left then is, what can be done to please those who do protest, and is it worth it?

Opera has been made the scapegoat here, so what are they asked, and what can they do? They are asked to either make it possible to identify those who use unregistered browsers, showing targeted ads (read: Help webmasters to block Opera users) or to read and act on a meta-tag (or similar) that says not to show any targeted ads. The first option should be unacceptable. The second – asks Opera to do things browsers doesn’t do. It’s possible, but not necessarily easy. Is it worth it, to please those who don’t like what Google do in its space in Opera? So far, Opera doesn’t think so.

Google is the one who is responsible for the ads and which ones are shown. Google is the one part to blame, should there be any to blame at all. I know Google has been contacted about this concern, but I have no idea how the problem was presented. But Google didn’t want to do anything about it, as they have a contract with Opera. Still, it’s Google that has the key to it all – maybe if it’s presented correctly, Google will agree that there is a conflict of interest they should address?

In any case, the problem isn’t a case of someone doing something wrong or unethical, merely a conflict of interest. If something should be done, it would be to create goodwill. And if someone should do anything about it, the most logically choice would be Google, that has the control over the ads – anything else would just be patches, and there could be many of them, with a potential struggle to get them in place.

PS. The other side of this problem is the webmasters who block Opera users from visiting their sites. It has been said that anyone are free to block whoever they want, and that it only would create a bad reputation for themselves. But is this correct? Not about the bad reputation, but that anyone can block whoever they want? Or does that depend on what type of web site it is? Some countries have regulations that says that certain types of web sites should be accessible, i.e. available to everyone…

After my previous article about banning Opera from sites, I have been thinking more, and have got a bit more information from the man behind the protest site, Kenneth Barbalace. First of all, what is the problem that has made him take the step of banning Opera?

The problem

The problem isn’t that Opera shows ads in itself, but the targeted ads. These works by letting Google read the visited pages, analyse them, and serve ads that are relevant for the content the user is viewing.

For a commercial site, these relevant ads may very well be a direct competitor. Whether this is a small or big problem, or maybe even not a problem at all, may not be the point. The fact that there is a potential problem, in that potential customers can visit a competitors site by clicking an ad when viewing your own site, is the point. While you don’t want to prevent potential customers to visit your competitors, you don’t want to advertise for them, either – but if your content is used to serve ads for competitors, this is basically what you do.

If you use Google Adsense, the ads served by Opera may be in direct competition and even identical to the ones on your pages. Unlike Google Adsense however, you don’t have the option to not showing competitors’ ads.

That’s the problem. If it is big, small or non-existent now isn’t important, the principle is. It’s a case of not wanting to “give the devil the little finger” out of concern for what the future will bring if you do.

A solution?

Since it’s the targeting that is the problem, a solution could be to opt out of the targeting by way of a meta-tag or something like that. Initially it looks like a good idea. Maybe it is, maybe not. Technically, both Google and Opera could do this. Opera could implement it for the browser, Google could do it for everyone who use targeted ads just the same way.

Google however has a contract with Opera, and can’t just change the product they’re serving, so the ball is with the Opera guys. From their point of view, is it a wise move to allow opt out from targeted ads?

Two reasons why Google ads are popular: First, they’re unobtrusive, and don’t annoy with flashy graphics, sound, pop-ups and all that stuff that makes ad blockers a good idea. Secondly, they’re targeted. What would happen if site owners could opt out of the targeting?

Probably not much. But then again – what if this is another “give the devil the little finger” – what if everyone opted out of it? The consequences of that would be – well, who knows. Bit this isn’t just about Opera – targeted ads are being served visitors outside of content writers’ control in several ways. In browsers, browser shells and extensions, desktop applications, mail programs…

If targeted ads were replaced by generic, random ones, I’m sure they would lose popularity with users, and it wouldn’t be as popular scheme with the advertisers either. That’s my guess. Could it be raising trouble for oneself? But I’m just speculating.

Gridlocked

As it is now, there seems to be a gridlock, or a trench warfare. It’s possible to ban Opera by checking the UA string, by checking for unique features in Opera by Javascript, it’s even possible to block Opera users if they come back with a changed UA string (or even a different browser…) On the other side Opera users can try editing the UA.ini-file, adding one line to pose as a different browser on that one site, they can use userjs to hide the unique features of Opera, use Proxomitron…

The question is of course who would try the different things to get access to a site that ban them? People who use Proxomitron are normal people who use a nice tool to get access to sites with sloppy coding. Roughly speaking. adding a line to get past a ban is no different than adding a line to get past sloppy coding. Editing the UA.ini-file is no different. As for getting a script that hides unique features in Opera, so that the browser can’t be identified through them – such a script already exists, and has done for a long time.

Bottom line is, users who want to gain access to sites that ban them will manage, so it’s no use to put in too much work in the banning. Most users will probably just not care. Maybe they’ll be a bit annoyed, maybe they take time to choose the “report a site problem” option in the help menu, maybe they just go on to a different site. But as for making a statement, to make people and Opera aware of it, I believe it’s enough already.

Is it really a problem?

I understand the concern Ken is voicing, I really do, and I even can understand doing something like this in frustration of being ignored by the people at Opera. I don’t agree with him, though.

First of all, I’m not convinced there really is a problem. True – something outside of my pages may show a link to a competitor based on my content – but then again, maybe the competitor shows an ad for me the same way. If I have an ad with Google, that is.

And is the use of my content to show relevant ads a problem? I’m not convinced of that either. I believe the content on the pages, including the ads, is what gets priority from the reader. If it doesn’t, the problem may very well be with my content.

Secondly, is banning Opera users a good move? True, it does make a strong statement, but it also stir up a lot of feelings, and not only with Opera users. And a reputation can be torn down ten times faster than it is to build up. Luckily, I’m not the one to take that decision, and to consider if it’s worth it.

One thing I do know though, is the no matter how much you disagree with a person, don’t get rude with him. That way he just get his shields up instead of trying to see your point of view, and no one’s the wiser.

Anyway, I think what Ken wants now most of all, is some words from Opera about what they think about it all. 😉

A mere hour ago I discovered a new site – or new for me, at least – Stop Targeted Opera RADs. I was directed to this site, or more specifically this page, when I tried to visit Environmental Chemistry, because I used Opera. Apparently, the owners of the site really doesn’t like targeted Google ads. (From when I hear, they don’t like Adblock in Firefox either…)

Since I heard they blocked access for Firefox users with Adblock, I first assumed they were just scared of competing ads. The first thing I read on the “banning” page shows this, too: Site owners being afraid of visitors seeing ads for their competitors, and that they can’t prevent it as it’s outside of their website and control. They may call it unacceptable misuse of their content, but – is it?

There are other services that does what these site owners are afraid of much better than the Google ads in Opera: There are sites that gather information from many competing sites, compare the products and prices, and give users many alternatives, without even showing ads from any of the sites in question. Scary thought, eh? Especially for those who are afraid of users comparing them with competitors…

I also thought a bit further. When visiting the sites with the Google ad banner in place, Google read the pages to see the content and serve related ads (unless it belongs to the exceptions defined in Opera). If the page isn’t in Google’s index, Google want to add it. Therein lies another problem: What if the site doesn’t want the pages indexed?

I must admit I can’t understand why someone don’t want publicly available pages that they want people to visit not to be indexed by Google. That however isn’t something I should speculate over, and it’s not really a point touched upon. It should be interesting enough to read what is written on the site I was redirected to, right? And the menu contains hints that I can find some answers and useful information there.

Someones brain must have been short-circuited. I was sent to the site because I use Opera, right? But – as an Opera user I’m prevented from seeing other pages than the one I was redirected to. Why? Afraid of showing Opera users the arguments? (The ones I read on the single page didn’t convince me of their view.)

Conclusion

All in all, all of this leave me with one impression: It’s better to ban Opera users and have them definitely visit competitors, then to allow them in and risk that they may be tempted by a competitors ad that may be shown in their browser.

If they’re afraid of being compared to their competitors – maybe their products aren’t worth the asking price in the first place anyway?

Who knows? Not Opera users…

The first browser to pass the Acid 2 test: Safari. It has been nice to follow the blog of Dave Hyatt and see how he proceeded, not only implementing in Safari, but also finding an error in the test itself. I’m sure he will continue, implementing even more of CSS. (And those pouty faces that insisted it was all a PR stunt from Opera should rather reconsider their view.)

How soon the patches or upgrade will be released to let normal users try it out, we’ll just have to wait and see.

So, should we just announce Safari as the winner then, or…?

As sensible voices have uttered earlier, the important thing isn’t which browser pass the test first – it’s about all browsers working towards standard compliance, and passing the test of course. But first of all implementing and supporting the standards, and not stopping the development and growing stale.

In this scenario, the winner isn’t the browser which first pass the test – the winners are us, the users of browsers; we will find that as browser support the standards, the browsing experience will be enhanced. Web designers will have to worry less and less about not being able to do what they’d like, because of lack of support in browsers. The Acid 2 test is a goal to work towards, and a pressure to actually work towards it. That Safari is victorious in the race to pass the test shows that we’re on the right track. We’re winning more and more as the browsers support more and more, all of us.

In this race, we’re all winners.

Many Opera users have noticed that some people, be they FFFB or otherwise, constanly makes silly claims about the best browser, Opera. These are for the most part claims that those same people themselved could’ve refuted easily had they only bothered to really try the browser they so much like to slag off. Those are the people blaming Opera for their own ignorance, and there are many answers to them around the web.

Then there’s those who blame Opera when it’s their own incompetence that’s at fault. Take the knowledge of CSS, and a simple thing like having the content of the body all the way out to the edge. Simple – just set “margin: 0” – right? Works in IE, works in Firefox, doesn’t work in Opera. Opera must be broken. And yes, this is an example I’ve seen a couple of times the last couple of days. I’ve seen similar earlier.

Not knowing the default values for margin, padding and so on in the different browsers is perfectly OK. Not knowing that there are default values? Well…You may find yourself with problems that puzzles you. At least until you give it a thought. Not knowing that both margin and padding influence ont the spacing between objects, such as the content in the body and the edge of the browser window, that’s incompetence.

There’s nothing wrong with being incompetent – we’re all incompetent in several areas. In the areas we have an interest, we educate ourselves to become competent. The aforementioned people with their problems with the margin in Opera could educate themselves further by reading or asking, and finding that the solution to their problem is something as simple as setting both margin and padding to 0. Instead of overcoming their incompetence in this field, however, they chose to blame it on Opera. And they’re not alone.

I wonder why so many are so quick to slag off Opera…

A while ago, before the WaSP Acid 2 test was ready, Opera challenged Microsoft to take said test in the upcoming IE7. Not everyone appreciated this challenge, claiming that WaSP is working for Opera and such (see among the comments after a post in Scobleizers blog.)

Well – the wait is over, and the WaSP Acid 2 test is here. How does the browsers fare? Well, not unexpectedly, IE6 fails miserably – but no one have claimed IE6 support CSS2.1; it’s the next version that’ll be interesting to test. Maybe to the glee of some, Opera 8 fails too, and so does FireFox. Every browser fails, truth to be told.

So, now we just have to wait and see how fast the browsers get the pieces implemented correctly. Work has already started on the Safari, for those of you on the Mac.