dW

A Perfect Ten

Anyone who knows me thinks of me as an Opera nut, not the shrill singing kind but the fast and excellent web browsing kind. From checking instant messaging chat logs I can pretty much pinpoint to my starting using Opera to the Ides of March 2003. It’s the best browser out there in my opinion, and nothing’s really going to change my mind on that. However, I want to give a fair review of the software. I’ve used Opera 10 in one form or another for quite some time now. I’ve seen it grow from a completely unstable alpha release to this powerhouse of a build that’s quite possibly the best release of the software yet.

Opera or any software for that matter has never been without its faults. My biggest beef with it in the past has been its monkey ass ugly interface and its equally ugly icon. Well, Jon Hicks and Oleg Melnychuk have taken care of that problem completely. As those two things are the first thing that’s noticable with this release I believe I’ll start on them first.

The Icon

The Opera icon that’s existed for as long as I’ve used the software has been finally changed. It’s been an object of much frustration by many people over quite some period of time just because of how god awful it was. I don’t need to go into grand detail on just how awful the old icon was as Brand New has done a good job diagramming many of the faults the original icon had. I can notice a few additional faults myself they failed to mention, but their assessment contains enough information to prove a point.

The new logo is a breath of fresh air. It’s been carefully designed by Oleg Melnychuk. Yes, Jon Hicks isn’t the only designer at Opera as Opera has a staff of very talented designers. Oleg’s a great guy and deserves recognition for his fantastic work here. I don’t believe anyone could have done better while keeping the current identity going. As an icon it’s still uninteresting compared to its competitors, but I believe that is its strong point. Its competitors’ icons are actually quite complicated, some bordering on too complicated. Just the other day I had to explain to someone that Safari’s icon is a compass because at small sizes it’s impossible to make out what it is, and no one knows what the hell Google Chrome’s icon represents. Opera’s is a red O, easy to recognize and read; now it’s pleasing to the eye, at the very least to my own.

The Interface

Okay. Let’s face it. Opera’s interface, while functional, has always looked like shit. Almost every release of Opera has received an overhaul to its skin or the appearance of its interface. The last few releases have had their interfaces increasingly simplified by default to allow people new to the software to grow and discover its features gradually rather than being hit with them at the start. This new release is no different. Opera is actually skinned and uses an extremely powerful INI based format to denote skin settings. I and Ralf Demuth, I believe, were asked early on what my suggestions for improving the skinning system were as Opera knew that to have a beautiful interface the skinning system needed a bit of weightlifting. I was regretfully not of much help to Jon and Petter Nilsen as the problems I had with the skinning system when producing my own skin had left me. Most of the problems I still remembered were so stupid or awkward that I couldn’t really describe them fully. In fact my thoughts were to scrap it in favor of CSS with -o-skin- additions for skinning purposes, something I believe would be incredibly difficult to pull off but incredibly rewarding in the end; it, however, would be too difficult to work on for a single release, requiring a lot of time to work on. The best word of advice I had for Jon who was going to have to take up the task of improving Opera’s interface was to work at it, and report the problems as he discovered them; that appears to be pretty much what he did. It was truely wonderful to watch each snapshot improve visually.

I believe Jon created a total of three different skins: a Macintosh native, Windows Native, and Opera Standard skin. On the Macintosh the Mac native skin is the default skin, but on Windows and other platforms the Opera Standard skin is default. A Linux skin wasn’t made because where would he start on one? The best he could do is pick one distribution and attempt to make a skin for it, but it’s just extra work for pretty much no return. It is of my opinion that the Windows Native skin is rather pointless now as the Opera Standard skin does a better job at integrating with Windows Vista or Windows 7 than that platform’s native skin does. Really what is native on Windows anyway? Microsoft themselves can’t even decide what they want to do, never have really. Perhaps the native skin should have retained the old Windows appearance as there are quite a few out there that disable Aero completely, using the legacy Windows appearance instead. The new Macintosh skin is excellent enough for me to say that I’m not going to be updating my own skin, Entr’Acte, as it’d be rather pointless and a waste of my time to do so. I might make minor modifications for my own usage, but nothing resulting in a complete skin is necessary anymore. Entr’Acte is French for “between the acts” and well, Jon Hicks’ skin is the beginning of the third act. Entr’Acte I believe has fulfilled its purpose.

Now, some Macintosh users have complained that Opera’s interface still doesn’t look “native”, citing Safari’s appearance as being so. If all browsers on the Macintosh looked the same — for example like Safari — it would be rather utilitarian. While Opera 10’s skin looks different than anything out there it isn’t so far removed to make it not fit in with the operating system. It’s not perfect as it does have a few visual faults, but they’re all faults which are a result of lingering limitations in the skinning system or results of screen-space-wasting old school Windows conventions which are nonexistant on Mac OS X. Jon is aware of all of this, so nitpicking his skin is unnecessary. Dragonfly has retained its ugly appearances, but that’s about to change as well, so I won’t harp on it either.

The most notable addition to the interface is visual tabs. Opera’s not the first browser out there to implement them, but as of this particular point in time it’s the most successful use of them. In the past browsers either forced them down your throats or made you choose either visual tabs or regular ones. Opera’s done neither by letting the user use either visual or standard tabs and by letting the user toggle visual tabs on or off by simply pulling on a toggler to make the tab bar bigger, therefore making the thumbnails appear on the tabs somewhat magically. Even the toggle bar can be removed. Personally, I won’t use them much as I usually have too many tabs open to make using visual tabs rather pointless, but I’ve found it useful at times so I keep the toggle bar on. The only negative thing I have to say about the entire setup has to do with the particular way I browse. As can be seen from the screenshot above I do not have any close buttons on my tabs, and to the casual observer that would appear to be absolutely insane. Well, in Opera you can middle-click to close a tab. I’ve grown so accustomed to using the shortcut having close buttons on tabs seems superfluous to my particular setup, and Opera allows you to remove them. It’s a shortcut so useful I typically start getting pissed off when tabs in other applications refuse to close on a middle-click. However, if you middle-click the toggler it expands the tab bar to show thumbnails, and I somewhat frequently miss the tab with my cursor and middle-click the toggler instead. I’ve yet to find a way to change those shortcuts, and I believe it would be beneficial to allow the end user to remove the middle-click shortcut from the toggler.

Features

Speed Dial, my favorite feature from Opera 9 and the favorite feature for Opera’s competition to rip off, has received a facelift both due to Jon’s excellent design skills and the developers’ thoughtfulness to allow the end user to customize Speed Dial with their own background image. It’s also added capabilities of having more than nine items, up to 25. Adding more than nine has been possible in the past, but only through manually editing the speedial.ini file in your profile (or preferences) folder. It’s one of Opera’s greatest features only made better.

However, the most notable new feature of Opera 10 has to be Opera Turbo. It’s a feature targeted only at a particular demographic. Opera Turbo is for people with slow internet connections such as people who do not have access to affordable broadband internet services or are using the internet in a crowded Wi-Fi hotspot. Essentially a transparent proxy, Turbo uses an Opera-run server to compress Web pages and images to drastically reduce their size, the result being that only a fraction of the usual bandwidth is used, and data arrives much faster for those with slow Internet connections. For those with fast connections the time spent compressing may negate the time savings, but the bandwidth savings are real for everyone. I’m personally blessed to have access to a broadband connection, so this feature is turned off on my browser. However, at work I am forced to use an unreliable ISP, and I have it set to “Automatic” — meaning Opera would automatically turn Turbo on when the network speed gets below a certain threshold and off when the connection is optimal.

One complaint about Turbo I have is how heavily it compresses the images. I understand it is for reducing size, but there are some instances where the images are so heavily compressed that it is nearly impossible to make out what the images were originally especially on small images, reducing the quality of the browsing experience in the process. I believe a middle ground between smaller size and image legibility needs to be reached.

Opera 10 has many more features, but these are the only ones I really want to touch upon myself as Opera already has a comprehensive listing and description of new or updated features in Opera 10.

Web Technologies

Being the latest release of the Opera web browser Opera 10 has the best support for web technologies to date for the Opera browser. As the image above dictates Opera passes the Acid 3 test perfectly. It’s not the only new thing in terms of web technologies Opera has been supporting as Opera has been rapidly working on Web Fonts, something that during Opera 10’s development cycle its competitors such as Safari and Firefox have released final builds with support for. It is of my opinion the single most important web technology that Opera has added support for in this release, and it is the only one I will discuss. Just like with its features Opera provides a comprehensive listing.

Web Fonts (or @font-face) is an old technology, coined over a decade ago in fact. What it entails is that a CSS stylesheet author can provide a hyperlink to a font file which is to be used in the design. The browser would download the font and display it on the website if the browser supported it. However, the only browser to have retained an implementation of it has historically been Internet Explorer, although not surprisingly using a font format only it supports. I’ve written a couple of times about web fonts, the controversy, and licensing problems associated with supporting them. I’m not going to reiterate much of what I’ve discussed in previous posts, so if curious read my previous articles on the matter.

This website has been using web fonts long before Opera has supported them, linking to Delicious — an excellent typeface by Jos Buivenga. Opera has great basic web fonts support, but past the very basics its support is nonexistant. It doesn’t support declaration of weights and styles within @font-face rules which prevent authors from defining bold or italic variants of web fonts. That is problematic in my opinion as I would consider support for weights and variants imperative in claiming to have web fonts support. It also doesn’t support overriding of installed typefaces using local() as Safari does which would allow a stylesheet’s author to do thus:

@font-face
 {font-family: Arial;
  src: local("Helvetica");}

This simple declaration when used within a user stylesheet would replace any instance of Arial with the local copy of Helvetica, a noble cause from any type-minded designer’s point of view. That is only something I would like to have seen supported, but Opera 10’s current implementation also doesn’t support split fonts, a problem brought forth when TypeKit was introduced which can potentially be a problem for Opera itself. Split fonts are multiple font files containing only parts of a whole font’s character set. They are listed in order in a font stack within font-family and if a character isn’t encountered in the first font listed it moves to the next and so forth. Opera 10 gets confused when encountering these. In my opinion this is a potential problem for Opera as TypeKit’s usage on the web is growing rapidly despite only being in closed beta. When it becomes public it could explode in usage, and because of Opera’s market share and many developers’ ignorance or ambivalence of Opera they’re not going to care whether Opera supports it or not. Every single end user using Opera viewing websites using TypeKit will be fed websites that look bizarre. Examples exist today. The average user isn’t going to know what’s going on and will most likely use another browser — potentially permanently. I know Opera is aware of this issue, but I hope they fix the problem before TypeKit becomes final.

TypeKit uses split fonts mostly as a security measure, but like local() they have more of a use than just defining typefaces. An author with the ability to split typefaces and permission from the type designer could create a font file containing every character in the font which contains titular (or uppercase) numerals and another font with just text (or lowercase) numerals. This is also something I will discuss in detail in a later writeup, but the author could define the text numerals for paragraphs and other uses for body text while defining titular numerals for headers based upon how those fonts are listed in the font stack. Therefore, Opera’s necessity for supporting split typefaces for the sole purpose of supporting TypeKit isn’t necessarily truthful as it does have its uses outside of a single potentially popular service.

What I Would Like to See

Aside from what I’ve already mentioned previously I would much like to see support for border-radius, and Opera does have support for it in the pipeline — if I remember correctly using SVG for rounding which will be quite interesting. That and any plethora of CSS 3 or missing/broken CSS 2 properties would be excellent. I would also like to see support for spoofing custom user agent strings. Opera already supports spoofing of browsers, used as a last resort for users who are experiencing compatibility problems on websites due to that particular website’s developer’s stupid usage of browser sniffing and their ignorance of Opera’s existence. In many cases spoofing Firefox on a particular problematic website does the trick. However, for additional compatibility problems and for testing purposes it would be beneficial to allow support for empowering power users to input a user agent string to spoof or mask as. Even Safari supports this out of the box — although it does have to be enabled. It’s a feature I’ve wanted for quite some time, and surprisingly a feature other browsers already possess either out of the box or through extensions.

Final Assessment

What drew me to Opera initially was its speed. Opera’s competitors can claim to be the fastest by using JavaScript benchmarks, but none of its competitors can touch Opera’s speed in the real world. In my opinion Safari is the closest in browsing speed, and with its being 64-bit in Snow Leopard it might even be close or have surpassed it. However, no browser can touch the sheer feature list Opera posesses. There are many features which are nearly as old as the venerable browser itself is which I could not do without and which are missing completely from its competitors or are only available by downloading extensions which do nothing more than slow an already slow browser down. Opera 10 only adds to the long list of personally necessary features; with the addition of a properly designed interface and icon I won’t have much Opera-related topics to bitch about anymore, and that’s perfectly fine by me.

Download

Opera 10 — Best Browser on Earth.