In the article, Ben Goodger, lead Chrome UI developer, states

[Google avoids] cross platform UI toolkits because while they may offer what superficially appears to be a quick path to native looking UI on a variety of target platforms, once you go a bit deeper it turns out to be a bit more problematic.” [... Your applications end up] speaking with a foreign accent.

But there’s something we’re not getting here. Obviously given enough brilliant programmers and a good team lead to keep the different codebases in sync, going with native APIs is the better approach. But the reasons Goodger is offering aren’t very convincing.

The problem is…. Google’s Chrome for Windows doesn’t look native. In fact, it’s about as far from native Win32 as you can get. We had originally explained away the non-win32 looks by assuming it was because Google wanted an interface that was consistent across the different platforms and different at the same time from any of the operating systems native UI toolkits: in line with Google’s vision of turning the browser into an OS, regardless of the platform beneath.

A non-native UI that looks the same on Mac, Windows, and Linux would be the answer to such a browser OS. It would indicate that Chrome is its own product – from the codebase to the user experience – and that to the end user it shouldn’t matter what OS you’re on. And that in the future Google could ship a standalone (OS-free) browser that looks like Chrome and acts like Chrome, regardless of the platform beneath?

Otherwise there is no good explanation for the horrendously-different user interface that comes with Chrome. It requires learning the tips & tricks to a whole new UI, and forgetting a number of “niceties” you may have been accustomed to (such as pressing the ’spacebar’ to OK pop-up dialogs, etc.).

With the preliminary screenshots of Chrome for Mac, the platform Chrome runs on begins to peek through.

Does this mean that Google’s vision of Chrome as its own OS has come to pass – with Google now content to just launch a cross-platform browser without attempting to lull users away from the platforms they’ve come to love?

Whatever the case, it’s sure to be interesting watching and waiting to see what Google has planned for its users. Whether its a cross-platform browser experience that’s different enough to be the same across all platforms while retaining a feel of the platform or if it’s paving the way for the OS to come it’s quite obvious that the gears are now in motion and something big just might happen.

We’ve long felt that Gmail’s approach was not befitting of the Web 2.0 service with all its sky-blue shades and flashy appearance – and now it seems that Google’s felt that way too.

Here’s the new loading interface… Subtle, simple, and effective:

(Click image to see more changes)
  

After all, first impressions are everything!

OK, perhaps it’s not a stereotypical “____ Screen of Death” in that it’s not a fatal error, it doesn’t make your computer restart, and it’s not as scary. But it’s still caused by a low-level problem in the kernel, and it does make you want to restart your PC.

When it happens: if your WDDM drivers partially crash, and DWM doesn’t properly deactivate. It’s very rare

Basically, when a window attempts to “gray out” because it’s unresponsive, the graphics drivers screw-up and paint it purple. Afterwards, all newly-opened windows will show the same purple tinting. We’ve seen this behavior on both ATi and nVidia machines, it’s something to do with the way the Windows Vista kernel handles a certain exception in the graphics driver subsystem.

You can see the full-size screenshot and many more in our extensive image gallery, in the Vista RTM album.

Apple [[AAPL]] has a new technology due to ship in Leopard that is pretty damn impressive to hear some of the developers (or at least those at liberty to discuss what they’ve been told) talk about it. According to this Wired News article, this framework will focus on making it easy for developers to add Next (Delicious) Generation graphics to their applications, bringing the concept of UI to a whole new playing field.

While it sounds just plain awesome, we can’t help but look at it with mixed feelings. Since this is Apple and not Microsoft we’re talking about, we’re pretty darn sure everything they say can and will be found in the RTM release of OS X Leopard *cough* WinFS *cough*, but whether or not it will revolutionize the concept of user interfaces everywhere remains to be seen.

Granted, the effects are supposed to be astounding. But that’s not the question, next-gen UIs (hereby dubbed “Desktop 2.0″) are almost always a matter of how and not what. As the Wired article states, these same “amazing” graphical innovations with the UI do already exist in certain programs (like Disco, the CD burning utility for OS X), the only difference is, Apple promises to make it much easier to get the same end results.

Is this just another RIA framework like XAML or Flex/Apollo? Well, it’s a framework, no doubt about that. Anyone remember the mock-ups and sample programs released when XAML was still in beta (and before the blasphemous change in nomenclature from WinFX to .NET 3.0)? Those were damn impressive, much better than some of the effects in Disco and other programs that Apple is pointing at. But look around you, and you’ll notice that no one is coding these “amazing next-gen” interfaces anymore. Why? That’s easy: they’re too insanely difficult to create!

XAML + .NET is an amazingly powerful combination. But where .NET in general is classified as being a pretty decent RAD (Rapid Application Deployment) language/framework, XAML most certainly isn’t. With XAML, it takes days of work to get things to look just right. That’s why we still have a rectangle, a bunch of controls thrown onto it (whether MFC/Win32 or .NET), and standard boring-yet-gets-the-job-done interfaces all over the place.

But Apple is notoriously famous for making things easy… for users at any rate. The best thing we can say is, they certainly keep their word, especially when it revolves around APIs, new technologies, and software vendors (yet the obviously-incorrect Developers, Developers, Developers still comes to mind!). This is certainly something to keep your eyes peeled for and your ears wide open, because if what Apple says is true and they can really deliver awesome animations in an easy-to-implement API, then “Delicious Generation” may very well be the next big thing and the gateway to Desktop 2.0!

But Geni is missing something – the big picture. I can create a family tree and propagate it as well as I can, and sure, I can invite anyone ON the tree to join in the fun – just put their email in the box, and they’re invited to join the family tree and start expanding. But the problem is, it’s lacking the one thing that makes genealogy so exciting: finding someone new.

It’s built on the same concept as LinkedIn, hell, it’s the same exact principle to the T. You invite users to your personal “tree,” and they put the people they know. The people they know connect to the people you know and you suddenly know the people you never knew to know. That’s great, so I can invite anyone that I already added to the tree, and they’ll expand the tree to contain the rest of the family – or what they know of it.

What about searching though? You know, like, search for someone that put xx and yy as their grandfather and grandmother with at least a 60% match for uncles and aunts. Isn’t that what genealogy is all about? It’s not about writing down who’s who, it’s about the thrill of finding someone new. Technically, the way Geni is currently set up, you’re unlikely to find someone you never new existed.

It’s not that the technology doesn’t make it possible, but that in order for you to find someone you didn’t know existed, someone you know must already know that person. So the likelihood that your relative is related to someone that you’re related to too and you didn’t know about it is what counts – and that’s quite a slim chance there.

Geni is great. It’s addictive, it’s certainly better than MySpace & Co. and it’s a hell of a lot more real too. But until Geni adds the ability to search for family trees that coincide with yours and let you know they exist, it’s not taking advantage of even a tenth of its potential. Maybe Geni is waiting for more trees first, or maybe they’re just waiting for someone to post and ask. Either way, here’s to hoping this gets implemented – and soon!

For one thing, Netscape doesn’t offer anything that’s not already there, provided by the now more-popular social web homepage services, like NetVibes, PageFlakes, Live, Yahoo!, and Google. The most important thing to keep in mind is: users have already left. If they have no reason to switch back to My.Netscape, why should they? They can get equal/better services elsewhere on the web, they’ve already configured them, and they’re more or less happy there.

When My.Netscape couldn’t catch up with the times, they left it for the competing services. The competing services haven’t given them a reason to look elsewhere, so there’s no reason to assume they will.

As a matter of fact, Netscape doesn’t offer anything exciting, not even a half-decent layout. The programmers at Netscape ought to be ashamed of themselves, they’re just blatantly copying ideas (such as modules and widgets), design (just about everything except the grey theme), and technique (everything) from the competition, with nothing new in the mix.

A site can’t succeed without adopting some of the concepts already out there, but in order for it to stand out, it must also innovate. It doesn’t work well if you have one without the other, but just adoption without innovation is suicide.

However, personalized homepages are more than just skin-deep. People use Google’s decidedly not-anything-special homepage service because it gives them one-click access to Google Reader, GMail, and (most importantly) Google Search. They use Live.com because of the original content by MSN authors, cross-compatible Widgets, and Live search. They use NetVibes because of the powerful development framework that makes creating awesome modules that easy.

Maybe people will use Netscape because of the new netscape.com digg-like site. Or maybe they won’t. It depends, but you almost guarantee that the average internet user who is already using a competing service won’t be switching anytime soon – not when Netscape doesn’t give them a good reason to!

The reasons vary, but whether it’s lack of user privileges, difficulty of installing new keyboard layouts, not having enough time, or simply using an OS that doesn’t support anything but QWERTY, the fact remains that a software-fixed keyboard layout isn’t always the best answer. It certainly is a very flexible solution – especially on the operating systems that let you drag and drop letters around to create your own keyboard layout that acts exactly the way you want it – but it’s not necessarily the only one.

It wouldn’t be difficult to create a portable (or not so portable) USB keyboard that types the way you want it to no matter what platform or configuration you’re on. While the OS itself is still configured to use the QWERTY layout, by using your alternative layout keyboard, pressing F gives you the U you really wanted. It’s as simple as changing the scan codes for each keypress; for example, making F return 42 instead of 53 – all invisible to the PC and the OS.

Imagine a keyboard with a scroller on the side with a little analog screen next to it. You can scroll to the layout you want and type – no BS involved. Even better, imagine a tiny cylinder that sits between your USB port and the Keyboard - with a scroller on it. It can (on the fly, of course) intercept outgoing scan codes from the keyboard and replace them with the right thing according to the layout you selected… How cool would that be!?

Core Features

• Ability to remotely upload a plugin (as zip file??) through the interface without having to FTP or SSH into the server.
• An “update detection” mechanism that can be manually or automatically run to check for updates for a plugin or more than one plugins and report back.
• Automatic-detection of “corrupt” or broken plugins – so you don’t have to FTP into the server to delete a bad plugin.
• Remote plugin management: delete plugins from the web.
• “Debug” Mode: All plugins are disabled. The key here is, once Debug Mode is turned off, your plugins go back to the way they were – some enabled, some disabled, just like you left them.
• Automated hooks for installing, uninstalling, and upgrading from previous versions.

Key Ideas

• Security. Don’t allow automatic updating of plugins because you have no idea what changes have been made to their source.
• Ease-of-Use. Never have to remotely connect to your server to install or remove a plugin.
• Stability. Bad plugins won’t fry your server (a la WordPress); you don’t have to frantically SSH in, and cd recursively into the directory just to rm the file.
• Compatibility. Define a specification that can be used by plugin authors to define for what versions of PHP, MySQL, and Habari their plugin(s) work and for which versions they don’t. Even allow for custom pre-requisites via a “compatibility check” function in the plugin. If a plugin specifically states it doesn’t work, don’t let it install.
• Bleeding Edge. By means of the update detection service, there is no reason for someone to be running an out-dated plugin.
• Automation. The first time a plugin is detected, it will call a install function in the plugin. If a newer version of a plugin that was previously installed is found, it’ll call an upgrade function in the plugin, and if a plugin is queued for uninstall, the uninstall function will be launched. These functions may have user interfaces, use other private and public methods, etc. but should be complete and self-sustaining: i.e. no secondary install steps will be needed. Configuration is different from installation.
• GUIDs. Each author, upon registering his or her plugin with the central plugin repository, get’s a unique GUID for one or more plugins. This GUID is unique to each plugin, and remains constant throughout all versions of the plugin unless it’s being shipped as a different product. This feature is used in the install/remove/upgrade automation.

These ideas and features are just the beginning of what’s possible for a plugin administration interface written from the ground-up with power, control, and end-user simplicity in mind. They mean nothing without a proper UI, and some good code to back it all up. But it’s a start.

These are not just “pie-in-the-sky” suggestions and day-dreams, this is all brain-storming for a very real and very tangible project. What do you think would make the ultimate plugin administration interface? What ideas are we missing here? What would you like to see in such an interface?

Although the behavior of margin- and padding- is very similary, there is one important difference: margin- is on the outside of the element, and padding- is on the inside. What this means is, one will apply the whitespace shift “after” the element begins (padding), and the other will apply the whitespace “before” the element begins (margin). Many times there is no visible difference, but sometimes there is – a big one.

• Colored backgrounds. Imagine a div tag with a colored background. If you apply “padding-top: 10px” you’re going to lower the text by 10px – but this “lowering” is going to take place inside the colored div, which will start in the same place as it always did. In the end, your colored section will be 10px larger. If you had used the “margin-top” property instead, you would have lowered the entire div, and the text along with it. Your colored section would not change in size.
• Negative values. You can use “margin-left: -10px” and end up moving an element 10px over to the right – actually making it float on top (or under) the neighboring element, because you’re changing where the element begins (-10 from the left). But you can’t use negative values with “padding-” because then you’d be telling it something like “start the contents -10px before the container for this text even begins;” which obviously isn’t possible. You can drag the entire element over to the right (margins), but you can’t place the number 2 before the number 1 (padding), that’s just illogical!

That’s all you need to know about padding and margins, but it’s useful to remember one more thing: when “zeroing out” a element (to get rid of all default margins), be sure to set both the padding and margin to 0. Depending on which browser interprets your page, it may place the default spacing as either a margin or a padding, and you have to be sure you get them both.

It’s still a Ford, though it looks more like a cross-breed between a Mustang (the top of the car), a Cadillac (the overly deep body), and a BMW (the wheels and the rear – from what we can see of it). It looks rather eerie, and only too fragile.

According to MSNBC, this is part of a new plan by Chrysler, Ford, and General Motors to attempt to regain market share lost to Japanese automakers. It seems they don’t understand it’s not just about the way things look, it’s also about how reliable they are and how good of a name they have to back them. Give the average American a choice between equally priced BMWs and Cadillacs, you shouldn’t be surprised by the result.

When Firefox 2.0 came out, we didn’t really care to review it – after all, there were plenty of reviews already out there from the Beta and RC stages. But now, a month into the RTM release of Firefox 2.0, we find a re-cap being called for.

The “New & Improved” Theme?

We’re not going to delve into this too deeply, suffice to say that when we first said Firefox’s excellent theming engine’s only shortcoming was the lack of a decent default theme we didn’t think they’d actually change anything. What we really didn’t expect was some washed-out toolbar icons that looked like they came from Internet Explorer 4 with some “Web 2.0” shading (gamma circles) applied. They just don’t fit Firefox!

Ugly, New First-Run Website?

Since the release of Firefox 2.0, the first-run homepage has undergone several iterations – all of which are absolutely appalling compared to the original “Welcome to Firefox” page that used to display in the old 1.5 series. To tell the truth, it looks like someone handed a 3–year-old a bunch of vector-drawing tools, and told him to ‘go Picasso’ – without the slightest idea regarding color theory, aesthetics, or knowledge of the generally-accepted location for body parts – except it didn’t come out quite as nice as Picasso’s work did.

Who blends orange and blue? It turns a ghastly shade of green, and especially when you add to it the improperly done “wave lines” it’s just a complete eye-sore to be showing around to someone turning on Firefox for the first time. Compare that to the default Internet Explorer 7 page: clean, aesthetic, and simple! But the problem is, that start page was changed several times since 2.0 was released, yet it still looks appalling.

Lack-Luster RSS?

Firefox was one of the first browsers to take RSS to the next level, and is largely responsible for the ubiquity of feeds today. But for some reason, Firefox has serious issues implementing a decent RSS reader into the browser. Before 2.0 you could view the headlines for RSS feeds and click-the-link to view the rest. Firefox 2.0 has an “improved” RSS reader that finally displays full-text articles for feeds. But it still lacks the ability to filter out entries based on categories and tags, nor dynamic searching of feed entries.