Warning: this is a nerdy code post, so proceed with caution

While modding my new site in Community Server 2007, I needed to know the item number in the asp:Repeater control so when I reach the last one (the repeater had a fixed number of items) it would not apply a certain CSS class so there isn't a redundant border. I didn't want to inherit the Repeater control and implement the logic in the Render event because I wanted to do it declaratively (because I was lazy basically). It was not immediately obvious as to how, and I still don't know why it is the case, but it works.

First, in the aspx page where the asp:Repeater control lies, add in the following somewhere:

<script language="C#" runat="Server">
//needed as cannot get Container.ItemIndex directly
int GetItemIndex(int index)
{
    return index;
}
</script>

Then inside the ItemTemplate tags of your asp:Repeater control, access the item number (ItemIndex) using the following code:

GetItemIndex(Container.ItemIndex);

You'll have to combine it with something else though because that code above does nothing with the number. Something like this to print it:

<%# GetItemIndex(Container.ItemIndex) %>

Or, this to use it conditionally:

<%# GetItemIndex(Container.ItemIndex) % 2 == 0 ? "print-this-if-true" : "print-this-if-false" %>

There's a ton of things you can do with this code, like create a new table row every 3 items, or make every 5th item red.

Why you can only access Container.ItemIndex when it's wrapped around a method I have no idea, but it didn't work until I did that. There are obviously some intricacies of ASP.NET that my hacking isn't picking up - I've never done ASP.NET coding before, just winforms stuff.

Hope this helps someone hacking through ASP.NET!

Some of you may know that I hate fanboys, whether they're Mac, Linux, Sony or whatever. I don't mind if your passionate about a particular company, in fact that's a great thing and if I startup, or work for a company, I'd be honoured to have such people. What shits me are those who are so blinded (or brainwashed) that they go around spreading complete FUD (fear, uncertainty, doubt) about other people's products. I don't care if they say, "Mac rules!" on an obviously Windows forum and wait for others to put him/her back in place. What gets me are people who say, "Windows Vista is just as buggy and insecure as XP, and there's nothing new. Get a real and innovative OS like a Mac. Micro$oft is EVIL." It's obviously untrue, yet they still feel it's necessary to plaster it over every blog they come across. Geez, if your product's any good, you won't need to sling mud over at your competitors - people will want it as soon as they see it. There's a fine line between playful jabs at your competitor, as Apple does so bloody well with their advertising, and blatant mud-throwing, and unfortunately fanboys can't see the difference.

Why am I rambling about this? I was reading a slightly unrelated post over on Scoble's blog, and saw an exchange between obvious Mac fanboy 'rory' and Scoble in the comments.

No, it is *NOT* DOJ to ‘blame’ it is the ones who made the piece of trash called ‘vista’….

Sure, the extra SW is no fun, but easily wipeable…

Vista is warmed over XP, and has the same security issues, just with a new face…..

Comment by rory — April 6, 2007 @ 1:02 pm

rory: if you think that Vista is only warmed over XP, that demonstrates you simply don’t know what you’re talking about. Vista has a completely rewritten networking stack. A completely rewritten audio stack. Substantial changes to UI. A built in search engine that actually works pretty well, unlike the ones I tried with XP, much better Tablet PC and Media Center features, and a ton of other stuff. And, it was completely recompiled with a new compiler that blocks buffer overruns. There’s a ton of security work that was done, too. Including that it’s not running in admin mode by default.

But, no, it’s just “warmed over XP.” Yeah, right.

Comment by Robert Scoble — April 6, 2007 @ 1:14 pm

Reading Scoble's reply made me realise what Vista actually did for Windows - it brought Windows on to the same level as Mac's OSX Tiger. With respect to operating systems, there is now a level playing ground. Vista now looks as nice as OSX, if not better in parts, works just as well, and now has the right infrastructure to build on to make it better than OSX.

Talk about Vista being just as insecure is complete crap. Vista will always have more security issues as Macs, because there are millions more Windows users than there are Macs and perfect 100% secure software is impossible for something this complex. Just because security issues on Macs aren't as widely publicised, doesn't mean there aren't any (check out what software updates you're getting - majority are security fixes). If in a parallel universe Macs and Windows swapped places, Macs will be in just as much shit. Of course, from a consumer perspective, they couldn't care less about the economics of things, more what they get from switching. And it's true then, that Macs have much less virus and spyware attacks. But what it's really doing is giving the user an excuse to become complacent about security. They're lulled into a false sense of security, and everyone knows a false sense is worse than no security.

We're worried about our physical security - we lock our front doors, have alarms & immobilisers installed in our cars, and have lighting to ward off potential crooks. Why aren't we worried about online security? Windows users are gradually being educated to be as vigilant as we are with our physical security. Macs on the other hand, give users an excuse to not be vigilant, purely because there are less of them. That's like saying Lexus cars don't need alarms and immobilisers because there's more Fords, Holdens and Toyotas out there. Sounds crazy doesn't it?

The one area that Macs have a one-up on Windows right now is hardware design. Macs are cool, well designed, and now, very competitive on price. I can't for the life of me understand why there's no cool looking Windows laptop manufacturers out there; surely Apple didn't hire every cool design engineer out there. Like I said in a previous post, people are now buying MacBooks then installing Boot Camp and running Vista on them. Another mate told me he bought a MacBook because he wanted something small, but in the PC world, the smaller the computer the more expensive it is. It's the opposite in the Mac world.

Until the next iteration of OSX, Leopard, comes out (in the next few months), both OSs are on pretty much even ground. Neither has anything particularly better than the other, and in that respect Microsoft has done a good job. Of course they need to start innovating very soon, but bringing XP up to speed with OSX was a big effort because OSX was leaps and bounds ahead of it. Apple has given Microsoft the jolt it needed. I expect Microsoft to release new Windows features more frequent now, so it can keep up and make Windows better than OSX. Whether they release a new Windows more often (as is the OSX trick), distribute/sell features via its Windows Update site, or distribute new stuff some other way I don't know, but they've laid the groundwork so the battle between Vista and OSX should get quite interesting very soon.

P.S. For another perspective on the differences between Vista and OSX, check out this article by an MSNBC reporter, who switched from a Mac.

While playing around with Paint.NET today (great free basic photo editing tool, essentially paint on steroids), and was pleasantly surprised when I got this as I went to close it:

Of course, if you have Windows Vista, you would've seen the former type everywhere, e.g.

To get one of the newer taskdialogs, takes much more than a line of code. There doesn't seem to be any managed code wrapper, so you'll have to dive into C++ or use P/Invoke. Unofficially, there's this library at the soon to be closed GotDotNet site, not sure how good it is though. Maybe I'm missing something here, or maybe it's coming out in .NET Framework 3.5, but it seems weird that there's no managed wrapper for this given that Microsoft is pushing managed code so much these days.

There's been talk of the browser being a dying breed when RIAs come along, and while I agree partially, I still believe there's a place for it in the short-mid term for browsing webpages, and information, as well as for testing out RIAs. I mean, imagine if you had to download a desktop app everytime you wanted to try out a new web app. It may not be in its current form, though.

The guys behind Trillian, the multi-platform instant messaging client, have created a Flash based version of their instant messaging tool, but given it the ability to break free from the browser once its launched, allowing it to act as if it were a desktop app, complete with tray icons. Check out their blog post, along with the demo video - it's a pretty cool example of the blurring lines between the desktop and the browser.

It works by leveraging Firefox as a platform, installing a Firefox plugin that allows it to break free from the browser. There are some security issues probably, e.g. how to make the plugin onlt allow the Trillian client to 'break free', and not spyware/advertisements, especially because it seems to have pretty deep access into the computer.

It's a demo of things to come though, and really just morphs Firefox into Adobe's Apollo platform, which is pretty cool and should please the open-source fans, plus it'll allow more than just HTML/JS/Flash, unlike Apollo.

No this isn't a post about how to use the current command line interfaces, which are power and cool, but insanely unintuitive for the average joe.

This post is about the command line making a comeback, after being shoved away into a corner by its bigger, better, more popular cousin, the graphical user interface. I stumbled on to an article titled "Command Line for the Common Man: The Command Line Comeback" [via SvN].

What it says is quite true - slowly, but surely the command line is becoming more relevant to our lives (assuming we're not all geeky linux nerds, or poor sysadmins trapped in a dingy room surrounded by servers). Not in the form that we all despise, but it has morphed. Chances are, you won't have even realised you're using it. Website URLs are one example, or when you're searching google, especially when you use its oneBox keywords.

All the major operating systems have built in search engines too - Vista, Windows Desktop Search; Mac, Spotlight; Linux, Beagle. And utilities like Quicksilver for Mac, and Launchy for Windows take this concept even further.

I agree with most of the article (except when it used Office icons as an example - check out Office 2007's new ribbon; obviously Microsoft's picked up on the ambiguity of some of its icons). It is very powerful, but the problem is remember the commands. Look at the weird command utilities in both Windows and Linux - hardcore users aside, you'd be hardpressed to find someone who could tell you what all the various switches for the copy/cp command are. Unleashing this power to the user is a great step forward, but typing and talking aside, which is what the article suggested, I think an even more powerful way to harness the power right now (i.e. without the perfect language understanding that computers will have in the future), is to offer Intellisense style prompting. Better yet, help the user 'build' the command string by showing them a list of common commands and letting them pick the required switches. Imagine if you took the CounterStrike style of numerically numbered menu items and implemented it here - you could just press 1-2-5 and have a blank email ready to be typed in. So much faster than typing in 'open blank email'.

Words still have a role in it of course, especially when you're searching for something. But menus ultimately save users that much more time by not having to remember the keywords. Then again, google's already shown that it is possible to parse a lot of our natural language requests into actions, so it probably won't be long until this picks up. The command line won't get to exact the ultimate revenge to its GUI cousin, but it'll make the playing field much more even.

The whole thing about whether Flash should be open-sourced has come up again - Ryan Stewart's summarised it pretty well on his blog post.

One argument for open-sourcing is that such an important platform (and it'll be getting more and more important) should not be controlled by a single, commercial entity, who god forbid, may decide to charge for it in the future. This argument used to apply mainly to the evil Microsoft empire, for its various formats and players, most recently being HD Photo. Its an argument that the open-source community wheel out for almost everything proprietary. In some cases, fine maybe it is a valid threat (e.g. Apple and Quicktime). But seriously, Adobe? It makes millions of dollars in sales from selling its Flash development environment, and the player has always be free for PCs and other platforms like Mac, Linux, Solaris (not sure for Flash Lite, the platform for mobile devices). Along with PDF, Flash is one of the most widely distributed multi-platform players. It is one of the few platforms, Java and HTML/JS/AJAX being the other big ones, that runs the same across all the popular operating systems. There is no reason for it to stifle this growth and adoption - it'll mean death to nearly its entire web push.

Now, assume that Flash became the new HTML, that is, it became the new language for the web. WPF/E and AJAX trail behind. Adobe is now in a position to charge for Flash, and extort millions more from everyone who's web experience is dependent on Flash right? Not necessarily. A certain gullible set of people will, as well the rich ones and others who need it desperately, but the vast majority of us will remain on the last free version of Flash. The web development community will realise that, and design accordingly, and in the long run,  change tact and use the various other free options, like AJAX, WPF/E, OpenLaszlo, Morfik (just realised they're an Aussie company, and of all places, located in Tassie - good on ya guys!) or the many other frameworks that will pop up, open-source or not. Web development moves that quickly anyway, look at myspace - its moved from perl, to ColdFusion, to ASP.NET now. New sites have popped up, old ones have died or been reinvented. So who loses out in the end? Adobe.

Then there's the argument that a single company controlling it prevents real innovation. This is true - the brainpower of the community is far greater than the brainpower of a single company. Adobe could do better to openly ask and faciliate various requests that web developers have (first thing that comes to mind is a digg-like site, like Dell and Yahoo have done, but surely Adobe could do better than that). But there in lies the problem with open-source. Developers are a cocky and tricky bunch, especially open-source ones. If they believe enough in their idea yet it gets rejected by the leading team (how is that decided?), or if they have a grudge or any other reason, they can easily create a fork of the open-source project and start going in their own direction. Nothing can stop them. Ok, maybe I exaggerated slightly, but it does happen. Not a very good example, but look at CMSes joomla and Mambo.

What's wrong with this? Well, the brand and platform is diluted. You have two very similar but slightly different projects going on that are mostly compatible, but begin to grow further apart. That results in confusion for developers, who have to work out which one is better, and more importantly, for users, who already couldn't care less about what Flash is. We'll end up with a situation like Linux at the moment - with hundreds of different distributions all vying for the users attention, all slightly different. The user installs one variation of Flash, only to go to another site and need another variation of Flash. Or worse, they'll go to another site and it'll turn out mangled because of their particular version of Flash, the user thinks its the site's fault and never returns. Confusion reigns, and sooner or later there's be subtle conflicts occurring between the installed plugins resulting in weird browser crashes. With so many incompatible projects, it just destroys one of the biggest advantages of Flash - its simplicity and cross-platform compatibility.

Ok, how about if we standardised Flash, then we can have many open-sourced Flash players out there yet everything will still work. Well, as all web developers will know, you only have to look at the disaster that is HTML, CSS & JS to see what the problem is. You'll end up with subtle incompatibilities between various Flash players so web devs will have to start hacking around and adding player-checking routines. Some will implement specific-player only functions, others will implement them in a different way. Ultimately, standards, just like laws, are just words on paper. There are hundreds of different ways of turning those words into reality, and with no binding agreement that a certain standard and its updates have to be followed, there is no way to ensure standards are followed in exactly the way that the words were intended.

Don't get me wrong, open-source has done wonders for the software world, and there are many great open-source projects out there that provide great alternatives. But some things just weren't meant to be open-sourced, and that applies to platforms, whether it be an application platform like an OS, or a web app platform like Flash or Apollo. Sometimes, its better if one entity has control of it, because at least then we know that there'll only be one version to code against. The more open that entity is, the better, but open-source is not the answer to everything.

That said, it'll be interesting to watch how the new open-sourced Java will eventuate. And it's worth noting that PHP and python are both open-source languages, yet seems to work quite well without too many forks. Maybe its just a big risk that may never turn out to be anything bad for the platform, but are the advantages appealing enough for Adobe to take that risk?

P.S. Forgot to mention that there's actually an open-source project going on to replicate the Flash player, called Gnash.

This is the last post in a series of posts about making Linux a more competitive in the consumer world. The first post, part 1, introduces the reason and some of the problems and solutions, like lack of innovation and boring UI design. Part 2 explores more solutions, including eliminating the restricted formats issue, improving the out-of-the-box experience and connecting the web browser to repository apps. They're not requisite reading, but worth reading I reckon (because I wrote them ).

• Compilation is a no-no. I should never have to compile anything that's released to the public (unless I'm doing something that absolutely requires it, like coding specific custom functionality, but then this is moving in to the geek domain). They should all be as easy to install as downloading them, launching them, and answering some questions if really necessary.
  
  This problem occurs mainly because when developers release apps, they don't have the time or resources to compile it for all the popular distributions out there - ultimately, there'll be people who'll miss out and the only way to get the software is to compile it themselves. That's why the recently publicised plan to simplify software installation across all supporting distributions by the Linux Standard Base project is such a great idea. "[ISVs want] to treat Linux as a single platform, which means they want to offer a single package for Linux, much as they do for Windows."
  
  There is also a time lag between the developer releasing a new version and the distribution's repository having an updated version - a time lag that sometimes people can't wait for (and for which there is no known timeframe for), hence forcing them to try compiling it. This is a disadvantage with the repository style of distribution applications that can be reduced if the developers ran their own repository source and the above LSB project succeeds - devs will then only have to do one release and all supporting distributions will have the newest release.
  
• Create alias for the file system root folders. What does /etc, /dev, /var, /bin, /root etc. mean? If I wanted to install a program manually (because the program doesn't come with an installer) where do I install it to? They should only have to deal with the home directory ideally, but these people are hobbyists, tinkerers - they will look around and the more you let them know what they're doing, the better. I understand it's a Unix legacy, but a better solution would be to create alias for each of those folders and hide those folders in the GUI. For example, instead of /home, they could call it /Users. Instead of /etc, they could call it /Configuration Files.
  
  
  
  
• Improve interoperability with non-Linux computers. In both Gnome and KDE in OpenSuSE 10.2, the shell smb:// support is still non-existent in most apps, e.g. Banshee, so the only ways to access files on another computer is to copy it to a local folder, or use mount in the terminal. Of course, Nautilus and Konqueror, Gnome and KDE's file manager support this.
  
• Reduce or hide choice. Geeks will scream at me for saying this, but when even the basics can vary so much, it becomes,
  
  a) very difficult and time-consuming to work out what I want (is Ubuntu better than Fedora Core, or is Gnome better than KDE?),
  b) very difficult to trouble shoot, especially over the phone, and
  c) confuses people when a certain program is not available for their distribution. The nature of the Linux ecosystem makes this very difficult but without any sort of across-the-board standard, trouble-shooting becomes very frustrating (e.g. only some distributions have pump & sax2, others have equivalents), and instead of knowing how to work Linux, you only know how to work a specific variant of Linux, reducing the community's ability to help everyone.
  
  Some sort of standard that dictates what command-line utilities must be included would be good too across all the main distributions. You can put more in obviously, but those basic ones must be there, so for example, there'll be no need to go digging around to work out what the command to renew a DHCP IP address is for every distribution.
  
  Also, distributions should be made such that you can change distributions without having to reinstall everything. If you give hobbyists the choice, they'd want to be able to switch. But if you force them to reinstall everything every time they want to switch distributions, they're not going to be happy.
• Stop labelling applications as for KDE or for Gnome if they can run on either (which most can). That labelling just confuses people, who are already struggling to grasp what KDE and Gnome are anyway as there is no equivalent in Windows or Mac (unless you count the terms Aero and Aqua respectively, but they're hardly used).
• Make the Linux Kernel version prominent. This will confuse users, and therefore some explanation is in order, however it is necessary because the only constant that hardware manufacturers can rely on is that for their hardware's system requirements, unless they're willing to test their product on all the popular distributions out there and list all of them.
• Finally, don't aim to make it as good as Windows, Mac OSX or anything else out there. That's a milestone. Make it better, unique and easy to use in every way and you'll have a winner on your hands (or in everyone's hands).

So Linux people, stop wasting your time spreading FUD (and distance yourselves from the communist fanatics at FSF who can't sell something even if their lives depended on it) and start making Linux a viable alternative to Windows. Its not a war, its competition - use your competitor's resources to leapfrog them if you can instead of reinventing the wheel.

For those who can't seem to be able to shake off the FUD spreading and the microsoft-bashing attitude, take this from someone within your community (its in relation to the recent Novell/Microsoft deal, but its applicable in general too) -

"I understand that many people don't like it as Novell is collaborating with the 'evil empire'. But I don't like this way of thinking. We are not working against somebody, but we are working for Linux. Fundamentalism always leads to pain. I think it is a good thing, especially for the users."
-Hubert Mantel, co-founder of SuSE Linux (quote from CNET, emphasis theirs)

This post is part of a series of posts about stuff that can be done to make Linux a more formidable competitor in the consumer computer world. Part 1 introduces the problem, and covers some points, including more innovation and designing better UIs. Here are some more:

• Make restricted formats (like MPEG2, MP3, AAC) easy. I know there are legal issues (as reverse-engineering MP3 is illegal in some countries, and the GPL doesn't allow binary-only package linkages), but as a consumer - I don't bloody care. Don't try to educate me, its not my problem. If I can't play my MP3 music collection on Linux, why would I switch? If I can't play my DivX movies or watch digital TV, why would I switch?
  
  This is one of the most asked questions on the net, but there's still no easy way on any distribution I've tried (no, searching the net for the right page, deducing the jargon in the instructions, then finding & installing some packages is not good enough). And by doing so, does that make me a criminal? Other OSs don't make me feel like a criminal for listening to my music or watching TV (yes, there's the Microsoft and WGA/Activation issue too if you want to argue but its irrelevant here).
  
  The best proposal I've seen so far is the Ubuntu solution. All desktop distributions wanting users should have something similar if they can't work around the problem otherwise. And no OpenSuSE, including RealPlayer is not a solution as it can't manage my music, and limits me to using RealPlayer only. Installing Helix Banshee to add music management functionality is a pain, and woefully underpowered, not to mention slow.
  
  While we're at it, if you want people to use your F/OSS stuff, get rid of stupid, arrogant attitudes like this (from Theora's FAQ):
  
  Q. Why use Theora?
  It's open and free. Do you need more reasons?
  
  
• Include binary-only drivers like the official nVidia and ATi drivers (not included as their source is not open, hence not GPL or compatible). Again, as a consumer - I don't bloody care. It's especially bad if my graphics card is crippled as a result, e.g. the F/OSS nv driver does not support the 3D features for my 7300 GT. I don't mind an 'encourage F/OSS software' attitude, but if you're crippling my hardware just so you can uphold your belief, you've lost me.
• Put a GUI to anything a user is likely to have to use more than once. Average people should not have to deal with terminal only apps. Some argue that that's just the Linux way of doing things. Geeks like it because its powerful. But its stupid for average people. Many studies have shown that most people generally interact better with graphical interfaces rather than command-line interfaces (the extreme being vi - yuck). Mac and Windows (well Mac did, then Windows copied) worked this out ages ago. Don't get rid of the terminal, just make GUI alternatives. It has and will continue to get better, but GUIs need to be factored in during the planning of an application, not as an afterthought.
  
  Also start promoting the GUI alternative in HOWTOs and forums. That way people are more likely to remember what to do next time or when there's a similar problem.
• Adopt the Ruby on Rails philosophy - convention over configuration. Most users just want something that works. They don't want to fiddle around for hours just to get it working. They couldn't care less about being able to tweak the tiny details. If its really necessary to be able to configure every last detail (which nerds love), don't make that the first configuration stop. Make it the last. Hide the tweaks.
  
  Here's a nice comment from a blog post I was reading that's representative of the current Linux situation: "Remember folks, linux and other complicated "free" solutions are only free if your time is worth nothing."
• Users need hand holding. Create wizards, tours, guides, tips etc. Hide error message details and give instructions on either how to fix the problem, or what the best recovery route is (unlike this from Vista/Expression Blend). As geeks, we hate them and think they're an obstruction - give us a way to disable them. But for others, make it default. Wizards, especially wizards with good explanations of the options presented are scarce in Linux. Firefox and the GIMP are good examples, and to a lesser extent, Evolution.
  
  

  The GIMP	Evolution

  
  
• Connect the web browser to repository applications like Synaptic. The web is ultimately a much better way of promoting an application than in repository applications. Create a new protocol (like skype, emule etc. did) and when a link of that protocol is clicked in a web browser, the specified package is automatically added to the repository application's download list (adding the repository to the list of sources if needed).

Spend enough time on digg, reddit, slashdot, and other similar sites reading tech news and you'll no doubt will see a Linux vs. Windows battle in play (not to mention PS3 vs. Xbox 360, Mac vs. Windows etc.).  The ferocity of these debates is amazing, often generating more comments than more real issues, such as the war in Iraq (hell, they're just operating systems goddamn it!). Supporters will tell you that Linux will take over the world when [insert random event here] occurs, like when Vista is launched very soon. Why hasn't it done so in the past? Will it in the future?

I'll try to tackle this as objectively as possible, but I'll admit I'm primarily a Windows user. That said, I have tried multiple Linux distributions before (Mandrake, Corel, Red Hat) and I've currently got OpenSuSE 10.2 tri-booting on my computer with XP and Vista, as well as Ubuntu Dapper with MythTV dual-booting with XP on my HTPC, plus a Linux NAS (Synology DS101) and a router (Linksys WRT54G with DD-WRT firmware). I believe in using the right tool for the job, so I'm planning on taking up Python to even out my programming skills.

Its also important to note that this series of posts is not about fueling the debate - I'm over arguing over which one is better because most of the time there's no way of changing their views and it just becomes personal.

Before going any further, its necessary to generalise and split up the world's computer users into groups. Below is a diagram of my theory.


Linux has conquered geeks, is slowly permeating through fixed-task (home) users as their geek friends convert them without them knowing/caring, and slowly permeating through fixed-task (business) through distributions like SLED/SLES and Red Hat. However, it hasn't made much of a dent in the largest and fastest growing market - the hobbyists. Right now, Linux isn't really a viable alternative to Windows for most of these people because Linux developers are not thinking like average consumers. In this series of posts (broken up so this post doesn't become too long), I'll run through some things Linux can do and should do if they want to make their OS more prevalent in the consumer world from personal experience.

• Stop plastering the phrase 'open-source' everywhere and force-feeding everyone the open-source philosophy. We don't bloody care. (I think there's a place for both open and closed-source, but that's irrelevant here.) Tell us that because it is open-source, its free, we can contribute ideas, and help make it better (although this is not unique, but better facilitated) - that's fine. None of this 'free as in beer, and as in speech' stuff everywhere - put it aside on a special page.
• Ramp up improvements and innovation. Create unique things. Improve existing things. Popularize it. Do what Microsoft has done with XMLHttpRequest (AJAX) and Tablet PC, and the upcoming Windows Sideshow (example) and WPF (example) technologies. Do what Apple has done with iChat and iMovie, and the upcoming Time Machine and Core Animation in Leopard. Xgl & Beryl are a good start, as is the new KDE start menu, but there should be more user-facing innovations.
• Stop bagging Windows about its security. Hackers love it because of its wide-spread usage, and 3rd party applications and social engineering contribute a lot too. If in a parallel universe, Linux and Windows switched places, Linux wouldn't be much better. Linux has exploits and patches now; they're just not mass reported by the tech and mainstream media unlike Windows. Every one makes mistakes. So just keep quietly releasing those patches - you already have a big advantage as the repository system keeps other apps updated too.
• Xgl & beryl are great, and are way better than anything I've seen on Vista and to a lesser degree, Mac OSX. But in terms of application UI design, its still very boring. As media players are usually the showcase app, here's a comparison between the three main desktop OSs:
  
  

  PC - WMP 11	Mac - iTunes 7	Linux - Helix Banshee
  	Photo by swanksalot from flickr

  
  Gnome/KDE themes make this a bit better (getting new themes should be easier btw), but the UI design in most applications is still back in the old days. Make the UI beautiful, usable and unique to Linux. After all, to capture this group of people apps need to be good-looking; features often come second.
  
  This isn't to say that all popular Windows apps are good looking - you only have to look as far as MYOB to see that - must be one of the ugliest ones ever; its still stuck in the Windows 3.1 days. Usability isn't much better.
  
  
  
  Screenshot from cnet

I've been pondering how to make all the various web services we use integrate better together. With the new trend being to create social networks for particular verticals (mums, fitness etc.) we'll all inevitably have a main general social networking site (like MySpace or Facebook), along with a range of other social sites and services like flickr, del.icio.us and blogger. Right now, the main way to post content or use features on these sites is to open up a new window and copy and paste the link. If sites want to provide better integration, often they provide widgets that other sites can use but the success of that ultimately depends on how well the other sites co-operate.

So how could this experience be better? How can you make your site look as if its integrated with another sites yet still have complete control? One answer is to create a browser plugin/extension of which there are plenty already - Firefox's addons directory even has a category for website integration. But this solution is annoying as the user has to install something which they may not be bothered to do (also involves a browser restart) or may be worried about given all the publicity about not installing anything from the net.

A better solution use bookmarklets, but create the interface as an overlay over the current page. Bookmarklets are simply bookmarks with javascript code instead of a URL, so just use the javascript to create some new elements on the document. Your code can access all the elements on the page that the viewer's currently seeing, and you're not interrupting what they were doing at all. Overlay the site with a small window that has your site's interface on it, allowing the user to utilise the content on the page they're viewing to contribute to your website. The overlay point is important, as most bookmarkets out there currently grabs the required content from the current page, gets rid of the current page and loads something else (or uses a new window which interrupts the user's flow). If its necessary to highlight the overlay, use the lightbox effect.

From a user's perspective, all they have to do is to drag your bookmarklet to their links bar once (which is unfortunately not visible by default on IE7) and whenever they want they can just click on that, without affecting the page they're currently browsing. It will appear very slick and not intrusive at all; in fact it could easily emulate a lot of what a browser extension can do, and is not restricted to the browser's UI design. The possibilities are endless, including:

• posting a photo on to flickr
• posting a blog post to blogger
• submitting a new item to digg
• checking out what others think of the current site
• finding similar sites and loading them in background tabs
  
• adding something to Google Calendar
• scribbling down something they just remembered (related to the current page or not) to Google Notebook
• finding out how far away a certain address on the page is using Windows Live Maps
• adding the current site's RSS feed to portals like Netvibes or offline browsers like AvantGo
• sending a link and quote of the current page to a mate using Yahoo Mail

The issue with this technique is that when you need to submit data, you run into browser security limitations, namely cross-site scripting (XSS). XSS is a risk because malicious sites can capture confidential information (like usernames and passwords) and send it back to base for future use or harvesting, so browsers instigated a 'same-domain policy' - that is, scripts and pages from different domains can't access each other. Javascript injected using bookmarklets (including external javascript files) however, can access the current page without an issue as they are assumed to be in the domain of the parent page (even if they're not), but they can't use AJAX to submit data back because the AJAX URL is in a different domain. An alternate method is to use HTML's form tag, or to pretend to add a script or image to the document, but make the source URL a URL similar to how the FORM GET method works, e.g. make the source URL http://somedomain.com/datasubmit.aspx?name=John%20Doe. It doesn't matter if nothing is returned - the request was the important thing. The problem with this is there's a limitation to the length of URLs in most browsers (smallest is IE with 2083 characters).

UPDATE: Another option maybe to look at using the bookmarklet to load a flash movie, which could then traverse the cross-domain issue. I'm not sure if this will work, given that flash has a cross-domain policy too.

Alright, enough talk. Looking around, the best example of this technique is the one on BlueDot which is a social bookmarking site (great site too btw). My blog's not letting me post BLOCKED SCRIPT links for obvious reasons, so go to BlueDot's bookmarklet install page, drag the 'Dot This!' button to your Links bar (the bar with the Getting Started and Latest Headlines buttons in Firefox; if in IE7 and there's no toolbar between the back/forward buttons and the star icons, click on Tools -> Toolbars -> Links), then navigate to another webpage and click the 'Dot This!' button you just added.

For the lazy ones, here's a nice video I found on YouTube demonstrating it, among other features of BlueDot.



These guys seem to have been able to work around the different-domain block of AJAX but unfortunately my javascript hacking skills aren't good enough - I can't seem to work out what their magic secret is. It'd be an interesting thing to know - someone fill me in!

More websites should offer this type of bookmarklet because aside from a browser extension, I reckon there isn't a smoother way to integrate a website's features into the user's browser so they can use it on any site they want. It'd be even better if the web browsers could work out a standard for securing bookmarklets so that it'll be easier to create bookmarklets like these - maybe only allow cross-site scripting if it was initiated via a bookmarklet, or make the user confirm/authorise the first time a bookmarklet is added, and then every time its changed.

P.S. Apart from the social features of BlueDot, another good thing about BlueDot is that you don't have to register before using it if you use it via the bookmarklet - just register the next time you visit BlueDot's website using the email address you used when submitting your bookmarks and they'll all be there.

I moved my digital TV card from my HTPC to my desktop recently after my HTPC died again . The card was detected and working fine in XP. Booting into Vista RC1 however, the driver installation was a mess because Vista first tried to install drivers from Windows Update but some how that went sour. The device drivers weren't installed properly and it didn't work. Rebooting Vista brought up the detected new hardware dialog again, so this time I tried the drivers from the card's website. These didn't install properly either, with Vista telling me that 'the system could not find the file specified'. Wow, how helpful. Attempting to update the driver didn't help - same issue, so I went digging.

It turned out that as soon as you use a driver, it gets stored into Vista's driver store. This store is protected so you can't go fiddling in it, even though its visible at c:\windows\system32\driverstore. I hate it when it shows me something but I can't edit it - opening Explorer.exe as Administrator didn't help. What I needed to do was remove the driver from the store so it would copy it again from where I downloaded the driver to (Yes, even if you specify update driver, if it thinks its the same as the one in the store, it will use the stored one).

I think I went the long way to solve this issue. The short way is to simply uninstall all the devices in Device Manager that relate to that physical device (e.g. the Tuner and crossbar devices for TV cards), and the driver should be removed from the driver store. In case that doesn't work however, I'll document the method I used below.

1. Open up c:\windows\inf\setupapi.dev.log in Notepad. This file is very useful as it logs everything that Windows does during the driver installation and tells you more specifically what the problem was if it fails. By default, Windows doesn't show absolutely everything, but you can change that using the instructions here.
2. All driver installations are logged here, so scroll down to the bottom of the file for the latest driver installation. Every time a driver installation is initiated, the following line is first written:
   

   >>> [Device Install ... followed by where the driver comes from and the device ID.

3. Find the line in this section that starts with !!!. This is where the error information is. I encountered the following error. If yours is different, this may not solve your problem, but there should be no harm in trying.
   

   !!!  ndv:      Driver Package import failed for new device...installing NULL driver.

4. If its there, find the following line in the section (where # can be any number) and make note of that path:
   

        inf:           Published Inf Path: C:\Windows\INF\oem#.inf

5. Go into your c:\windows\inf folder and see if you can find that inf file. If it doesn't exist, find a line in setupapi.dev.log in the section you located before that starts with:
   

        inf:           Driver Store location: C:\windows\system32\driverstore\filerepository\(inf file name and random string)\(inf file name).
   

   Make note of that path, find that file and copy it into C:\windows\inf. Rename it to oem#.inf where the # is the number you found in step 4.
6. Go to Start -> All Programs -> Accessories, right-click on the Command Prompt and select Run as Administrator.
7. Type in pnputil -d oem#.inf and press ENTER, replacing # with the number from step 4.
   

This should remove the problem driver from the driver store. Now try reinstalling the driver for the device.

This trick is useful for fixing issues if you try to install alternate drivers (e.g. unofficial drivers from another source) or drivers from Windows Update which do not install successfully. Its also good if you want to force Windows to use a certain version of a driver, instead of letting it pick for you.

When actually updating drivers (i.e. newer versions of the driver) though there shouldn't be an issue - assuming that the driver writer updated the version number Windows should realise that its newer and use it instead.

http://www2.jeffcroft.com/blog/2006/dec/17/iso-funny-acronym-for-flash-revival/

After web devs have been putting up with 'web standards' being implemented slightly (or in some cases, grossly or not at all) different in the various web browsers out there, a combination of web bandwidth and computer performance improvements have resulted in a new viable competitor, Flash.

Gone are the days when we'd have to wait 5 or so minutes just to wait for a flash site's menu to load, before having to wait another 5 when you select a menu item. With the bandwidth most net users have nowadays (there are exceptions, like my friends in Berwick who are still stuck on 56K dial-up), its feasible to have flash sites for the general public, that is, websites entirely driven by flash (in fact, a lot of the newer movie and car websites already are). There are still some issues with a lot of flash sites out there which can't handle bookmarking and browser history moves, as well as inhibiting search engine crawling and accessibility access. However, Adobe is addressing these and the advantages are apparent, in particularly, the fact that each flash renders in nearly the same fashion on every browser, across all platforms. No more ugly CSS/Javascript browser detection hacks.

And for those who argue about standards being better than proprietary formats - here's proof that standards are no where near as effective as proprietary formats. There are so many variations on web standards possible as more and more people interpret it, no browser out there is perfect (which itself is subjective) and there probably won't be, especially with something as complicated as a HTML/CSS/JS. Standards have shown to be useful in many instances, but it's difficult for something as fast-moving as web design.

(Even attempts at an open-source web standards rendering library, including WebKit and Mozilla's Gecko, aren't working out because the issue here is getting the browser devs to decide and use one. Given the work that's gone into both, and also into IE and Opera's engines however, that's unlikely to happen. The W3C, which maintains the standards also don't seem to be doing their job either...that or browser devs are just rebels.)

With Adobe's Apollo platform coming to take flash apps to a new level (along with Flex), as well as WPF and WPF/E arriving as Microsoft attempts to crack into the market, not to mention the open-source OpenLaszlo's effort (essentially an open-source equivalent to Flex), there's no doubt where the web is heading. With respect to Rich Internet Applications (RIAs), AJAX is currently the technology of choice, but it is really just an in-between technology, still bounded by the eccentricities of HTML/CSS/JS implementations of browsers. In fact, the humble web browser itself may not be the web browser we know and love(hate) in the near future.

The problem now is to wait and see an increase in crazy, annoying Flash animations and unworkable, confusing flash navigation systems, as flash abuse continues to grow, except this time, there'll be no turning it off...

More:
http://blogs.msdn.com/shanemo/archive/2006/12/22/designer-crunch-time-2007-the-year-of-the-rich-internet-app.aspx
http://blogs.zdnet.com/Stewart/?p=204

Well, after getting my Vista installation dual-booting with XP again, I needed a new challenge, so I decided to try my hand at Linux. Past experience told me this would be a rough ride, and it was. I used OpenSuSE 10.2, as it was the only major distribution to use the 2.6.18 kernel version so it would work with my new Intel 965 chipset computer.

Installation went relatively smoothly, but once it tried to boot, problems started occurring. I seem to have a history with Linux and its video system, and this was no exception. On first boot, the screen went black and I got the "Out of Range" message from my monitor. After a bit of digging and luck, the solution is:

• While the screen is black and the monitor is in 'out of range' mode, press Ctrl-Alt-F1 to get to the terminal.
• Press Ctrl-C to kill off the failed GUI loading attempt.
• Log in using the credentials you made during setup.
• Now, try using SaX2 to configure your monitor (this didn't work for me, but its worth a try). Do this by typing (pressing Enter after each line):
  • su
  • enter your root (admin) password
  • init 3
  • sax2
  • Fiddle with the settings here, in particularly the monitor settings. Test, save, exit.
    
  • reboot
    
• If that doesn't work, repeat the first 3 steps again to get to the terminal.
• Now instead of using SaX2, we'll try using the display configuration file used for the setup process. Type the following (pressing Enter after each line):
  • su
  • enter your root (admin) password
  • init 3
  • cd \etc\X11
  • mv xorg.conf xorg.conf.faulty
  • cp xorg.conf.install xorg.conf
  • reboot
• That should get you into the Linux GUI (which ever you chose).

Ok, now that I have the display running, I wanted to get xgl working (one of the funky special effects engines on Linux). I knew it wouldn't be easy - turns out the default drivers I had for my NVidia GeForce 7300 GT did not have the 3D acceleration stuff in them (they were the open-source nv ones, not the proprietary nvidia ones - geez, when will the linux people realise we don't give a damn about license conflicts as long as it works ). To install the 3D accelerated ones, follow the instructions at http://www.suse.de/%7Esndirsch/nvidia-installer-HOWTO.html#1. Note that from my experience, the pre-built packages downloadable and installed through YaST don't work, so I did it the 'long way', being the following (repeated and commented from the link above):

• In SuSE, opened the Control Center, scrolled to the bottom to YaST, then clicked on Software Management.
• Searched for the packages kernel-sources, make and gcc and ticked the box.
• Clicked accept.
• Waited for the download and installation...
• Closed the software management, YaST and Control Center windows.
• Opened up Firefox and navigated to the NVidia site to download the latest linux driver (http://www.nvidia.com/object/unix.html). Note that IA32 = x86 = your standard computer unless you're running a 64-bit one. Remember where you save the .run file - I saved it in /home/sam/Desktop.
• When the download's done, close Firefox. Press Ctrl-Alt-F1 to return to the terminal. Press Ctrl-C to kill off the GUI.
• Type the following to start the installation:
  • su
  • enter your root (admin) password
  • init 3
  • cd /home/sam/desktop (replace with the path where you saved the NVidia download)
  • dir
  • record the name of the file you downloaded, starting with NVIDIA-Linux.
  • sh (name of the downloaded file), e.g. sh NVIDIA-Linux-x86-1.0-9631-pkg1.run
  • Follow the instructions, and it should install successfully.
  • sax2 -m 0=nvidia
  • Check your display configuration again - note that it is now using the NVidia driver now, instead of the nv driver. Test, save, exit.
    • Tip: Press the Options button next to where it displays the name of your video card, and click the nologo option in the list - this gets rid of the huge and annoying NVidia logo every time it is used.
      
  • reboot

Ok, I thought I was done now, but I wasn't. I was stuck with a resolution of 800x600 on my Viewsonic VG500 LCD monitor. After fiddling around for a while, I discovered a neat troubleshooting trick for these problems (which I've had on almost every linux installation I've done).

• If you are currently in the GUI, press Ctrl-Alt-F1 to return to the terminal and then Ctrl-C to kill off the GUI.
• Now reload the GUI, however, this time, type startx -- -verbose 6 -logverbose 6
• Once in the GUI, open up the file browser and navigate to /var/log. Open up the Xorg.0.log file in your favourite text editor (if its vi, then by all means, damn weird linux nerds).
• Most of the stuff isn't that important - locate the section in the file that contains a heap of lines starting with (**) NVIDIA(0) - the asterisks could be two other characters to indicate the meaning of each line.
  
• Find something resembling this:
  
  (--) NVIDIA(0): --- EDID for @@@ (CRT-0) ---
  (--) NVIDIA(0): Maximum Image Size           : 65535mm x 65535mm
  (--) NVIDIA(0): Valid HSync Range            : 4294967 kHz - 0 kHz
  (--) NVIDIA(0): Valid VRefresh Range         : -1 Hz - 0 Hz
  (--) NVIDIA(0): EDID maximum pixel clock     : 0.0 MHz
  
  If your section (there's one for each monitor you have connected) is anything like mine, it means its old-ish and isn't quite as smart as Linux thinks. It is not providing Linux with the figures needed (called EDID data) for it to work out how to talk to your monitor.
• A bit further on after that you'll start seeing these lines repeat over and over again for the various display resolutions.
  
  (II) NVIDIA(0): --- Building ModePool for @@@ (CRT-0) ---
  (II) NVIDIA(0):   Validating Mode "1024x768":
  (II) NVIDIA(0):     1024 x 768 @ 60 Hz
  (II) NVIDIA(0):     Mode Source: X Configuration file ModeLine
  (II) NVIDIA(0):       Pixel Clock      : 65.00 MHz
  (II) NVIDIA(0):       HRes, HSyncStart : 1024, 1048
  (II) NVIDIA(0):       HSyncEnd, HTotal : 1184, 1344
  (II) NVIDIA(0):       VRes, VSyncStart :  768,  771
  (II) NVIDIA(0):       VSyncEnd, VTotal :  777,  806
  (II) NVIDIA(0):       H/V Polarity     : -/-
  (WW) NVIDIA(0):     Mode is rejected: HorizSync (48.4 kHz) out of range
  (WW) NVIDIA(0):     (4294967.000-0.000 kHz).
  
  The last two lines may be slightly different, but nonetheless, it will say 'Mode is rejected'. The reason it is rejected in my case is because the EDID data from the monitor doesn't exist but the computer hasn't worked that out and its using the dummy figures returned as valid values. Therefore all the modes get rejected because of silly conditions, like the one above, where it says the Horizontal Sync rate has to be below 0 kHz.
• As my monitor was not providing the data but the computer did not know this, follow the following steps to tell the computer.
• Open up the Terminal (in GNOME, go to Computer -> More Applications -> search for Terminal).
• Type in su, press Enter then enter your password followed by Enter again.
• Type in gedit /etc/X11/xorg.conf (replace gedit with your favourite text editor)
• Locate the line:
  
  Section "Device"
  
  And put in the following line before the following EndSection line,
  
  Option       "UseEDID" "FALSE"
  
• Save the file, close all other windows, press Ctrl-Alt-Backspace to restart the GUI, and with any luck, you should be able to use the resolution you selected in SaX2 earlier, and choose alternative resolutions using the Control Center.

If the above trick doesn't work for you, check out Xorg.0.log again and see what's going wrong - it is a very useful file for tracking down screen resolution problems. Note that if your EDID data is not empty, do not add in the above lines to xorg.conf - your monitor is providing data and this is a good thing.

For more information about what's actually going on behind the scenes in the section explored above in the log file, check out http://us.download.nvidia.com/XFree86/Linux-x86/1.0-9631/README/appendix-j.html. For various settings that could help fix your screen resolution problem, see http://us.download.nvidia.com/XFree86/Linux-x86/1.0-9631/README/appendix-d.html. For anything else related to the NVidia drivers, check out the other bits in the readme file at http://us.download.nvidia.com/XFree86/Linux-x86/1.0-9631/README/index.html.

Windows with all its issues doesn't come close to making a graphics card work this difficult. All you have to do is turn your computer on and the display works (with the inbuilt drivers or you can just download and install the newer drivers from NVidia)... Linux clearly still has a fair way to go before being a real challenger to Windows in terms of user-friendliness for those of us who aren't experts but enjoy experimenting and poking at it - I shouldn't have to touch config files, or the terminal at all.

P.S. For those who are wondering why I'm not using Ubuntu Edgy, its because its kernel version is old, and Fiesty isn't out intil April 2007. For the record, Ubuntu Dapper wasn't much better. It refused to boot until I manually installed the NVidia drivers using the terminal.

P.P.S. If your video card doesn't appear to be recognised in the Desktop Effects dialog in Control Center, and you're sure that 3D acceleration is enabled (check SaX2 - checkbox in near the bottom), you can enable it anyway by firstly holding SHIFT and clicking on the warning triangle and/or the word disabled against 3D acceleration until it changes to enabled. Then just click the Enable Desktop Effects button, and restart.

I've been fiddling with my computer recently, trying to get it to all work nicely after I inherited some new parts. I have a seperate partition on my hard drive for the purposes of testing Windows Vista RC1. After having to repair my XP installation due to a change in motherboard chipset and CPU (AMD to Intel), I realised I could not boot into Vista anymore - the Vista Windows Boot Manager no longer came up.

To fix this, I decided to try the Vista recovery process. It worked, restored the Windows Boot Manager, but I was no longer able to boot into Windows XP nor Windows Vista. Attempting to select the "Earlier version of Windows" gave me a message that essentially told me /ntldr was "missing or corrupt". Choosing the Windows Vista option gave me a BSOD, which was understandable given Vista didn't have the drivers setup for the new motherboard and CPU. So I had to reinstall Vista from scratch, as I could not find an option anywhere within the recovery environment to repair my installation (system and image restores were useless as I needed Windows to re-detect all my hardware again). Luckily it was just a testing OS and there was nothing important on it. It does make me wonder though, what should I have done? Maybe it was fixed before Vista RTM-ed.

Anyway, reinstalling Vista gave me access to Vista again, but still, I could not get into Windows XP; the same message appeared.

After some digging around the internet, I found the solution:

1. Boot up Windows Vista, and navigate to Start (the Vista orb)-> All Programs -> Accessories -> Command Prompt, then right-click on Command Prompt and select Run as Adminstrator. Either enter your password if needed or just press continue. The new window should be titled "Administrator: Command Prompt".
2. In the command prompt, type in bcdedit /enum active and press Enter.
3. Locate the sections titled "Windows Legacy OS Loader". Note the identifier (all of them if there is more than one matching section); my identifier was {ntldr}. If the problem is the same as mine, the device property should be currently set to unknown.
4. Use the command bcdedit /set (ID) device boot to change unknown to boot, that is, the partition that is active and is booting from. By default, Vista does not change which partition boots, so if your XP partition is set as active, it will add the new booting code to that. Therefore assuming your XP partition is set as the active (boot) partition, the above command will be correct. Apply the above command for each identifier, replacing (ID) with the identifier itself, including the curly braces, e.g. bcdedit /set {ntldr} device boot. Press Enter after each one - it should report that the operation was completed successfully.
5. Rerun the command in step 2 to check that the changes were applied correctly.
6. Test!

That should fix the problem. I assume that when the device was set to unknown, the boot manager didn't know what to do to boot that partition, hence the message.

Microsoft has put up a handy reference for bcdedit too: http://technet2.microsoft.com/WindowsVista/en/library/85cd5efe-c349-427c-b035-c2719d4af7781033.mspx?mfr=true

Note that for basic booting option changes, such as the default OS or timeout, you can run msconfig or the System Properties, Advanced tab.

If you discover that there is no "Earlier Version of Windows" option in your boot manager at all, try these instructions (also repeated on the above Microsoft site): http://www.nukeation.net/2006/07/05/5456+Update+2++XP+Goes+Missing+After+Installing+Vista.aspx

UPDATE (6/12/2007): Clarified some steps and fixed a few spelling errors.

Ever since Microsoft gave me one of their fingerprint readers, I have never used it for anything more than logging on and off on my computer (which is quite convenient), even though it can be used for authentication on websites, programs etc. Why? Because the software doesn't work with Firefox. Ok I thought, Microsoft is the big aggressive monster that consumes everything in its path, and Firefox is an obstacle. After all, I got it for free anyway - I'd be returning it if I had bought it though as Firefox is probably the most used app on my computer. Interesting, it didn't even work with MSN Messenger/Windows Live Messenger, which I thought was odd.

Just recently however, to coincide with the release of Windows Vista, Microsoft released version 2.0 of the fingerprinting software which included Vista support. They also made an effort to make sure Mozilla Firefox worked well with Windows Vista way before this, to the point where they invited the Firefox developers to their campus to work on issues.

Guess what - the fingerprinting software still doesn't work with Firefox 2.0. It works perfectly fine with IE7 (which is a great leap from IE6 - and by the way, please upgrade if you're using IE6, for one, my site will look much much better ) but again refuses to see the Firefox window as a login screen.

Where was the hardware division when the Firefox team visited? Or was it really all just a show...

What makes it worse is that apparently before Microsoft re-branded the fingerprint reader from DigitalPersona, the software actually worked with Firefox!

There is a workaround however - the FingerFox extension, which detects the login forms on websites and re-exposes them in a seperate window so the fingerprint reader's software can work with it. This is ridiculous however - great initiative and solution shown by the extension developer - but it's still clunky and nowhere near as smooth as the IE7 implementation. As the second most popular browser out there, its a shame that the hardware division didn't see it as imperative that their devices worked with it - maybe Microsoft hasn't changed that much after all...