Silverlight 4 PivotViewer…I like it…A lot…

There aren’t many controls that I see that make me sit up and think “wow”, however this really is one of those. For a while now I have been coming up with new ways in which to allow users of our workFile ECM product to access massive amounts of data visually, allowing the user to search through them in a graphical way that is both intuitive, but more importantly in a fashion that is usable rather than just looks pretty. We have had quite a lot of success with this, however in certain areas our solution just isn’t perfect, and for these areas we revert to more traditional methods of understanding / navigating data. The main issue is usually the potential amount of data the user may have to look through…

Let’s have an example. A while back I was asked to consult for a company that were looking at a solution that needed to look through vast amounts of pictures and graphics. The vendor had a very visual solution; it looked great and performed ok, for small sets of data. It involved using nicely animated thumbnails of data, which the user could then navigate, very similar to the standard tree view with details pane. However, the problem arises when your node could contain hundreds of thousands of images (which the customer did have). While the demo looked great and all the representatives from the client thought it would do the job perfectly I had to speak up and ask just how it would cope with say 500,000 images in a single node….. The vendor looked a little shocked and said “you would navigate in the same way or break the node down into several other nodes”…The client looked fine, however again I had to point out that scrolling up and down searching for a single image in 500,000 or popping into every sub node to find my image would take a hell of a long time. It was at this point that the vendor and the client both realised that this solution wasn’t going to work for them….Now this may seem obvious, but the issue with cracking graphical interfaces in this fashion is that, while they work great on smallish subsets of data, they really do fall down to massive vast amounts of information (you can’t beat a good text based search can you…..or can you…)

This morning I decided to finally look at the PivotViewer with Silverlight 4, and, I was very impressed. I have to say I like it a lot…All of a sudden I have a tool that can give me just what I have been looking for (without anyone here having to spend a long time coding).

What is PivotViewer

Basically it is a Silverlight 4 based control that allows you to graphically represent collections of data. You can filter and sort this data quickly and move between different views of your collection allowing users to quickly identify just what they are looking for, or perhaps trends in data for example…

What could we use it for?

Well there are many things it could be used for. A couple quickly spring to mind. Imagine a way of quickly and accurately reviewing your quarterly sales in a graphical format. Then imagine being able to break that data down (using the same graphical format) into each month, then week, then day. Or perhaps you want to compare cars based on fuel economy, price, and number of seats. As a user, you can filter and re-filter all this information graphically so you can quickly and intuitively find what you are looking for…

There are many uses for this particular tool, and it could indeed lead to RIAs that deliver a completely new way of providing us with data and even navigation…

Interested?

Well if you are interested, have a look at http://www.silverlight.net/learn/pivotviewer/

HTML 5 – It’s not the end of internet plug-ins

I have posted a number of times now about HTML5 and my concerns that people see it as a complete replacement for internet plug-in such as Silverlight and Flash, allowing RIAs to be delivered in pure HTML 5. One of the main people who keep going on about HTML 5 is Steve Jobs (though I think a lot of this is trying to convince the users of iPhones and iPads that Flash has a short life ahead). However, it seems that more and more people are sharing my opinion that HTML 5 will not kill of Flash and Silverlight, and that its adoption is a hell of a long way off in general…A recent report and article from Forrester illustrates this…

HTML 5 traction and buzz….

There is for sure a lot of buzz around HTML 5 in the past couple of months, least not because of Jobs, but also because Google has recently open-sourced its VP8 video codex. To date, abilities and licensing issues surrounding such video converters have been one of the sticking points for beta HTML 5, however this is not the only issue. Though there is a lot of internet buzz, it seems that adoption of HTML 5 is a long long way off, with browsers only supporting small fragments of HTML 5 currently. It seems that for wide spread adoption, as users we will be waiting until 2020 or sometime around then…That’s not exactly close is it…Its again another reason why I am not at all “hyped up” about HTML 5, it’s just so far off….

So while HTML 5 is a long way off, just think how much traction Flash and Silverlight will gain in this period. Silverlight is the new boy on the block, but has already around 60% adoption across all machines. That’s rather impressive, all this while HTML 5 is in beta releases and going through a lot of, development pains and issues shall we say…

There is also the issue of cross browser issues. Just like HTML 4, HTML 5 will suffer at the hands of different browsers. The author of the Forest report (Hammond) stated “Until you get consistent behaviour the question will be why you would use HTML 5 when it actually creates more challenges than it solves from a testing and deployment perspective.” I have to say, this has always been an issue with HTML in general, especially when delivering internet applications, and it is one that won’t go away for HTML 5…Though HTML 5 is supposed to be intended as an enterprise-class product, the reality is that the architecture of HTML 5 with the browser has a number of issues and draw backs, even when talking to web services. Though the aim is for HTML 5 to allow easier building of “applications” the fact is that HTML and that side of the web architecture was never designed with this in mind….

Test once…You are all done…

Ahh, well this is not the case is it with HTML. Unfortunately you will need to perform tests on all the browsers out there, and no doubt, place “HTML fixes, CSS fixes and JavaScript fixes” into your application depending on what browser is running it. This does make life a lot harder for testing and development, oh, and of course ongoing support. However, this problem is just not there for Flash and Silverlight, because their architecture is completely different and in many ways separate from the browser and the web in general, indeed you can run Silverlight out of the browser fine…

Hammonds recent report – titled “Does HTML 5 Herald the end of RIA Plug-Ins? Not Really” – found that application delivery through RIA amongst businesses rose to 34% in 2009, up from 26% in 2008. This illustrates the increase use of RIAs amongst businesses, especially with technologies such as Silverlight develop further, all this while HTML 5 is still in its draft phase…

For traditional website material, you could still use HTML and HTML 5 if you wish, however for complex functions and applications, I would always recommend the use of Silverlight, there are just so many hurdles you negate while being able to use a technology that isn’t restricted by the browser web architecture.

Open aspect of HTML 5

So many people claim they love the idea that HTML 5 is “open”. And there are some good arguments made for this, however I have yet to see one example where these arguments are valid. Especially arguments that users may have to pay for Flash or Silverlight use, that you can become restricted to what browsers you can use, or that you are dependent on them for your support…I don’t see an issue or potential issue with any of these arguments, they are just created so people try to feel more safe with an “open” technology controlled by many rather than a single company…

However, this open aspect of HTML 5 may also work against its progress and adoption, especially as open standards are very slow to develop. HTML 5 has been in development for a decade now, and though early candidate releases are recommended for 2012/13, it is a while yet before we see HTML 5 as the standard version of HTML being used. On top of that, cross browser issues and W3C adoption is even further off…

Architecture…

RIAs require processing on the client, or at least they should do. Users expect “thick client” performance and usability in an RIA and on top of that, access to hardware components, such as storage, web cams, other applications running on the client etc etc. The architecture behind the web and HTML jsut doesn’t allow this. Though HTML 5 will bring us a richer web, with easier video playback, website animations and improved usability (a little like Ajax has done), it will always be behind Silverlight for example, that can take advantage of hardware on the client, keyboard interactivity, integration with other applications and the ability to work in a “disconnected state” from the internet….

My own view on the use of HTML 5 in the future…

It simple, for typical web content, HTML 5 will provide a greater level of interactivity, animation and improved user experience. It will no doubt be used for “simpler” RIAs, however its adoption as a serious RIA for businesses is plain fantasy. RIAs need to deliver more, and therefore organisations will continue to look to plug-ins, especially Silverlight more and more. I also believe that websites available to the general public will also have more aspects delivered in Silverlight, even once HTML 5 has gained traction, simply because Silverlight can deliver a better end user experience without many of the hassles associated with web development and HTML, CSS and JavaScript across multiple browsers…..

Do we need a web browser?

There have been a lot of discussions I have seen floating around on Twitter etc with regards to HTML 5, and will it kill Flash and Silverlight. To be honest, there is no way this can happen, simply because both Flash and Silverlight do not rely on a third party to make them work. In addition neither has to conform to a generic standard which can hinder their functionality. Both have product roadmaps and both move forward at a rate that such a generic implementation could never hope to achieve. This means, the user experience will always be (potentially) better, and that’s the main aim.

However, both Flash and Silverlight based web experiences do rely on a browser. A browser has to be used by the end user to locate the web site, and then for the Silverlight / Flash plug-in to be executed. After that, the browser is pretty much redundant…

In the beginning

In the beginning of the Internet, a browser was simply used to locate, access and display basic documents, that were formatted in a particular way in which the browser would understand. (I know, I am making this very simple, but I want everyone to see where I am going with today’s post). This allowed people to access these documents that were stored somewhere and read them. If you think of a browser as Microsoft Word for example, and the HTML as the actual document, you start to see where I am coming from…

Browser wars…

Jumping forward, and into the web as it was a few years ago (before social media, videos etc),  the browser started to become an integral way of accessing content on the internet. Using HTML format for the documents, the browser allowed users to use an address to find that content, then interact with it (move around the website etc). Now this is all fine, if you have one browser, or a set of hard and fast rule of standards that everyone conforms too. But we don’t, in practice that is…

There are many browsers out there, which essentially have the primary of displaying HTML content to you, the user. However, as users we want more. We want to have options to store favourites, access feeds, personalise my browser etc etc. We also want websites to do “things”. We don’t want to just read content. So what we end up with is companies fighting for us to use their browser, which in turn turns into a bit of a nightmare for web developers as their supposed standardised HTML gets displayed differently in different browsers. Worse than this, some functions just simply don’t work in some browsers…

Does browser wars actually help end users?

Old way of thinking…

For me the web has moved on. We are already saying goodbye to web 2.0, and some smart person will term web 3.0 before long (which will actually mean nothing different to web 2.0 or even web 1.0…) my point is, the web hasn’t changed its implementation, only we as users have changed the way we use the web and what we expect from the web.

The concept of using a third party application to access content on the web is old. I don’t like it at all. I also think that using HTML or any standardised format to deliver applications is plainly wrong. As a developer you are always being “shoe horned” into a way of thinking and working which hinders the application look, feel, interaction, and therefore detracts from your users experience.

Internet websites are no longer formatted pages of information; many now act as applications and with Flash and Silverlight, deliver highly rich, interactive user experiences. With such websites, the browser is simply used to find the RIA (rich internet application) and start it. The application isn’t run by the browser at all. So do we need a browser for this?

HTML 5 is supposed to deliver the ability to show video for example. However, the same issues will still apply between browsers and websites; they will just now be even more complicated.

A new way of using the web

In my own mind, HTML should remain as it is today, however, with standards (especially regarding CSS) tightened. HTML is fine at delivering content, that’s after all what it was designed for. However, delivering complete websites, rich user experiences should be left to bespoke software, such as Flash and Silverlight. This form of distributed computing power helps the end user, and enriches their experience. I see no place for a browser on my machine, and would rather see the ability to browse the web as part of the underlying operating system.

Websites can then be developed in whatever technology they require, such as Silverlight or Flash. These technologies then display the website / application as they should. The web is used to provide access and download the application / content, no need for a browser…

I hear some of you crying at this point “how will a search engine pick up the content”, which is a good point. However, search engines must adapt. Why can they not interact with Flash and Silverlight? With the latter, the content essentially is stored as xml, so it’s not a massive leap. Also, what’s stopping search engines from picking up on tags that describe the content fully, still within the hosting HTML?

HTML shouldn’t be seen as just something a browser understands, rather a format the operating system itself understands. Once this happens, and we use the web to distribute applications and information in this fashion, many of the headaches of the web will be removed, and we can truly open up the potential of distributed and mobile applications / rich experiences…Silverlight 3.0 already delivers an out of browser experience, so are we far away from this ideal?