I’ve also been discovering the nuances of the different web-browsers. Chrome and other webkit based browsers are generally my browser of choice. Mainly due to their rapid adoption of rapidly developing standards and also the user interface design which is minimalistic and easy to use.
Although I haven’t even taken the 70-480 exam I already have my eyes set on the next goal of taking the next two exams which would get me the ‘Microsoft Certified Solutions Developer (MCSD): Web Applications’ qualification.
Study continues …
Some of the resources that I am using to help with my studies are:
- http://www.microsoftvirtualacademy.com/tracks/developing-html5-apps-jump-start – This is the recommended Jump Start course for the 40-780 exam. It is very heavily Windows 8 centric although it’s a very good base of information. I would be careful here though as some things which they discuss are not yet avaliable in most browsers or at least not how Microsoft implements them. An example of this is the flex-box and the grid box. At time of writing the flex-box does have support in other browsers but behave very different to the Microsoft implementation. The grid box is only implemented in IE10. Also the Error object is implemented very differently in IE10 than other browsers (I checked Webkit and Firefox).
To give myself interactivity and to let me rapidly prototype things I am using the open source editor Brackets.io which is really starting to come along, easy to use, configurable and intuitive. It also works exceptionally well with Google Chrome so I can very easily change HTML & CSS files and see the results in real time.
Some things which also dawned on me was with getElementById and querySelector calls. So to get the element with id=”myElement” would be getElementById(“myElement”) or querySelector(“#myElement”). The querySelector looks great and is used by quite a few people, but I was wondering about performance. Taking a look at a few projects on JSPerf demonstrates that in many cases querySelector is 58% slower than getElementById which is staggering.
Also in the Microsoft Jump Start they talk about the NodeList returned by querySelectorAll is live, so if you add a new element to the page then the NodeList should automatically update with the new element. I tested this on Chrome and found that it doesn’t, and that I needed to call querySelectorAll again. I would therefore be very wary about some of the content of the Microsoft Jump Start as not everything they spoke about will work on another browsers.
The studies continue.
In the name of creativity and plan curiosity I’ve been looking at lots of different things recently. Quite simply anything which caught my eye or seemed interesting.
WebGL – This has been around for a couple of years now and is simply 3D graphics for the web. Up until recently the only real way to get 3D graphics in browser has been to install a third party plug-in or bloatware software which in most cases doesn’t work and if it does will only work for a very small fraction of devices. WebGL is an open standard being developed by the Khronos Group and supported by Google (Chrome), Opera, Apple (Safari), Firefox a few others. With no great surprise Microsoft Internet Explorer doesn’t support it (though what does it support really?).
WebGL is something to get excited about, and their are lots of people currently developing for it and creating technical demonstrations (see Chrome Experiments and Three.js). I’ve recently started learning how to use WebGL from the book ‘WebGL: Up and Running’ which does a little bit on WebGL and quite a lot on the WebGL wrapper Three.js library. WebGL is development is certainly not an easy one to learn (much like any 3D or OpenGL programming) but there are fantastic libraries such as Three.js which takes quite a bit of the heavy lifting and help to explain some of the more difficult concepts.
Windows 8 Modern UI – I decided to take a look at developing for the new Windows 8 UI, mainly to find out what was involved, how easy / hard it is and to create a comparison between it and Android programming. As this is quite a large change to traditional Windows programming there is a reason amount to take in. Though it works heavily with existing Microsoft programming concepts such as XAML and objects and is generally fairly easy to navigate. I’m still not sure about the new Windows 8 probably as it’s such as major change to a operating system which hasn’t really seem such as dramatic change since the introduction of Windows 95.
Personally I’m sticking with Windows 7 as my main dev environment, why? Well my general philosophy is don’t hide anything from the user, but to show things when they need to be shown. Windows 8 hides a lot, with hidden menus around the screen I can see how it works for tables but it doesn’t work for desktop. I’m one who works in Virtual Machines quite a lot so you have to be quite precise if Windows 8 is on my right screen, the host Windows 7 on my left and I want to click on that precise point in the bottom left to bring up the start menu or top right to bring up this menu, or top left to bring up that menu …. Yes I could just use the Windows key but that not the point. Simply it’s bad design.
Hack Design – I may be a software engineer but I love design, even more I love good design. I also like to play with pens, pencils, paper, scissors, post it notes, glue and all those things you used in primary school. I also love to draw as a hobby at home, I may not be up to some peoples standards but I love it. Hack Design is a web site devoted to engineers, programmers, hackers or anyone who wants to move towards design or at least what design is, what it involves and how to start doing it. Hack Design is a weekly digest which uses text, pictures and videos to talk about and explain design, and design ideas.
These are just a selection of some of the top ones I’ve been looking at. In addition to these I’ve also been looking more into Node.js, Underscore.js, Backbone.js, Phantom.js and many others.
Today I heard about Bootstrap from Twitter and decided to take a look. Bootstrap is a collection of HTML & CSS conventions and was developed by Twitter as a single conventions library. Bootstrap is pure CSS and HTML which has been tested and works on all modern web-browsers. It is very well documented with some great examples. I’m amazed at how little code can be written which achieves very powerful results. The example\hero.html page alone gives a good basic example of using the library to rapidly create well designed, well structured feature rich designs. The docs\index.html page though is amazing and really demonstrates the simplicity and power of the Bootstrap library.
Another tool is Less also from Twitter which is a powerful style sheet language which extends CSS. Less is basically CSS with the addition of global variables and nested styles, two things which are most definitely missing from CSS in my mind. You can then run the .less file through the Less tool and it will generate the CSS for you (my preferred option), or include the .less file and it’ll generate the CSS from the on the fly. Less is certainly something to be looked at by anyone doing web development as it massively simplifies the task of editing and getting right your CSS files.
Using Bootstrap for only a couple of minutes and taking a look though the code has already taught me a lot. I’m certainly going to go back over my implementation and reimplement several things. Less is a pretty great tool too.
Another great advantage to webapps is that it’s easy to “try before you buy”, you can login and try the demo or a low function version of the software, if you want more functionality you can subscribe to the the service, if you don’t want it any more then just stop the subscription.
This does have a few draw backs, traditionally when you buy, download and install you have a copy of the software, forever. So it the company goes bust, the licensing model changes or support is stopped you still have the software on your computer and you are free to use and reuse it. Webapps though don’t have this luxury, if the company goes bust, the licensing model prices you out of the market, changes to Terms and Conditions or the server goes down then you simply don’t have access any more.
Webapps also have other trade-offs. The biggest problem for webapps is speed. Web-applications are written using interpreted code, this means that the web-browser reads the code using a language close to what a human can understand. Typically desktop PC software is written in a programming language and compiled to something understandable by the computer, in the computers native language (hence native code). For interpreted code the web-browser has to read the code, work out what each instruction means and convert it to a form that it understands each time the software is run. This does provide a bit of a performance and memory cost. Many modern browsers are though very good at doing this task, desktop computers are becoming increasingly powerful, so most users day to day don’t notice the performance bottlenecks. It’s mainly when you have complex data-processing tasks, or where you have embedded devices which don’t have the performance processing capabilities that these problems are observed.
Webapps though can be stored and used on devices. For example HTML 5 now introduces offline mode so if you loose internet connection the webapp is cached in your browser, and your files and settings re-synchronised with the web when you reconnect.
The future is certainly bright for the web and all those who can see opportunities to harness its advantages and capabilities.