Posted July 25th, 2010 20:16 by Bob
Recent attempts to work smarter rather than harder have been an utter failure. As the business grows and we get busier and busier I’ve needed to look at our business practices and workflow. It’s so far not worked out so brilliantly.
I used to write a fair amount of our tools and all of the frameworks and code for our websites and applications. Originally this was because there was little option, there were no free or cheap versions of the tools we needed and the frameworks available were challenging at best. Of course this wasn’t all bad it gave me a greater understanding of the underlying technology and probably increased my coding skills far more than downloading plugins and other people’s code ever would. But, as many people pointed out it did make development a relatively slow and arduous process which in the new age of Agile Development would see me left behind.
It took a while for me to change my ways, I still have an OC like obsession with understanding exactly how everything works and will still rip a framework or plugin to pieces on occasion to make sure it does what it claims. I’m now a complete jQuery convert and massively happy with WordPress. I went halfway with PHP, I tried a bunch of frameworks but just couldn’t get on with any of them so wrote my own, it took about six months (mostly research) but works perfectly for us and allows us to put together full online applications in a matter of hours. It’s not ‘finished’ and never will be, I constantly add to it and improve it but it’s lightweight and flexible and does exactly what we need (and nothing more).
So with frameworks for php, javascript, actionscript and a shedload of standard setups for html and css I’ve embraced the agile, flexible world. Next step was to improve workflow and project management. This is where it’s all fallen down a bit. I really don’t want us to go back to the bad old ways of writing every tool ourselves, I need to sleep sometimes. I’ve spent a lot of hours over the last couple of months researching and testing project management and workflow tools and I have not found anything I would be happy to use. There’s a lot of dross out there combined with a lot of well written but feature-weak applications. The grind now is that I could have probably used the time better to have written something in-house. I still don’t want to but if I can’t find the right tools soon it may be the only option.
We now have two choices: 1) Use a whole bunch of unrelated tools in a haphazard manner, tying them together with apis where possible or 2) write it ourselves.
Working smarter and faster is fine if the quality tools are available but if you have standards you may be better off doing it the long way.
In case you may be able to suggest anything, our primary requirement right now is a decent project/task manager which supports:
That’s it. To be fair we did find one piece of software that covers it pretty well but it was way outside of our current budget.
Posted in Coding, Work | No Comments »
Posted May 22nd, 2010 19:36 by Bob
A bit more arsing about with HTML5. Purely decorative, completely non-interactive this time. An animated fountain based on one I wrote in AS3 which in turn was based closely on one from the absolutely brilliant Foundation ActionScript 3 Animation: Making Things Move! (seriously, if you need a good basis in programmed animation/interaction this book is a godsend, explained in simple terms and with a brilliant separation of concept from code so the information and tutorials can be easily applied to languages other than Actionscript 3).
Converting to Javascript was pretty straight forward, after all the syntax has the same origins. It’s slightly easier to achieve in Actionscript as Flash has built in the concepts of independent visual objects, wheras the canvas tag does not and so you need to roll your own, which for a simple example as this is not a problem, and I’m guessing there will be some nice JS frameworks around fairly soon which will make the job easier. Strongly tempted to convert a load of the other tutorials from Foundation AS3 to Javascript as well but not a great use of time and the author may (understandably) hunt me down and beat me to death.
The app is pretty simple, it generates a load of circles, each with it’s own direction, velocity, colour and size which move away from the center, when they are off the screen they are reset with new velocity, colour, size, etc and put back to the center again. That’s it. Here’s the code, it’s not very polished but it works…
Just the required bits…
That was easy. You just need a canvas tag. I’ve given it the original id of ‘MyCanvas’. Next up…
Once again, I’ve used jQuery to get things moving quickly, you don’t need to but it makes life easier.
First we set some global variables as follows:
Define a ‘Ball’ class. The xPos/yPos determine the ball’s current position in the canvas. vx and vy are the Ball’s x and y velocity. Wind and Gravity and horizontal and vertical force. The strangely named ‘sizeWobble’ is the amount the ball radius changes each frame. The reset function sets up the ball, most parameters are randomly generated, xPos and yPos are set to the center of the canvas.
The core of the script. Wrapped up in a nice jQuery document ready handler. First we set the globals for the canvas dimensions (as used by the reset method in the Ball class). Then we loop as many times as we want balls, create a new ball each loop instance, add it to the balls array and reset it.
Once our balls are created we use setInterval to call the ‘drawBalls’ function every 50 microseconds.
This is the function that’s called each ‘frame’ by the setInterval command. First we clear the canvas. Next we get a context for the canvas. Then we loop through the array of balls. For each one we:
That’s pretty much it, the remaining three functions are helpers. clearCanvas clears the canvas by setting it’s width (must find a more logical method of doing this). randomColour returns a random colour in the appropriate format and drawCircle draws a circle to the canvas…
Posted in Coding | No Comments »
Tags: actionscript, Animation, canvas, fountain, html5, javascript
Posted April 27th, 2010 23:17 by Bob
Been having a play with HTML5, very excited by the new semantics and especially keen on the new canvas tag. Aware that it’s not going to be supported everywhere in the immediate future but it’s a great step in the right direction. I don’t think it’s going to be the immediate flash killer some have made it out to be anytime soon (I love flash and AS and have no desire to see them go), but it can certainly work as a good alternative to flash for simple animation and interactivity. Once the decent JS frameworks add more functionality to support it it’s going to make life a lot more interesting.
I’ve put together a couple of experimental demos with HTML5 this week, this is the first. It’s a dynamic graph viewer. It can accept data via javascript and update the graphs on the fly. The data in question in the demo is supplied by the sliders but it can come from anywhere you like. I’ve kept the graphics basic for the sake of the demo.
I’ve not written any sort of tutorial/code article before, so bear with me, feel free to just download the source.
I’ve used jQuery and jQuery UI in the name of agile development. The jQuery ui for the sliders and a couple of jQuery statements to get measurements and deal with events, these could be easily swapped out for alternative frameworks or more verbose standard javascript.
View the source of the demo for the full HTML, I’ll just include the pertinent parts here:
This sets up two canvases, one for each graph, I’ve named them the imaginative ‘Graph1′ and ‘Graph2′. The controls div contains four divs to be converted to jQuery UI sliders (one for each point / bar on the graphs).
First off we set up a couple of global variables:
graphPadding is just a number to use for spacing items of the graphs out. I’ve just set a single number to use for all margins, spacing, etc to keep things simple. colours is an array of colour codes, one for each of the graph points.
Next we initialise with the nice jQuery document.ready function:
The first command sets up the jQuery UI sliders including giving them the appropriate event handler to redraw the graph whenever any of them are changed. The second command just sets each slider to a random value. Next we call drawGraphs to draw their initial state.
Now we have the main graph drawing function, this is set up to draw both graphs in one go.
First we get the canvas dimensions, these could be stored as constants/globals somewhere or hard coded but I like the flexibility of checking on the fly. Really we should check each one individually and split the graph drawing into two functions but it’s a demo and I don’t care. Then we clear each canvas to start drawing on it using a function which will be defined below.
Prepare to draw the first graph by getting a reference to the appropriate canvas then calling getContext on it. Then we set the stroke style to black.
Draw the axis, a simple case of drawing two straight lines. We use a combination of the canvas dimensions and the graphPadding global to work out the positions.
Work out how much space we actually have to draw the points in. The first point is two sets of graphPadding from the left and the last is two from the right hence the * 4. Same applies vertically.
The points will be equally spaced horizontally so we just need to get their vertical position. Loop through the sliders, get the value for each one (0 – 100), work out the percentage of the available vertical space and add it to the points array.
Draw the graph lines. We separate the line drawing and point drawing as when we draw a circle it calls closePath which knackers up our lines (there’s probably a better way). To draw the lines we loop through the points and draw a line between each one.
To draw the points we just loop through the points array again, drawing a circle at each one, again the horizontal position of each point is fixed.
I’ll cut out most of the code for Graph 2 as it’s identical. The part of interest is:
Here we work out the available space for the bars by subtracting the appropriate number of graphPaddings from the total available space then dividing the result by the number of bars. We work out the max height of each bar by deleting 4 * graphPadding from the total available height.
Then we loop through the sliders, get the percentage value from each one and set it’s bar to the given percentage of the total height.
That’s pretty much it. I’ve used two additional helper functions which are based on those found at the Mozilla Canvas tutorial, they shouldn’t need any explanation:
So, there you go. Dynamic graphs without flash or server side graphics libraries. These could of course be achieved pretty much entirely with CSS and JS but it’s been a nice intro to HTML5 and the canvas tag.
I’ve not given this a huge amount of testing (or thought) as really for my own entertainment but hope it’s of interest.
Posted in Coding | 5 Comments »
Tags: Coding, css3, graphs, html5, javascript, jquery ui
You are currently browsing the archives for the Coding category.
advert Anxious Silence Art book Books Books of Joy Branding Business Cards Church day trip Design Everything Starts freelance Help History Of Guns html5 Illustration InkyStuff javascript Logos london mac Maidenhead Marquee Moonshine PR Music Photography quoting rant reincarnationfish rough school self-initiated Self Employment Self Invoved Bullshit Shinytastic Stationary Template This Is Radio Silence Time Lapse Trees Video Web Design Wordpress Work
