Well, all my original work on the Surface got erased from existence by a hard-drive failure coupled with the Uni’s Surface being replaced. Unfortunatly, the result was supposed to be shown to the BBC this Wednesday, so tonight I’m working on piecing together a rough copy of what worked so that a demonstration can be filmed and shown.
Let’s hope this goes well…
I’m probably jabbering nonsense, though if you fancy being more specific I’ll be glad to try and be less vague.
My rough little screencast. Any feedback would be appreciated.
I guess I can use this as an all purpose devblog, right? I mean it’s not like anyone’s reading (and if they are, it’s not like they care)
So some more work came through from the argumentation department. I’m making a screencast to demonstrate the basic features of Araucaria; a lovely little java application (platform independence FTW) that allows you to analyse arguments, import/export images and AML (Argumentation Markup Language, based on XML) and generally pull apart arguments like a skilled surgeon (or in my case, a small boy armed with a magnifying glass, a pair of tweezers and a sadistic streak.)
I’ve got Camista on evaluation and I don’t quite know what to say. I mean, on one hand it’s easy to use and makes great looking screencasts with a bunch of pretty sweet features. On the other, it keeps crashing and has dumped at least 2 hours total of work. It’s like it’s actively trying to test my understanding of the fact that you can’t smash virtual things.
Anyways, assuming I ever get to the first draft, it’ll go up here. Wish me luck.
Been a few days, eh?
Well, basically I was completely baffled by a bug that was introduced with the relative filepath update due to the fact it only appeared on the surface. Lo and behold (after tearing out some of my Samsonesque locks) it was a silly error to do with assembly paths and substrings that produced some pretty baffling side effects. Example: I loaded three previous versions of the project onto the surface at once and suddenly the error was gone. Oh well, what can you do.
Since then, I’ve changed the arguments from being hard-coded (because that’s useful to virtually no-one) to being read from an XML file in the base of the image folder. Next step is to make a dodgy wee program that updates the file and puts images in the right place (useful to computer-literate argumentation folk)
I dare you to say that aloud three times fast.
Just a quick note at today’s/yesterday’s work. Just a bit of tidying up, really. The file path is now relative (meaning I can shove it on the Surface really quickly in debug mode should I need to).
Being as I need the “Path” (System.IO) class to do this, and later need the “Path” (System.Windows.Shapes) class to connect stuff up, that’s going to lead to quite the clash later on. Luckily for me, I’m only really dealing with files in the one place, and even then only when demoing; so I just used a throw-away reference (i.e [when calling] System.IO.Path.Combine(assembleyRefrence, filepath))
Anyways, Today is working out the quirks of multiple images, then figuring out whether I need to connect them up or just start working on DOT processing.
I spent yesterday deliberating what it is, precisely, that a Button (Surface or otherwise) provides and narrowed it down to two things: Context (i.e what the button does) and an Event. The trouble with buttons in this context is that it hijacks the feeling of control that is provided by the ScatterViewItem, makes SVI contact detection difficult and takes a lot to make look nice.
It’s with this in mind that I’ve decided to replace the button entirely for some stylised text (Context). When the SVI is expanded (read: When you expand on a point, geddit? :D) it fires an event almost identical to the one fired by the button.
I hope it’s obvious what’s going on in the picture: A tag is placed and one of it’s arguments expanded. The image for the argument is shown on the right as before.
In other news, a grey background was annoying me. Here’s to royal blue (for now!)
I’ve not done anything since the last update, so here’s a quick look at the end result of the exploding animation. Note the wee boxes around the (soon to be replaced) buttons. I’ve decided to make a gesture for the ScatterViewItem (a.k.a the wee box) that will expand the box, loading the picture. Later on, this can be expanded so that it just becomes quite large (a sort of title, if you will) and all of it’s argument graph can exploded from it onto the surface.
Also introducing tag #03: A taxi driver. I needed a lot of arguments to test the explosion, and who’s got more opinions than a taxi driver?
Animations are the work of the devil and are to be treated as such.
Upon reading a tag, the code now plays a short animation of the topic areas “exploding” (if you can call it that with three objects :P) from the centre of the tag.
THAT’S RIGHT ANIMATIONS. I’VE BEATEN YOU. I WIN.
Tomorrow holds more fun with Animations (that is, animations coming from the topic area selected) and some thinking on what to replace my SurfaceButtons with.
Err, time for some well needed sleep I think :D
In preparation for my meeting with Chris Reed and Simon Wells, I spent the afternoon crushing the few obvious bugs I was on about earlier. Now:
Satisfying! Next step is to “make it more flashy. A waterfall or something”. You have to love briefs like that! So I guess tonight is looking at animations, library stacks and the like and seeing if I can make something flashy looking burst out.
To test it, first I’m going to use a cut up version of Georgie’s argument, but connecting the pieces might be an issue. Maybe I’d be best to just start on DOT processing first. I’m sure It’ll be detailed in the next post.
Oh, and my visage is the face of the NCR competition for the QMB advertising video by sheer accident. I, uh, can only hope for a good edit :P
