|Hooray for funny web 2.0 names (See VYEW…a project I worked on). Gliffy is a visio knockoff written in flash using OpenLaszlo. It has a feature where you can publish a JPG of your drawing, but everytime you save your drawing, the JPG will be automatically Mobdro APK updated! So I’m testing that out here:
[EDIT]: Ok…it seems to work. I just change the drawing over there, hit save, and when I reload this page, the drawing is updated.
Post a comment w/ a valid email address mobdro if you want to get into the gliffy beta.
Click for larger
March 31, 2006
March 29, 2006
This is a procedural landscape I generated using Perlin Noise, and the beauty is that it’s all done in Flash. I’m writing this post from within flickr, just to see how it’s done.
You can see the actual swf at my flash page, but beware, most of the stuff there is beta, and I don’t have any fancy flash embedding set up, so it may not even work for you.
Here I am posting the schematic of the current design, and my proposed schematic for design revision 2. The main changes are:
- Directly drive the ROW inputs with transistors, instead of using shift register. I have enough I/O pins that there is no need for a shift register here, and the transistors will allow a decent amount of current through.
- Limit the current coming out of the COLUMNS with resistors. This is needed now that there will be more current flowing through. Before I got away without them, probably because each row was being driven by a single output pin of the shift registers
- Use shift registers on columns as before, but this time with a latch, so you wont see the data being shifted in. Thus the ‘off’ LEDs will really be off, instead of the faint flicker/glow
Schematic Rev. 1
(What I currently have built)
Schematic Rev. 2
(What I am planning to change next time)
March 19, 2006
I ported the code to C. Now it is easier for me to work with.
This makes it easy to scale things, so now I have 10 boucing lights instead of one:
bounce3.mpg – 3MB
I’m no ASM master, so I could probably be doing it the hard way, but this code in ASM:
March 15, 2006
I finally got the code worked out for bouncing a little ball around. Here’s a short movie:
bounce.avi – 200KB
MPG1 format – 1.1MB
I think next, I am going to rewrite all this code in C. The ASM is getting too complex for me to manage efficiently, I spend too much time tracking down ASM bugs versus actually writing code.
March 14, 2006
Ok, now I’ve written the code to display patterns from RAM, but I haven’t written the code to actually update those patterns. So I’m not sure exactly where this pattern is coming from. But it’s slightly more interesting looking now:
Still the same problem with the ‘off’ leds lighting up slightly, but I think it’s not too bad for now. Next up, I think I’ll animate a simple 1-pixel ‘ball’ bouncing around or something.
March 13, 2006
Here it is you can see the different colors possible:
One problem, is I am using shift registers without a latch. So as the data is shifted into place, it appears on the incorrect LEDs for a very short amount of time. Which causes them to light up a short amount of time. I’m not sure there is anything I can do about this, besides shifting the data into place faster: currently my AVR is clocked at 8mhz, but it supports up to 16mhz (I just need to put in a different crystal). I think best solution is to use a latch though.
Finally finished wiring it! (Man that was a lot of work):
Unfortunately, my programming cable is about 3 inches long (parallel port). I have an extra 6′ parallel cable, but it’s the wrong gender. So currently, I have to crawl behind my PC everytime I want to reprogram the AVR 🙁
I’m building a correct gender parallel cable now, so I don’t have to keep doing that. (Yeah…I know you can get them for $4…but I’m cheap)
Everyone keeps telling me I need to put resistors in series with the LEDs to limit current, but it seems to work ok without them, and thats 16*2 fewer solder joints I need to make. We’ll see if anything burns out hah.