| |
|
Puddlejumper
An elecroluminescent raincoat that showcases hand-silkscreened luminescent
lamps and embroidered rain sensors. The goal of this project is to frame
technology as a medium for fashion that can be softly toned, playful and
exuberant rather than cyber-styled gadgetry.
7/28/01 - See Puddlejumper in progress!
Soft Hard Ware
Based on my own experience
designing and building the projects for my master's thesis from scratch,
I'm now working on ways to erase the separation between software and hardware.
We need architectures for programming physical devices in tandem with
screen graphics, with seamless interaction and communication between both
aspects.
6/29/01 - PCB design of the final boards has commenced. The boards will
be optimized for small size, easy attachment/incorporation, modularity
and flexibility.
5/7/01 - Efforts are being temporarily redirected to experiments in electroluminescent
silkscreening. The C compiler for the Tiny boards is still being planned.
Some Tiny-board features are being rethought and designed, for instance
the analog IDs; hopefully a screen-based simulation of how these IDs might
work as part of a general communication/broadcasting scheme will appear
soon.
4/17/01 - Programming directly in JetSet instructions is enabled via a
pc-based program that includes cues for sending code to the Tiny board.
Although right now code must be entered with Jetset mnemonics that are
packed into the jetset 2-byte instructions, programming capability in
C is on its way. Also, when code tagged as a 'program' (rather than single-exectuion
'instruction') is transferred to the Tiny board, it will automatically
be written the PICF876 built-in EEPROM for persistence through power-down.
4/1/01 - Jetset, our instruction set for the new Tiny boards, has been
implemented on the PIC16F876. All the core stack, heap, logic, math, register
and branching functionality is up and running. PIC-specific special calls
for setting I/O pins high and low have also been added, as well as printing
out to serial port for debugging. The next step, currently underway, is
building the link from the PC to the Tiny board; right now i am working
on an interface for coding directly in Jetset mnemonics, which are converted
to 16-bit instructions and sent to the Tiny board via serial line. The
Tiny needs to know how to write to the internal PIC eeprom for program
persistence. Eventually coding will happen in C, and get compiled to Jetset
code via Jared Schiffman's PERT system.
3/4/01 - The lastest major development of the soft hard ware project,
which I'm calling Tiny for now, is the specification of an instruction
set that will be used by myself, Jared
Schiffman, and Tom White.
I've moved from Microchip's 16F84 to the bigger, brawnier 16F873. I'll
be writing a VM that will run on the PIC and receive code via a serial
line from PC. Jared's C interpreter, Pert, will be expanded to compile
into instructions that will run on the VM. I've also started thinking
about the form factor and flexibility of the end-result boards, i.e. how
to allow for expansion modules, communication, easy hookup and connection,
etc.
2/15/01 - The past few months have been completely absorbed by Atmosphere
(see above), a humongous
endeavor by scores of people. As part of this, I did some WinCE programming
for the Compaq IPAQ3600, a palm-sized PDA with a bright 16-bit color,
320x240 display. I'm hoping to do more stuff with the IPAQ soon, perhaps
involving a wireless network connection. Maybe a hook-in to the soft hard
ware system I've resumed building?
10/18/00 - My current
project, Lint, is
mostly finished technically. On
the soft hard ware front, Lint utilizes the Cricket bus protocol to control
tri-color LEDs. The next step, now that hardware/hardware communication
is established, is to start working on the software/hardware link. This
is the meat of the problem. I'll be starting to design an interpreter
specification for the PIC chips, while coding serial communication from
the PC to the PIC.
10/7/00 - My Edgewise00 presentation went pretty well. I brought an in-progress
Lint to show, as well as still images which can be viewed online here.
9/18/00 - The current model is
one of interpreter-based firmware (Microchip PIC) as the core of hardware
program modules. PC-based software can send instructions/programs to the
program module serially. A standardized set of i/o sensors and actuators,
with normalized values, can be driven from the PC or the program module.
As references, I am looking at the Epistomology & Learning Group's
Crickets project, which is also interpreter-based and modularized.
|