The Táve Software team set out to make a full featured graphics
and photo editing application for BeOS in 1997.
More coming soon! Jamie is sending me some of our old screenshots
and documents; I'll be sure to add them to this page.
Lessons in Project Management
Sadly, Táve Software was disbanded in 2000, when Be Incorporated
unwisely changed its focus from BeOS, which was quickly gaining
ground at the time, having received placement in Best Buy and
with sales growing exponentially (I know; I rented a room from the
man responsible for the growing sales), to the unexciting and ultimately
doomed BeIA platform.
The most painful part of all of it, was that in late 1999 the application
finally started coming together. I was young and inexperienced when
we started the application; unable to properly lead such a large and
complex project. Several years into it things began to click. Having learned
the hard way just how dangerous "Wouldn't it be cool if..."
is to a project.
I also learned how important it is to plan, outline, and monitor progress.
I realized that "Good enough" doesn't mean "Perfect",
and that until you realize that, your project won't go anywhere.
If I knew then, what I know now; BeOS would exist, Táve would have
a handful of applications on a variety of platforms, and I'd be wealthy.
As you probably gathered, many of the features parallel those of Adobe Photoshop
and The Gimp.
Back when we were designing and developing Imagination, there were several features
in Imagination that were lacking in Photoshop. For instance, we had nestable layers
long before Photoshop... allowing you to create folders (and subfolders, which Photoshop still lacks) of layers.
We also had multiple layer types before Photoshop. Initially only Raster (paint) and
Vector (draw / text), though we had plans to include Video, 3D, and mixed-media layers.
That functionality, or ability to grow, was one of our key design points. Imagination
was just the first part of a larger puzzle. We planned to have a suite of multimedia
applications that could all work together; more so than Adobe applications can today.
For instance, we imagined you color correcting a movie while somebody else (potentially
in another part of the world) edits it, somebody else sweetens the audio, another person
adds subtitles, and a group of artists work on compositing. Another real world application would
be a teacher presenting a lesson while every student in the lab watches the same document...
seeing the teacher edit the document on their own machine.
Of course, that would be far down the road. You'd only now be seeing some of that.
You could see this though. For instance, you could have multiple windows open showing
the same document (similar to Photoshop's Document Views). However, unlike Photoshop,
as you drew, every single window would update live as if you were actually drawing in each
every window with a separate mouse. Everything was extremely responsive, as was customary
with BeOS; change a slider, and watch your document change live as you move it.
We had also enabled a tiling system, where the image was actually broken up into a number
of smaller images. This allowed you to open extremely large files, and have many different
layers in it, without having terrible RAM and swapping issues (which were common in
BeOS due to the lack of mmap and frequency of "double buffering"). Another example
of how we had to re-invent the wheel due to the young age of BeOS... you see, BeOS lacked things as simple
as Combo Boxes and Twisties to complex features like a mature development environment or mmap.
To my surprise, I later came to see a decent number of our "original" ideas
implemented in extremely expensive editing systems like Avid and Quantel
while attending Full Sail.
The core team, who stuck through with it from start to finish or contributed
a significant amount of work, are:
- Adrian Ziemkowski, Project Manager
- One of the founders of the project, I oversaw the
development, was the lead designer, and am responsible
for the failure of the project.
- Jamie Wilkinson, Coder Love God
- A founding engineer, Jamie was only a freshman in
high school when we began the project. While young, Jamie
was a very talented (having spent a summer at Johns Hopkins University no less),
vocal, and active member of the team. Jamie was responsible
for engineering and designing many of the custom classes which we
needed in order to develop the application, due to BeOS deficiencies.
Jamie's inovative "icon on folder" designs and icon builder, was
eventually adopted by Be Incorporated and is now a common occurence
in MacOS and Windows.
- Matt Lewinski, Color Systems & Algorithms
- One of the most productive and hard working members
of the team, Matt worked on many of the difficult core systems.
His color system work eventually became an important part of
the Document Kit I ended up spending most my time on near the end.
I think Matt, more than any other member, had the best understanding
of what needed to get done and how, in order for us to succeed.
Sadly I didn't realize that until too late.
- Zeid Derhally, Quality Assurance & Engineer
- Zeid, author of famed BeOS FTP application, NetPenguin,
worked vigorously on making Imagination a solid application which
could succeed in the commercial marketplace. Being an employee of
MetroWerks, meant he also managed our build environment and attempted
to bring a higher level of professionalism to the team
- Shafer Stockton, Systems Administrator
- A founding member, Shafer ran the systems we developed on.
He also provided feedback, testing support, and general assistance
during the process.
- John "Soco" Robinson, Core Engineer
- Soco was responsible for a good portion
of the design. He worked on many exciting facets of the application
and actively lead the way for Imagination.
- Eric Lanehart, Graphic Designer
- Eric developed the Táve logo, which has been
an important part of Táve's image ever since. He also
provided critical feedback of interfaces, icons, and user experience.
- John Huerta, Theorist
- John was another founding member. Although he didn't have
an opportunity to contribute as the application matured, he and Shafer
helped design the distributed and modular design of the original "Mozart" suite.