I discovered an error in the AndroidRobot 2.2 documentation last month,
and the prospect of correcting it prompted me to make some other changes
Last fall, while testing AndroidRobot 2.2 with an orthographic camera,
it became evident that Karl's model was noticeably different from the
images from The Google's Web pages. I toyed with the idea of remodeling
it, but tabled it in favor of other projects, such as a RoundEdge
update, two brand new modules to deal with spectral rendering, an
operating system upgrade, brushing up on my Spanish, (re)starting an
exercise program (again), and recovering from the worst toothache of my
life. But finding errors in documentation makes me crazy. (/Another/
documentation error explains the rapid update from 3.0 to 3.0a. This is
why professional writers have someone else proofread their work.)
I found an SVG version of the logo, which is claimed to have originated
from a Web page on the Android developer site, although in keeping with
The Google's typical practice, the page no longer exists. (I know
someone on these newsgroups works for The Google. Tell 'em to stop
that.) I ran the coordinates from the SVG source code through a
spreadsheet so that the new model would have the same size and position
as Karl's (burning bridges also makes me crazy), but with official
Speaking of not burning bridges, Karl's model may be summoned from 3.0a
by setting a version switch.
The branding page (linked from the AndroidRobot 3.0a User Manual) has an
Adobe Illustrator image, which may be the best source for the exact
proportions. It is loaded with neat stuff such as CMYK data for your
properly profiled color printer. Alas, I don't have access to Adobe
Illustrator, and the SVG image was good enough for my purpose.
In addition to the revised model, AndroidRobot 3.0a has new tools to
assist with creating user-defined eyes. The 3-D model's eyes are just
dark holes, and leave a bit of a spooky impression. Hopefully, the user
can mitigate this soullessness with these new tools.
Note: The standard Object Collection README file says that you may use
images that you generate using these files without restriction. As the
Android robot is trademarked, this is obviously not true for this
particular module. The User Manual has the proper licensing notices.
I am curious to know if anyone has used this module in association with
their app for Android.
Post a reply to this message
On 2015-04-27 3:22 PM (-4), Cousin Ricky wrote:
> I found an SVG version of the logo, which is claimed to have originated
> from a Web page on the Android developer site, although in keeping with
> The Google's typical practice, the page no longer exists. (I know
> someone on these newsgroups works for The Google. Tell 'em to stop
> that.) I ran the coordinates from the SVG source code through a
> spreadsheet so that the new model would have the same size and position
> as Karl's (burning bridges also makes me crazy), but with official
Just when you think you sufficiently documented a past project, or
thought you could get away with not declaring magic numbers...
The new head design issued last year has changed from the previous heads
to the extent that I need to go back to that spreadsheet to
re-proportion the entire robot. I've reacquainted myself with both the
SVG file and the spreadsheet, added comments <!-- --> to the SVG file,
and created identifiers in androidrobot.inc for most of the magic
numbers that correspond to the spreadsheet numbers. But I've run into a
strange snag: two of the numbers in the spreadsheet do not show up in
the relevant macro, and the two would-be corresponding numbers in the
macro do not show up in the spreadsheet!
I will have to figure out where my head was 5 years ago, and how those
two macro numbers relate to the spreadsheet.
Post a reply to this message