From: Gary Czvitkovicz <gac_at_usa.com>
Date: Wed, 27 Mar 2002 10:51:00 -0600

>Right, we can now see the need for creating and keeping a proper source
>repository. It seems stupid that the "real" lcd code we have
>now is in the ui
>simulator, in Alan's code at home and in Gary's code at home.
>And they are
>all mailed around in various archives...

Thanks for setting up the CVS project. I promise to start using it :-)

>Have you kept to your initial buffer-approach that draws
>everything in the
>buffer and then writes the complete thing to the lcd in lcd_update()?
>
>If you have, the ui simulator is gonna work pretty much
>unmodified with your
>new code.

Yep, it still works the same way. The main change was to generalize the
character drawing code to draw arbitrary bitmaps, so I could draw icons,
etc. A character is just a bitmap, so lcd_string() now calls
lcd_bitmap() for each char.

I did change the format of the memory buffer to make the code easier.
(X and Y indices are swapped from the old code.) So if you have code
that depends on this, a small change is needed.

>Good job! Can we get the windows simulator added to CVS too? We could
>possible make the uisimulator in two directories, one for X and one for
>Windows. I don't think there's any point to put them in the
>same dir, as they
>won't really share any code. AFAICS.

Sounds like a good idea to keep them separated. Maybe another directory
for the firmware implementation? As the project grows, I'm sure we'll
want to separate out other major pieces of the code.

>I thought about toying up a filesystem api and simulate that
>(until we have a
>real file system), then we could even start fiddling around
>with file lists
>and directory hiearchies etc.

Sounds good. We could use this to get started working on browsing UI.
There's plenty of work to do.

Thanks,
Gary
Received on 2002-03-27