Clock
BNig
bernd.niggemann at uni-wh.de
Fri Jun 5 05:19:28 EDT 2015
Roger,
this is probably due to the limited processing power of the android-watch
This is similar to iOS back when I made the scrollers. At the time iPhones
had trouble mixing all the blend modes fast enough.
So I build in a cheat. The command is "makeiOSReady" it is located in the
script of the group. What it does is to make a snapshot of the scroller and
hide all the graphics. What is left on is an image "mock" and layered on top
of this field "fDisplay-Scroll"
For the first 3 of the clock faces try
on mouseUp
dispatch "makeiOSReady" to group "Hours"
dispatch "makeiOSReady" to group "Minutes"
dispatch "makeiOSReady" to group "moveSeconds"
end mouseUp
once you are done with your resizing of the groups because a resize of the
group, e.g. "hours" will remove the image "mock" and you do your resizing on
the graphics.
That is the reason why you issue the "makeiOSReady" command at the end, when
you are done configuring the groups.
This should put considerably less burden on the processor of your
android-watch.
If that does not make it work try disabling the card script that set the
color.
Also you may play with the wait x milliseconds with messages that is issued
twice, once in mouseUP handler and once in "advanceSeconds" handler of the
start/stop button
send "setWheel tHours + 1" to group "hours"
send "setWheel tMinutes + 1" to group "Minutes"
send "setWheel tSeconds + 1" to group "moveSeconds"
wait 5 milliseconds with messages -- increase the milliseconds to
wait, gives more time.
What I see in the video is very much what took me long to debug and which
turned out to be the wait... being too short.
Note, "makeiOSReady" only works on the non-egg shaped scrollers. For
egg-shaped scrollers I could look into them next week when I am back from a
weekend trip and see if I could tweak them to work more efficient. But there
is a lot of computation going on behind the scene.
Try these suggestion in the given order and see what makes it work. Please
tell us if you get it to work.
Kind regards
Bernd
Roger Eller wrote
> Bernd,
>
> I finally found a few minutes to play. I think possibly the layerMode may
> need a tweak to get smooth movement on Android. It looks great and smooth
> on the computer. Here's a video of the first run on my watch.
>
> https://dl.dropboxusercontent.com/u/54789013/Android/SmartQ/LordEllerWatch.MOV
>
> ... and this is the shrunken stack (last modified in 6.6.2).
> https://dl.dropboxusercontent.com/u/54789013/Android/SmartQ/LordEllerWatch_0_2.livecode.zip
>
> ~Roger
--
View this message in context: http://runtime-revolution.278305.n4.nabble.com/Clock-tp4692455p4692972.html
Sent from the Revolution - User mailing list archive at Nabble.com.
More information about the use-livecode
mailing list