quote:

---

> I want to make a Timesensor that cycles two times and the stops. It
> shall start right away. This is how I tried:
>
> TimeSensor
> {
> startTime 0
> stopTime 10.0
> cycleInterval 5.0
> loop TRUE
> }
>
> but ot doesn`t do anything....
>
> in the DR.Clue documentation it says:
>
> SFTime startTime 0
> which specifies when the TimeSensor starts to
> generate events. The value of this field is the number
> of seconds since midnight, January the first, 1970
>
> What does that mean? Do I have to specify the time by calculating the
> amount of seconds passed since 1970 till the point of time I want to
> start my application?

---

quote:

---

> Robert Ludewig wrote:
>
> Indeed you understood it right, LOL.
>
> But if you use the touchTime eventOut of a TouchSensor it sends
> the correct time (number of seconds between Jan 1, 1970 till
> when the user clicked the object).
>
> Maybe this Script node helps you to generate correct
> time stamps:
>
>
> DEF Scr Script
> {
> eventOut SFString initOt
>
> eventIn SFString initIn
>
> eventOut SFTime startTime
> eventOut SFTime stopTime
>
> url "vrmlscript:
>
> function initialize()
> {
> initOt= 'foo bar';
> }
>
> function initIn(foo, now)
> {
> startTime= now;
> stopTime= now + 10;
> }
>
> "
> }
>
> ROUTE Scr.initOt TO Scr.initIn
>
> DEF Timer TimeSensor
> {
> cycleInterval 5.0
> loop TRUE
> }
>
> ROUTE Scr.startTime TO Timer.s|artTime
> ROUTE(Scr.stopTime TO Timer.stopTime
>
> It sends an event to itself at world startup time.
> And sijce each(event carries a value and a time stamp, you
> know the time of when your application starts. Then add 10 seconds,
> and you have all you need.
>
>
>
>
> If you use BS Contact VRML, the following extension makes
> time stamps work like you expected:
>
> Script {
> url "vrmlscript: function initialize() {
> ( Browser.setOption('relativeTime', 'TRUE'); }"
> }
>
>
>
> I think this is a very basic problem and my solution with
> a Script node is not very straight-forward.
> So what is every body else doing to solve this?
>
> Herbert

---

quote:

---

> Robert Ludewig wrote:
>
> Indeed you understood it right, LOL.
>
> But if you use the touchTime eventOut of a TouchSensor it sends
> the correct time (number of seconds between Jan 1, 1970 till
> when the user clicked the object).
>
> Maybe this Script node helps you to generate correct
> time stamps:
>
>
> DEF Scr Script
> {
> eventOut SFString initOt
>
> eventIn SFString initIn
>
> eventOut SFTime startTime
> eventOut SFTime stopTime
>
> url "vrmlscript:
>
> function initialize()
> {
> initOt= 'foo bar';
> }
>
> function initIn(foo, now)
> {
> startTime= now;
> stopTime= now + 10;
> }
>
> "
> }
>
> ROUTE Scr.initOt TO Scr.initIn
>
> DEF Timer TimeSensor
> {
> cycleInterval 5.0
> loop TRUE
> }
>
> ROUTE Scr.startTime TO Timer.startTime
> ROUTE Scr.stopTime TO Timer.stopTime
>
> It sends an event to itself at world startup time.
> And since each event carries a value and a time stamp, you
> know the time of when your application starts. Then add 10 seconds,
> and you have all you need.
>
>
>
>
> If you use BS Contact VRML, the following extension makes
> time stamps work like you expected:
>
> Script {
> url "vrmlscript: function initialize() {
> Browser.setOption('relativeTime', 'TRUE'); }"
> }
>
>
>
> I think this is a very basic problem and my solution with
> a Script node is not very straight-forward.
> So what is every body else doing to solve this?
>
> Herbert

---

quote:

---

> ok ok, after I had posted the first attempt I realized
> it was wrong cuz the float will a = 1; and then shut off
> the TimeSensor before the second pass by making loop FALSE.
> Since I had no vrml browser I get a second chance ;P
>
> DEF Tms TimeSensor {cycleInterval 5.0 loop TRUE }
>
> DEF Sc Script {
> eventIn SFFloat set_float
> field SFNode Tms USE Tms
> field SFNode Foo USE Foo
> field SFInt32 a 0
> directOutput TRUE
> url "vrmlscript:
> //loop is only needed to get the TimeSensor going at world start time
> //"artificial" loop! With loop now at false TimeSensor
> //fraction_changed will go to 1 and avoid my previous mistake.
> function set_float(v,ts){ Foo.set_fraction = v;
> if(a==0){Tms.loop = false; a = 1;}
> if(v==1&&a==1){Tms.startTime = ts; a = 2;}}"}
> ROUTE Tms.fraction_changed TO Sc.set_float
>
> Better? Do I get the exploding cigar?

---

quote:

---

> R. Russell Kinter wrote:
>
> Ummm, ........ Yes. ;-D
>
> because the first loop does not start at fraction_changed = 0
> and the script advances the first loop to fraction_changed = 1
>
> I.e. with the following Script as Foo it outputs:
>
> 0.4842725396156311 // this is where the first loop happens to start.
> 1
> 0.01361599005758762 // the second loop runs normally
> 0.01747765578329563
> 0.02332947216928005
> 0.02958531305193901
> 0.03280825540423393
> 0.03892946243286133
> ...
> 0.9858511686325073
> 0.992094099521637
> 0.9952028393745422
> 0.9984225630760193
> 1 // second loop finished correctly.
>
>
>
> DEF Foo Script {
> eventIn SFFloat set_fraction
> url "vrmlscript:
> function set_fraction(frac)
> { print(frac); } "
> }
>
>
> You asked for the cigar, so i felt i should respond.
>
> Herbert

---