add Loop, NoLoop, IsLooping and FrameCount

[psampaz]

closes #32 and #39

[codecov]

Codecov Report

Merging #41 (6ea7f0e) into main (6ea7f0e) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main      #41   +/-   ##
=======================================
  Coverage   80.59%   80.59%           
=======================================
  Files           9        9           
  Lines         567      567           
=======================================
  Hits          457      457           
  Misses         89       89           
  Partials       21       21           

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 6ea7f0e...fb788c7. Read the comment docs.

[psampaz]

@sbinet even if this test https://github.com/go-p5/p5/pull/41/files#diff-e65edbcd5a006378e11ec33d7b121f21a20ec2631433f0b75bfa54e11768e9caR85 passes, if you manually try to run one of the repo examples with NoLoop() the window that open has just a white background.

Is this somehow related with how Gio works?

[sbinet]

Yes and no.

As Gio is an immediate mode toolkit, one has to redraw the background for every new frame.

That's done here:

p5/proc.go

Line 293 in 6ea7f0e

paint.Fill(ops, clr)

... in 'draw'.

I'll check tomorrow what the equivalent p5js code produces.

[sbinet]

I've just checked with p5js.

the following:

function setup() {
  background(0,0,0);
  noLoop();
}

function draw() {
  background(255,0,0);
  text(frameCount, 10, 30);
}

produces a red canvas with the number 1 being displayed.
see on https://editor.p5js.org/.

so it would seem p5js draws at least 1 frame and then enters the "frame loop".
ie: (in pseudo-code)

proc.Setup()
// create Gio window, then draw one frame.
proc.Draw()

for evt := range events {
   // ...
   proc.Draw()
}

do you want to give it a stab?

[sbinet]

I'd move this inside the setup func.

also, I think that to test the noloop/doloop state, I'd use the value of FrameCount as I did in the little code blurb in p5js.

WDYT?

[psampaz]

I updated the test.

if the incFrameCount is done with defer the image draws 0. I removed the defer and it is now 1 which is normal.

This does not solve the issue I have here #41 (comment)
Running the actual real example like solar-system it does not work.
The code works, it calls draw once, it increases the frame count by 1, stop the loop but the output is a white screen.

[sbinet]

I'd propose to merge this in and file a new bug report regarding this behaviour.

WDYT?

[sbinet]

I'll take that as a yes and proceed with merging that PR.

[sbinet]

with the code as is, framecount_golden.png displays 1, as does the p5js code.
but according to the documentation of FrameCount (that returns the number of frames displayed since the program's start), some people could interpret the correct answer to be 0.

some (re)wording of the documentation could be devised (but could also be deferred to another PR) to address this 1-off effect.
I suppose it probably isn't too much a big deal however.