Let me tell you about my Viser

[senyai]

Hello everyone! Ten years ago, I created a program with an idea similar to Viser. Unfortunately, it is based on proprietary libraries and cannot be published. It allowed for an infinite canvas for visualization, so one could:

c.add_text(str, font_and_size, brush)
img_id = c.add_image(numpy)
c.add_line(..., line_width, parent=img_id)
point_id = c.add_point(x, y)
c.set(point_id, "x", new_x)
# and a little bit of layout management

Can viser development move a little bit in similar direction? The Three.js viewer would then become just another widget. I see people wanting to add images #532. I would use it to visualize AI image training progress.