Clipper1

[Qengineering]

PaddleOCR with the Paddle-Lite framework still uses the original Clipper1 version.
I tried to migrate to CLipper2, but it become very cumbersome.
Is there a way to obtain the 'old' Clipper? I'm using mostly Linux aarch64 machines (Raspberry PI, RockPi5 or Jetson Nano).

[AngusJohnson]

Without a LOT more detail, and from my POV you haven't provided any, I won't consider this as an Issue.
(That's why I've transferred this to Discussion.)

I tried to migrate to CLipper2, but it become very cumbersome.

Please explain what you mean.
Has the migration process been cumbersome, or are you finding Clipper2 performance 'cumbersome'?
Or is it something else?

Is there a way to obtain the 'old' Clipper?

Sure, it's still here.

[Qengineering]

To be clear, neither Clipper1 nor Clipper2 is cumbersome. Both are high-quality software, with state-of-the-art coding.
Replacing the many Clipper1 calls in the PaddleOCR cpp code is time-consuming and error-prone.

The provided link to Clipper1 solved the problem. I compiled the cpp library and got the PaddleOCR software running in a jiffy