Known issues

While we appreciate all feedback, please remember that we offer no official support for beta \ preview \ developer channel versions of Windows. If you report issues while in any of those early release channels, you should not expect them to be addressed unless it still exists in a retail version of Windows.

• tbd

Excellent - so far, so good - the history works for me now and has kept all the old entries.

Thank you for the positive feedback, Frank, I will pass it along.

Sean Drohan
Stardock Product Lifecycle Manager

maybe we need can customize dark theme , coz most windows user like dark theme

I have a trial version of ObjectDock and so also DesktopGPT. I am using the GPT-4 model, and sometimes the GPT-4o model.

I am enjoying testing it out, using it for getting information about coding with Python and other programming tools in Foundry Nuke. This is a pretty niche area.

However, I am often getting very garbled output.

It will often cut off the top sets of lines.

For instance this is the top part of some output.

the current pixel's x and y coordinates, respectively. The kernel then inverts the color of each pixel by subtracting each color component from 1.0.

Thisget approach is_global_id applicable to is all versions a function of N used withinuke that Blink Script support Blink in The Script, Foundry but it's N's alwaysuke for a good working with idea to kernel functions refer to. Blink the specific Script is version's a GPU documentation you-accel're workingerated with for node for any version image processing-specific features that allows or changes.

you to write custom effects. It's designed to be fast and efficient for high-end visual effects production.

Its missing the top number of lines that should be there. 

And the next two paragraphs are garbled.  Text within text.

I ASSUME that Chat GPT-4 wouldn't return such garbled text....

If I run the exact same command again, it will usually return information that is not garbled.

Here is an example. You can see that the text is not the same, garbled or not. But the information is relevant.

get_global_id is a function used in Blink Script within The Foundry's Nuke, a node-based digital compositing application. Blink Script is used for writing custom image processing operations in Nuke, allowing for high-performance, GPU-accelerated effects.

In the context of Blink Script, get_global_id is a function that returns the global ID of the current processing element (such as a pixel) that the script is operating on. This is particularly useful in parallel computing contexts, such as when running on a GPU, where you need to know the specific element (e.g., pixel) out of a larger set (e.g., an image) that the current thread or work-item is tasked with processing.

The function typically takes a single integer argument, which specifies the dimension (0 for x, 1 for y in a 2D image) of the global ID you are querying.

Not sure what to make of this...

I do know that its wasting my tokens, because usually the garbled return is useless for understanding what's going on, and it also often garbles the actual example code.

The template and command I'm using are as follows:

I'm using an Inline Template

Shortcut: /nuke 

AI Directive: Return information about The Foundry Nuke application. Using TCL, Python, Blink Script, and the Nuke Python API. Assume knowledge of the Nuke interface. Give example code, and say if this applies to specific versions of Nuke.

-------------------------

Example command: 

/nuke What is get_global_id in Blink script?