Following the descreening tutorial using Fourier Transform but when I try to do the initial (Direct) transform I get the following error

*** Error in ./gimp_fourier/ *** Command '-endlocal': Not associated to a '-local' command within the same scope.

If I do Fourier Analysis that seems to work (gives two layers) but if I modify the relevant one and try to do the Inverse transform on the modified layer it says that the image does not appear to be a fourier transform.

Any suggestions?

You are right Harry, this filter has been temporarily broken.
It should be fixed now, just refresh your filters and it should work again.

Thanks Ronounors,
Now works a treat.
Just a question. How far should you go with blotting out the 'hot-spots' in the transform. I did about 40 of the strongest ones and the result is quite good. Would it be worth doing more of the minor ones or is it likely to be a case of diminishing returns?

That's a good question Harry, and unfortunately I don't have a clear answer about that.
The fact is that the main thing you see in the Fourier transform (upper image) is actually the logarithm of the module, and thus, the brightness of the hotspot should not be perceived in a linear way. Two hotspots with small difference in the amplitude can be in fact represents two frequencies, one being clearly visible while the other is not.
I think that's why only removing the very brighter hotspots often render an acceptable result. Probably the more you remove, the best result you get, but this quickly becomes invisible when the brightness of your hotspots decreases.

I detected the same error message when using this filter.
After refresh the message did not reappear and the filter worked well again.
HarryB thank you for pointing out this flaw, so it could be corrected.
Ronounours thanks for the quick resolution of the problem!
I really like this filter!