thumbnail mtime, ctime

[Dr. Michael J. Chudobiak]

> You are correct about this. But in my opinion the consequence of this
> should be another change of the spec. It should be rephrased to say that
> the empty PNG file should only be created if thumbnail generation failed
> even though the file was opened successfully. It seems wrong to me to
> save a failure PNG only because the file cannot be accessed. And it
> would be relatively easy to change the implementations accordingly.

As a follow-up, Nautilus and gThumb have been patched to skip 
thumbnailing entirely on files with insufficient read permissions, so no 
failed thumbnail is read or written.

Now... is anyone in a position to modify the thumbnailing spec? Or is it 
maintainerless?

- Mike