On 6/28/2024 5:04 AM, Eli Zaretskii wrote:
> I'd prefer to defer the decision about skipping until we fully
> understand what happens here.  If the output below doesn't give a
> clue, could you perhaps explain to me the idea of this test in more
> detail, so I could perhaps also think about the reasons?

I had a sudden realization about what the problem might be. I *think* 
this patch should resolve the issue. Could you try it?

The issue (I believe) is that on MS-Windows, the "file:" URL to the 
image we're trying to load wasn't formatted correctly. That meant that 
'url-retrieve' never successfully loaded the image, so the test stalled.

If that *doesn't* fix the issue, here's a quick summary of what the test 
is doing:

1. Load a simple HTML page with an image (using one from the test directory)
2. Wait for SHR to insert the real image, which happens asynchronously.
3. Put point on the image and call 'shr-zoom-image'.
4. Wait for SHR to insert the updated image.
5. Check that we have a new (sliced) image in the buffer.

(The test repeats the above steps for a few different cases, each with 
different values of the HTML "alt" attribute.)