[Vm-dev] [OpenSmalltalk/opensmalltalk-vm] sqUnixXdnd: Don't record SQDragLeave when XdndDrop is handled (#508)

[Christoph Thiede]

Hi David,

thanks for sharing your observations, this fits in with my own results.

> By the way I did the same 3 tests with a VM which I compiled from the OpenSmalltalk Cog branch (not from the PR).
> 
> I seem to be having the same results.

So there appears not to be any regression.

> But basically I don't understand what you are trying to fix.

Here is how you can observe the erroneous under-the-hood behavior this PR aims to fix:

1. In a recent Squeak image, browse `HandMorph >> #generateDropFilesEvent:` and insert the following line after the assignment of `dragType:`
   ```smalltalk
   Transcript showln: dragType.
   ```
2. Open a Transcript
3. Drag a file into the image

Transcript output on an X11 platform using the latest released plugin version:

```
1 (dragEnter)
2 (dragMove)
2 (dragMove)
...
2 (dragMove)
3 (dragLeave)
4 (dragDrop)
3 (dragLeave)
```

Transcript when compiling the plugin with the proposed patch applied:

```
1 (dragEnter)
2 (dragMove)
2 (dragMove)
...
2 (dragMove)
4 (dragDrop)
```

The current image implementation does not care about the additional, false dragLeave events, but my changeset which I am planning to put into the inbox will do so.

Is this comprehensible? 😅

PS: For some reasons all your posts arrive multiple times in the repository. Also, they would be easier to read if you could use codefences (` ``` `) :-)

Best,
Christoph

-- 
You are receiving this because you commented.
Reply to this email directly or view it on GitHub:
https://github.com/OpenSmalltalk/opensmalltalk-vm/pull/508#issuecomment-676129131
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squeakfoundation.org/pipermail/vm-dev/attachments/20200819/37fcb777/attachment.html>