Thomas Adam

This also the case on fvwm2, because the server is grabbed performing the function; and with new_desk packets, the same operation blocks (therefore queuing the events) until the button is...

Hi @harridu Could you provide a short video demonstrating the issue you're seeing, please?

@harridu Can you provide further details, please?

Hi, I can't reproduce this on FreeBSD-12. Rather -- I *can* reproduce it if I use `fvwm2` from FreeBSD ports. But if I compile either `fvwm2` or `fvwm3` myself, then...

Hi @pm215 Yes indeed -- however, I do have to also consider a degree of portability. My understanding from reading about the `poison` pragma before is that it's C99 only....

Cool, then I'll use the pragma. Indeed, we've been using things like `inline` for ages -- so I'm not at all convinced c89 compatibility is important. However, even compiling with...

> Well, I don't personally care whether the code base is strict C99-only. -std=gnu99, ie "C99 + GNU extensions" works. At least some of the problems with a -std=c99 compile...

Hi @NsCDE Hmm -- this seems to be working for me... Here's an example: ``` DestroyModuleConfig FEn:* *FEn: monitor_focus Echo $[monitor.current] Module FvwmEvent FEn ``` Then in a separate terminal:...

OK... Are you able to reproduce this by using a minimal config which isn't using NsCDE? I still can't reproduce this.

Hi @NsCDE Is this still a problem? For me, the following works as I would expect: ``` DestroyModuleConfig FEn:* *FEn: Cmd echo *FEn: monitor_focus "MONITOR FOCUS CALLED $[monitor.current] on $[desk.name0]...