virtualbox.org

loukingjr wrote:How intermittent? I haven't noticed anything.

Either your VMs have GAs installed or it's just not showing up. It probably depends on the host system and how a VM is configured plus some randomness factor.

well the OP and myself are both on OSX 10.10 Yosemite and I thought you had to have the GAs installed for mouse integration, so I don't get it.

is there a bugticket or another post about this?

Hi, yes it is intermittent - sometimes the mouse works and next time it does not. Same with the Mac track pad; it never worked before but now does intermittent with the latest release.
If this is going to be fixed in the next update, I'll wait for it.
Do u have any idea when that will be released? Tis

Strange, I've never had the Touch Pad, Magic Mouse, or the Apple Mouse not work with any version of VirtualBox or OSX

Edit: On three different Macs.

MichaelJF2 wrote:If this is going to be fixed in the next update, I'll wait for it.
Do u have any idea when that will be released? Tis

Yes, it's going to be fixed. I can't tell you exactly when the next update will be released but it's likely to be a matter of days rather than weeks.

Can you tell me how to break mine?

loukingjr wrote:Can you tell me how to break mine?

No If you have a VM that uses the USB tablet emulation (not Guest Additions) and don't have any problem moving the mouse cursor inside the guest, consider yourself lucky

I've seen the problem on at least 3 different systems and it was never consistent. Others never saw the problem at all. With all bugs that have some kind of timing component, some systems are far more likely than others to exhibit the problem. It could well be that yours is on the near-zero side of the probability scale.

actually all my guests have the GAs installed. I think I have about 25 (didn't feel like counting). All my guests are set to USB Tablet.

I have seen what I thought was a timing issue on occasion. My Windows 7 guest either immediately connected to the network, took 15-30 secs to connect, or never connected (unless I didn't wait long enough). I never worried about it but I thought it might be caused by Avast trying to call home. I uninstalled it and it hasn't been an issue since.

@loukingjr: Take a look at my original report two days ago in the Discuss the 4.3.24 release thread as well as michaln's explanation. Until he stepped in to say that "yes, it exists", I thought I was going nuts. I was trying to get something repeatable for 2 days while banging my head at the wall.

socratis wrote:@loukingjr: Take a look at my original report two days ago in the Discuss the 4.3.24 release thread as well as michaln's explanation. Until he stepped in to say that "yes, it exists", I thought I was going nuts. I was trying to get something repeatable for 2 days while banging my head at the wall.

I had read that just after you posted it. Since it was exclusively about running OSX guests on OSX hosts I didn't pay much attention to it since I don't run OSX guests in VB anymore. I just chalked it up to the lack of GAs for OSX guests.

Michaln just confused me when he said it was an intermittent issue. Apparently it's not intermittent at all but occurs 100% of the time with OSX guests and as near as I can tell, 0% of the time on non-OSX guests.

loukingjr wrote:Apparently it's not intermittent at all but occurs 100% of the time with OSX guests and as near as I can tell, 0% of the time on non-OSX guests.

Close... It's intermittent with guests that do not have GAs (like OSX) and 0% with the rest, AFAIK. It does not happen 100% with OSX guests. I'd say about 1/3 to 2/3 of the time. Start it three times and you'll get lucky once.

I see. I'm so used to installing GAs in everything and not running guests that have no GAs I would never see it.

::scratching head:: I thought I had a handle on this "bug" but apparently Dementia has set in. My understanding was/is mouse integration never works without guest additions being installed so I don't understand how mouse integration can be intermittent in guests without guest addition support. It shouldn't work at all.

I installed Sparky Linux this morning and purposely left the guest additions uninstalled. I was not surprised that mouse integration didn't work. So I just disabled mouse integration (Host+I) so when I clicked inside the guest it captures the mouse and I have to press the Host key to release the mouse to the Host. It's worked that way as long as I can remember. Maybe my memory is not too good.

The only reason I keep trying to figure out what people have been saying is because one workaround mentioned is to fallback to VB 4.3.22. But that has it's own issues.

Anyway, mine works normally in VB 4.3.24.

loukingjr wrote:::scratching head:: I thought I had a handle on this "bug" but apparently Dementia has set in. My understanding was/is mouse integration never works without guest additions being installed so I don't understand how mouse integration can be intermittent in guests without guest addition support. It shouldn't work at all.

That's how it was originally. Several versions back (I don't remember exactly when), VirtualBox added USB tablet emulation which provides mouse integration for most modern guests even without GAs installed (or before they're installed). And the USB tablet is on by default for almost anything remotely recent (for example with Windows guests it's Windows 2000 and up, or all OS X guests).

The fallback to PS/2 mouse emulation is still there. If the guest never initializes the emulated USB tablet device, the emulated PS/2 mouse can be used as long as the mouse is captured by the VM. Or the mouse integration can be manually disabled, in which case the PS/2 mouse emulation again works the same way it used to.

That said, OS X guests are tricky because they don't have any PS/2 mouse driver, so if the USB tablet emulation doesn't work, the fallback won't do any good.

I see. I didn't realize the USB Tablet setting had implemented mouse integration without the GAs being installed. I learn something new every day, or week, or month, okay, sometimes years.