Re: OT: more Win7 fun

"Snit" <usenet@xxxxxxxxxxxxxxxxxxxxx> wrote in message news:C718A69B.524CC%usenet@xxxxxxxxxxxxxxxxxxxxxxxx
Daniel Johnson stated in post Tvqdncm3LcHEzG7XnZ2dnUVZ_qWdnZ2d@xxxxxxxxxxxxx
on 11/5/09 3:27 PM:

I do not know what happened; perhaps it is a bad interaction between your
virtual machines 'resize the screen' function and IE.

Could be. Also use qres to resize... might be something funny with that?

I do not know. That is kinda old.

Parallels has a mode where windows from Windows show up on the Mac desktop,
right? That's got to be intrusive. Maybe that's installed and is somehow
interfering?

No... was not using "Coherence" or the new "Crystal" modes. Rarely do.

But were they installed? I imaging there's a component that must be installed in Windows for this to work.

[snip]
I see in your attached movie that it is lagging really badly. Your VM is
exhibiting a performance problem, to be sure. It is *not* that slow on real
hardware.

(Nor was the release candidate, just to cover that base)

Everything else is quite snappy... but could be wonky video driver.

Yes, but you'd need a completely awful driver to get what that movie shows. Think "VGA frame buffer"- and I don't think you get the glass if you have a driver like that.

Anyway, resizing is one of the toughest things to make snappy.

[snip]
The RTM is build 7600. That may explain some of your troubles, though this
is exactly the sort of cosmetic bug I'd expect MS to not fix at the release
candidate stage.

I still favor the "Parallels integration is doing it" theory.

Could be... but that would mean drivers can lead to such wonkiness... and
things will not be consistent from one computer to another.

Drivers can do all sorts of awful things, though I would be more inclined to suspect Parallel's user mode components. It surely must have some.

[snip]
Yeah. Certainly performance will always be affected by that, and dual CPU
support seems like a likely problem area. The window resize/placement issues
may also be caused by Parallels' desktop integration features.

I merely guess, but it sounds plausible.

Not using any of the integration features. They keep tossing more and more
of those in. I can understand why, but likely leads to a reduction in
productivity and the like.

They are certainly invasive for Windows. It's a clever hack, but it's a hack.

The "official" way to do this stuff is with RDP, but the performance of this is likely to be inferior.

[snip]
I do not think it's using 2 cores. I can't see enough of task manager to
know, but I have tried this here and I can't get it more than a bit above 1
core's worth no matter how hard I wiggle.

See the two bars in the task manager... that means two cores... right?

If you have two cores (or hyperthreading) you should see multiple CPU *graphs*; the line charts to the right multiply. The bars on the left do not.

[snip]
Anyway, clearly your 'Move' command is busted in the same way as your 'Size'
command. I assure that this is *not* normal Windows 7 behavior; when the
window is maximized, "Restore" should be enabled, and "Move" and "Size"
disabled.

Maybe fixed in the final. If so then no big deal. Even if not, it is not
exactly a show stopper, but it would be a sign of not focusing on details.

I still think it's Parallels. MS does not focus so much on these kinds of details- and so, they would not have fixed this between release candidate and RTM.


.

Relevant Pages

  • Re: OT: more Win7 fun
    ... Also use qres to resize... ... Parallels tools were installed. ... my download folder as a drive for Windows. ... but you'd need a completely awful driver to get what that movie shows. ...
    (comp.sys.mac.advocacy)
  • Re: Wireless cards
    ... > Under Linux a Linux driver is a good thing. ... Under Windows a Windows ...
    (Fedora)
  • Re: ESET NOD32 or ESET Smart Security?
    ... Even amateurs can break a WEP key in a matter of minutes with software downloaded from the Internet, and thereby make you the neighbourhood free bandwidth supplier. ... an updated driver for my wireless card; ... driver updates are better downloaded form the manufacturer's support site than from Windows Update. ...
    (microsoft.public.windows.vista.general)
  • Re: How to intercept traffic on different OSs?
    ... > to write intermediate driver, but what should I do if I want to modify ... libdnet WIN32 VC6 project ... Yea though I walk through the valley of worms, I will fear no Windows. ...
    (microsoft.public.win32.programmer.networks)
  • Re: [opensuse] natively supported external wireless cards?
    ... found: KeReleaseGuardedMutex ... Which Windows version of the driver are you trying? ... I believe you should NOT use a Vista driver with ndiswrapper or Linexant, but should use a compatible XP driver. ...
    (SuSE)