Yes, there are some differences from RC1 to RC2:
RC2 is much improved in general. I am getting far fewer Event Veiwer log
entries about things failing. Notably the snap-in problem I noted in RC1 has
not reappeared. That one was particularly annoying because I couldn't load
snap-ins to try and further any diagnosis.
The operating system also feels a little quicker. A little quicker, not a
lot, but this is perceptual and I have no hard evidence either way, though it
would be logical for this to occur.
I turned off UAC and signed driver enforcement (well, sorta, I gotta keep
hitting F8 at boot). The experience is so much more pleasant without UAC. I
stand by my earlier critique of UAC. I won't belabor the point.
Indexing also seems to banging away at my hard-drives with less ferocity
than RC1. I got skin burn from the photons emmitting from the hard-drive
light before. But on that note, indexing is SO SLOW. At this rate it'll
take nearly two days of solid activity to index my drives, and I turned off
everything I could except indexing of multimedia files (mp3, wmv, etc.).
There were a few extraneous file types that couldn't be unchecked, so those
of course remained.
I haven't really started banging away at the O.S. yet. That's where it
really fell apart with RC1. This also gives me a clue as to why some people
feel the experience is perfectly acceptable with Vista. If you don't do too
much, or expect too much, its just fine. I can pop around and do thsi and
that and its reasonably responsive. But as soon as I start to do anything
really heavy, it grinds to a halt in ways that XP Pro x64 doesn't. That
points to fundamental issues with the O.S., and not cosmetic fixes prior to
going "Gold."
All in all, I'd say Microsoft's programmers have made remarkable
improvements. RC1 was no Release Candidate. RC2 is much closer.
I still have some issues.
There are the usability issues (which I know others are echoing) I mentioned
before. Navigating with Explorer feels clunky and difficult. Some of that
is from me being so used to XP explorer, however as previously noted, aside
from this version of Windows each prior upgrade presented the user with a
*more* intuitive interface. This is the first time its been a less intuitive
interface. IMHO a LOT less intuitive.
I liked the ability to customize explorer. Here there is less ability to do
so. I had hoped this version would be ultimately configurable. Many things
are more configurable than XP, but I suggest that is more a failing of XP
than a nod of confidence for Vista. But many things are also less
configurable, and the O.S. is certainly less configurable than it ought to
be. There's a myriad of things this applies to from the Sidebar through to
how dialog boxes display files. I poked around in the registry abit, but I
didn't want to look up Vista's skirt that much. A couple explorer functions
I really miss are copy-to and move-to. Maybe they're there, I didn't see
them. And there's no button for them like you could add to Explorer in XP.
Other niggles: Explorers estimations of how long it takes to do things are
sometimes pretty close, and sometimes wildly off. Much worse than XP's
optimistic view-point of file transfer times. For example, I copied 1GB of
data from one drive to another. Both were reasonably quick SATA drives. The
estimation came back that it would take 1 day and some amount of hours. Part
way into the copying, Explorer now reported 30 seconds remaining. That
seemed just as optimistic as Explorer was pessimistic only moments before,
and sure enough it took another 5 minutes or so. So that algorithm needs to
be refined.
I also note that even though I'm running the x64 version on a Dual Core
processor that the O.S. does a sub-optimal job of allocating tasks to
processors. It heavily favors processor 1 (out of 2 processors, 0 & 1). I
would have thought were it to do this at all it would favor 0, but it
doesn't, and its very consistent about its preference. THAT is functional
flaw that goes to the core of the O.S. There aren't a lot of things more
basic than processor scheduling. I mean there are things much more basic,
but processor scheduling on a multi-processor rig is still a very basic core
function. Stuff like that needs to be addressed. Should have been
addressed. I've known companies whose only focus was processor scheduling in
multi-processor environments. It shouldn't be an afterthought, and "oh yeah,
lets take a look at that after we get some more glass in this window border."
The system isn't without its crashes. Explorer has crashed once. That's
better than XP, but not as good as XP X64 which I've never had Explorer crash
on.
So that's it. In a nutshell:
-Much improved
-Needs more testing to determine stability
-Interface could still using some usability modifications
-Aero is very pretty
-UAC es no bueno
;-)