Vista - Joomla & Vista

dmex,

Okay, turning off UAC slows down boot time. That's a lot different than slowing down Vista performance.

Where is there any documentation from anyone that shows it slows down Vista after it has booted?

S-

The documentation is the "Operational" Performance log found above, If you want to check it yourself do (StartupTime + With UAC) and (StartupTime + Without UAC) = AverageBootTime with and Without UAC and you have your answer...I found the Average booted faster with UAC

dmex,

Wow. Did you even read my question?? Slowing down boot up time is a lot different than slowing down Vista performance.

Where is there any documentation from anyone that shows it slows down Vista after it has booted?

S-

I did and why do I have to repeat myself?

If you check your Operational Log, do your math and calculate the average time with and without UAC for boottimes and applications than you will have your answer...I dont know what documentation you want because that average will not apply to every VistaPC since each one has different hardware and software configurations but there is a noticeable difference in the results for application and bootime performance.

dmex,

You keep repeating the same stuff about boot times and that does not answer my question.

All that you have shown is that boot time is slower with UAC disabled. The link you provided discusses boot times only. There is nothing about Vista performance after the boot process is finished and that is what I am asking about.

The Event Viewer log on two different system shows nothing about applications taking longer to start up.

I can find nothing on the Internet about these delays with UAC off either. You would think someone would be talking about them.

S-

dmex,

Look at this document from Microsoft:

http://download.microsoft.com/downlo...Vista_perf.doc

When doing performance testing, Microsoft recommends leaving UAC enabled but set to "Elevate without Prompting".

So that kills the idea that leaving UAC enabled but set to "Elevate without Prompting" will affect performance. I still can't find anything that says turning off UAC causes any performance issues other than a slower boot time (big whoop!). Do you have anything???

S-

Did you follow Microsoft's example and leave UAC enabled while doing benchmarking? There is a variable difference in program performance depending on IO usage (amount of running processes, drivers and other code) while benchmarking.

Do you know what virtualization actually does to x86 code on x64? It reorders all instruction sets (basically the whole program) and this does cause performance issues on x64 if the application is not coded correctly.

The code im using for a soon-to-be-released utility does have issues with execution time and do vary depending on both UAC and the virtualalized driver being enabled on systems.

If I tried to minimize the number of locks taken out to guard the consistency of data structures accessible from multiple threads it causes the execution time to vary from instant to 10 minutes...the x86 memory model provided a level of safety by guaranteeing that all write operations occurred in order and on time.

This guarantee does not happen with 64-bit because it explicitly requires it before accessing the 64bit instruction altogether and while this is good for end-users...It gets worse (for poorly written applications) when virtualization gets involved and code that reads data without the protections of locks can experience unexpected results caused by the processor reordering instructions (virtualization) and the potential for unexpected code execution results arises.

Now this isnt a problem for end-users since developers refine the code untill its executing timely and properly without variable results but it does cause me headaches trying to code directly for x64 systems with virtualization envolved.

---Facts aside---

If you want me to prove how virtualization makes applications slower go try running a few resource-hungry programs in a virtual machine, thats the only way you will see first hand how virtualization slows performance down without building your own application.

I also dont want to continually argue the point of UAC being disabled makes application performance & execution time longer, I know it does from first hand experience coding applications myself...So Im pulling rank and thread closed