If full emulation mode is supported, why can't I load my Dell P1000 Laser Printer driver on my Vista machine? It works on WinNT through WinXP. The laser printer is only about 4 year old. It's not like it's a recycling refugee that's been hiding in my closet for 10 yrs. If Vista supports compatibiltiy modes, this should not be the case - I should expect to maintain base functionality, just lower performance. Much like not having a 'proper' driver for your graphics card will result in a generic, non-optimized driver loading. You still get basic performance. High Speed SCSI tape backup drivers are non-functional in Vista, scanners of all sorts and shapes. Why can WinXP (both 32 and 64 bit) handle legacy drivers, but Vista doesn't? We could move on to the area of factory automation and DAQ cards, or any other number of very expensive, very customized PC hardware. This is a basic software Quality issue.
All MSFT has to do, is inform the customer that there is limited legacy support. Forcing the consumer, as MSFT attempted to do, to abandon perfectly good equipment, so they can have a new OS, is negligent.
If you updated your Linux distro, and suddenly found your entire RAID array corrupted and unmountable, your tape backups and your network going down and not coming back up - I doubt you'd have very nice things to say. Now imagine not having the ability to retreive or obtain the previous Linus Distro that had worked fine for 10 years. That's my point.
Be careful on how you interpret 'prefetch'. Basic question, what runs the PC? Software, or the hardware? The hardware does the pre-fetching, the software sits at a much higher level, well above the embedded level that is iniated by BIOS to drive the South Bridge. A pre-fetch in software does nothing more than push a pointer to a memory stack, that is ALREADY sitting in Hardware (the pagefile, RAM, cache or in the South Bridge). The Software (OS or not) does not have memory, it does it's best to manage memory found in the hardware - but all it can do is direct the hardware. WinXP has a pre-fetch, and most benchmarks have found that a maximum of 3 Quadwords was all that was required for optimal performance. Now a pipeline controller will obviously flush the pipeline in a unconditional branch or JMP - but if you think that a software "Prefetch" is magical, you are mistaken. That's why we have multi-level L2 and L3 cache. All the software does is write notes to itself, as to where the hardware stashed the code. The hardware then manages which memory locations are in the Read, Modify or Write phases, whether the data is stale, or if a core is presently preparing to update a memory area (modify flag). Don't believe me? Dive into your BIOS and disable the pre-fecth in your South Bridge, and crank the OS prefetch. Without the SB pre-fetch turned on, OS prefetch or not - you will take a significant performance hit.
I have yet to see any benchmarks that show the performance bump the Super Pre-fetch has over the standard prefetch, compared to the super-duper-double-fudge brownie pre-fetch. I question their 'real world' benefit. If your disk optimation program places frequently used applications, recent documents and the OS in high speed areas of the hard drive, the hardware is able to grab it a wee bit faster. Will it make your jaw drop? Probably not. Will you see the difference in a benchmark? Maybe. Will you 'feel' it in everyday use? I doubt it.
Superfetch has to do with the loading of information into main ram before it is needed. You should read up on what it is before you post again as you aren’t helping your credibility at all this way.
Dell has a vista 32 driver.
Due to superfetch apps like word, excel, VS2k8 start in much less time than they do without it. See - I actually tested.