Doug's Mailbag: Microsoft-SAP, Windows 7 Speed
Recent rumors have Microsoft trying to buy up SAP, but a couple of readers see some problems in that match-up:
About Microsoft and SAP -- what a culture clash! Marketing into the big enterprise is sooooo different from Microsoft's current sales channel. Now, at the enterprise level, the C-level staff would probably feel better with SAP buying Microsoft.
Can't they do better? SAP is horrid!
Doug recently mentioned a recent PC World review of Windows 7 that put the new OS on the same level as Vista, at least in terms of speed. Redmond readers, on the other hand, pegged Windows 7 as faster. Readers weigh in:
As a matter of fact, with Windows 7 on one machine only (a Fujitsu U810), the speed of the operating system has improved, but Office 2007 performance has decreased under Windows 7. Some tests that PC World ran appear to test applications rather than Windows tasks, so perhaps the results are accurate but poorly described.
Nevertheless, this is an important issue for Microsoft. The vast majority of home users have no idea where Windows ends and Office starts. While fast boot times are nice to have, an issue with Outlook or Word will result in users claiming that "Windows is too slow." I would guess that 90 percent of the complaints I heard about Vista were Office 2007-related. Thus, PC World's results matter, and should be taken more seriously. Why? Because people like my father, a Windows user since 3.11, are now running Office on Apple.
Interviews with 50 users and tests on three computers are not comparable. One is subjective, impressionistic, qualitative "research" and the other is quantitative measurement. Any researcher can attest to the built-in biases and inherent unreliability of interviews and surveys.
Windows 7 may be faster than Vista or slower, and it no doubt depends heavily on who is running what, on what and at what time. Test suites and test protocols can certainly be contested and usually are. But in the end it is a quantitative question that can only be settled by direct measurement, not by interviews, even with thousands of respondents.
I just saw your item on the speed of Windows 7 in the test being equivalent to Vista. I can't really comment on that as I don't use Vista. I can comment on one aspect, though. I run Virtual PC with separate XP instances for testing and such and created an instance of Windows 7. In my real-world experience (just use of various applications including custom-built), Windows 7 appears to be faster than XP in a virtual environment. I've really been impressed with how quickly it boots and how fast programs load. It might be worth having someone run the numbers on in a true test.
In my opinion, Windows 7 is snappier, not faster. The input lag is reduced and it feels faster. Perception is reality when it comes to these things.
I think the problem is responsiveness vs. speed. Vista is plenty fast, particularly after SP1. On performance tests it equals or betters XP nearly across the board. However, it just feels slow because the UI is unresponsive quite often. I have not yet tried Windows 7 but my hunch is that it is no faster but much more responsive. Users typically report responsiveness as speed/performance when in fact the two are not the same and can diverge quite a bit.
As is often the case, the slow reportees probably report less enthusiastically. There is a lot of milk-age behind the "Vista is hopeless" lobby (I speak as an XP user).
Or maybe their installation is so slow that it is retarding their e-mails?
And Doug recently expressed his concern about rumors that Windows 7 will be released in October, saying that OSes aren't like cheese, which have a known "ready" date. James thinks that's beside the point:
You need to clue in the rumor mill about the difference between cheese and software. Although it is true that cheese cannot and should not be released before its time. Microsoft doesn't make cheese, so it doesn't matter if the software is ready or not; it can be, and usually is, released before it is ready. They just figure, "Hey, if it ain't workin' now, we'll just send out a patch/upgrade or SP later to fix it." If we had to wait until software was truly ready, anyone running MS would still be back around 3.1. It ain't about "ready" -- it is more about "ready enough" and how MS can maximize their bottom line (or as the like to say around here, the ROI).
Tell us what you think! Share your thoughts below or send an e-mail to firstname.lastname@example.org.
Posted by Doug Barney on 05/20/2009 at 1:16 PM