Windows 7 or Windows xp?

[Anonymous]

posted on January 28th, 2011, 4:10 pm

Last edited by Anonymous on January 28th, 2011, 4:14 pm, edited 1 time in total.

I thought you were uneducated in science because you don't seem to understand the idea of keeping a variable constant to eliminate the influence of its effects on changes in data. Introducing a second variable (and keeping certain variations of it paired with certain variations of the other variable) just because it's like that in the real world will just screw up the data.

ive studied statistics from 4 separate perspectives, so i understand reducing the effect of other variables, so your assumption was rather bad.

what often happens if u make a test too clinical is that you generate lovely accurate numbers that u can feel happy arent biased hugely by natural variation, but are completely and utterly useless because u have gone away from what is real and have tested something that is far too restricted and sterile. inferring things from statistics is an extrapolation, so if your data is far from real life, your extrapolation isnt strong.

avoiding the random chance/variation present in real life is also a challenge, thats why statistics can sometimes be misleading. a lot of statistics use incredibly low significance levels and end up drawing conclusions when the data arent that conclusive.

the industry doesn't need or want such a thing.

i think that is the key here. these "statistics" arent wanted or needed.

many of these arent really statistics they are just raw data presented by someone who has the book Excel for dummies.

ive seen so many examples of teenage geeks with a computer and a GCSE in maths that had a stats component, they get so far ahead of themselves and go mess around in excel or get a warez copy of spss and start trying to draw conclusions. they have all the answers to questions that nobody asked, nor wanted the answers for.

Whether or not 7 or XP will ever be on a particular system in the real world is completely irrelevant.

its completely relevant. whats the point of gathering data that wont allow you to infer information about the population? if you do so, u have wasted your time and money on something that wont help anyone ever. the point of statistics is to infer something about real life. thats the key. if u cant turn data into inferrence you have nothing, just a bunch of meaningless numbers that wont help anyone do anything they want to.

a comparison between xp and 7 on the same machine is silly because they are designed for different machines. comparing them on their own machines, what they were both designed for, introduces more variability, but with a high significance level you can gain some useful insights. u wont get anything conclusive on this thats for sure.

tests of this kind in general arent helpful since users dont sit at their computers and run benchmark programs over and over again, they all do different things. some browse the web and watch porn, some pirate dvds to sell in tesco car parks, some use it to talk to family on skype, some use it to write asinine business efficiency marketing jargon reports, some use it design chairs in cadcam, some use it to play games, some use it to run a website, some use it to troll people, some just use it to look at pictures of cats on ichc, people use computers differently and its impossible to extrapolate to their uses effectively.

i leave you with a picture of a lolcat, i wrote a huge long post, i deserve a spammy image dont i? meh ill do it anyway

[img width=300]http://icanhascheezburger.files.wordpress.com/2011/01/c56bccce-8fcc-42fe-8bea-f0651568e07d.jpg[/img]

---

EDIT: changed image dimensions to seem less spammy.

posted on January 28th, 2011, 5:35 pm

Basically it comes to this. Windows XP and windows 7 32bit will run on roughly the same systems. XP service pack 3 will continue to be supported until 2014.

-XP has several fundamental security flaws, but it is currently written for and supported by the vast majority of software writers as its APIs are more broadly understood and more widely implemented. basically it's an XP world, you're just living in it. XP can also address storage and boot drives of up to 2 terrabytes completely unaided in NTFS, so mondo storage is still not a problem.

-W7 is a much more secure platform from the get-go, taking a closed security approach as opposed to the open approach taken by XP (XP started in an era when remote control attacks couldn't be executed, primarily for the fact that most users had dialup making it unfeasible on a large scale) of course, this means a more intensive user experience, with UAC for better or for worse. Going forward, however, 64-bit w7 is more compoatible with more than 3.3 gigs of ram, drives over 2 tb, and modern interactive bios setups.

but, unless you NEED greater than 3.3 gigs of system memory total (that's mobo and video combined) or you execute a lot of 64-bit only applications, currently the 32/64 is moot, and in fact, much software only comes in 32 bit flavours to this day.

in theory, if you have two systems, 32 bits, with under 4 gigs of memory, you COULD choose either one. In practice however, what CPU you're operating makes a difference. Any single core processor, hyper-threaded or otherwise, will operate either OS perfectly well, and with no noticeable difference in performance. However, if your machine has anything that is multi-cored; that is anything newer than an Athlon X2, a Pentium D, or anything with Core in the name, (or the rare pentium 3 or 4 or xenon dual processor system) then you're better off running Win7, which provides several benefits on multi-processor systems. Basically, in side-by-side operation, a Core Duo at 1.6 ghz, 2 gigs of ram will perform noticeably better in windows 7 than in XP.

Now to Linux... Linux is always 2 years or so ahead of innovations seen in Windows. Linux has had support for advanced bios, drives over 2 terrabytes, multi-core procs, and other innovations for some time that are just starting to appear in the windows world. Of course, the problem is cross-platform support. Just as with macintosh, running many popular programs can be hit and miss. while some popular systems such as Office suites and browsers are available in both windows and linux flavours, others are relegated only to one OS or the other. While there are emulators such as VMware, and Wine, which can enable cross-platform program running is still hit and miss. Wine support improves everyday, but it still not perfect, and VMware and similar programs are expensive, with free versions having crippled specs such as no optical drive access. (of course VMware pretty much requires multi-cored processors to work anyway)

Macintosh? it's a non-starter. It's visually impressive, and simple to operate, but it lacks broad software support, lagging behind both windows and linux in its walled garden where it comes to running anything other than an apple-device or browsers. And honestly, where it comes to the machines, you're paying cadillac prices for chevrolet hardware. Of course, owing to this, you can operate windows on mac hardware with no problems, and mac's on "non macintosh" windows hardware with some hoop-jumping.

---

Anyway, my two cents and personal experience.