How do you choose which categories matter most? That’s subjective.
User satisfaction and user serveys are subjective.
Also the stability of features often depends on where along the life of Windows you measure. For example XP post-SP3 is a very different OS to XP pre-SP1. So how do you decide where to measure? That’s also subjective. Not to mention the subjective question about what “stable” means and how you compare stability
There’s been enough arguments on HN regarding that last point alone.
For example Windows 2000 has crashed fewer times for me than Windows 7. Does this mean that Windows 2000 is now “objectively” the better OS?
You might argue that the sample size is too small, but who decides the sample size? Are you including or excluding specific datapoints? Of course you are, which means “better” is then defined by the sampled data (for example the power-user vs novice ratio for windows 2000 will be different to Me and 7). And thus it’s subjective again because who decides which datapoints are important and why?
Are we talking about “better” for web browsing? Home server tasks? Software development? Video editing? Etc. who decides that and why?
What about “better” for memory or CPU footprint? That matters more to other people.
We might think that because we are making measurements that this isn’t subjective but those measurements are still a matter of personal judgement.
Edit: as an aside, this is precisely why I never use absolutes in a professional capacity. It’s easy to “say X is better than Y!” but it demonstrates far more professional experience to say “it depends. Tell me what your requirements are.”
I find it fitting here to reassert the last point from my previous reply: Your reasoning here implies Windows versions, or any operating systems can not be ranked objectively, which suggests an objective criteria can not be established. I disagree.
The approach to accomplishing that can be a discussion, but not one relevant to the point I made above - it need only be acknowledged that it is possible.
> I’m not saying operating systems cannot be ranked objectively. I’m saying the GPs statement was far far far too broad to be objective.
> If they said “highest rated for performance X according to Y” (where X and Y are defined measurements), then it would be an objective statement.
So you are saying OSes can only be compared objectively on specific features, and not in an overall capacity?
> However just saying something is “objectively the best” doesn’t make it objective.
True, which is why I said I believe a case could be made. This being a HN comment I'm not taking too seriously, I'm not going to put in the work to do so, I'm more interested just to see casual discussion and see people who might agree or disagree, without worrying too much about the semantics of the claim.
> So you are saying OSes can only be compared objectively on specific features, and not in an overall capacity?
Depends. “Better” is a subjective term so if you were to discuss overall capacity objectively then you would need to explain why you arrived at that conclusion. Otherwise it’s left to interpretation (ie subjective).
For example: I could make several well reasoned arguments why I consider Windows 2000 to be better. Others could do the same for XP. We are all correct for our own interpretation of “best”.
> “Better” is a subjective term so if you were to discuss overall capacity objectively then you would need to explain why you arrived at that conclusion.
Well, like I said, I need not describe the methodology that would result in objective measurement - it need only be possible to measure things objectively to the best of our ability. And it is.
User satisfaction and user serveys are subjective.
Also the stability of features often depends on where along the life of Windows you measure. For example XP post-SP3 is a very different OS to XP pre-SP1. So how do you decide where to measure? That’s also subjective. Not to mention the subjective question about what “stable” means and how you compare stability
There’s been enough arguments on HN regarding that last point alone.
For example Windows 2000 has crashed fewer times for me than Windows 7. Does this mean that Windows 2000 is now “objectively” the better OS?
You might argue that the sample size is too small, but who decides the sample size? Are you including or excluding specific datapoints? Of course you are, which means “better” is then defined by the sampled data (for example the power-user vs novice ratio for windows 2000 will be different to Me and 7). And thus it’s subjective again because who decides which datapoints are important and why?
Are we talking about “better” for web browsing? Home server tasks? Software development? Video editing? Etc. who decides that and why?
What about “better” for memory or CPU footprint? That matters more to other people.
We might think that because we are making measurements that this isn’t subjective but those measurements are still a matter of personal judgement.
Edit: as an aside, this is precisely why I never use absolutes in a professional capacity. It’s easy to “say X is better than Y!” but it demonstrates far more professional experience to say “it depends. Tell me what your requirements are.”