Originally posted by thibs Geez using old lenses on mirrorless (whatever the brand) is like a geek's computer.
It's very nice and fancy but it is the exception. On those users are satisfied the body doesn't sell (as any body from any brand) and you don't sell lenses 'cos your fracking (sorry, BattleStar Gallactica as of late) users use... old lenses.
Quick money is on the bodies.
Real money is on the lenses. Something Sony still did not get.
Not sure what your point is. My point was if your claim to fame is all your legacy glass, a decent percentage of which is MF, other people run that glass as well or better than you. Sure, the only people that matters to are camera geeks, therefore they're the only relevant audience for that topic.
If you're talking about what drives profit, then sure, lens sales do, but that's really a separate topic. And on that topic, legacy glass is a problem, because it means people can buy lenses from someone other than you. And the lens compatibility of MILCs breaks mount lock-in and makes taking a flier on buying into a new system easier. Would I have bought into another mount if I had to re-buy all of my glass to have an effective system? Probably not. I did the math and it would have been a $5-7K move for me building a competent system all at once. As I was, I could buy a body and a prime, and leverage my existing Pentax glass, which made it a $1.5k trial (less than that because I could have recouped some on trade if I hadn't liked it). Much more palatable. And then they sold me 2 more bodies and 3 more lenses, because when I'm lazy I want AF. Yep, they work the same why as everyone else.
I personally know a decent number of other photographers that have moved into Fuji, Sony, or Oly systems via that same route. No sure how much of an exception is is.
---------- Post added 09-22-2014 at 02:16 PM ----------
Originally posted by monochrome You are aware that my old and still good Monte Carlo backtest simulator running on a new HP box works just fine in Win7 (in a virtual XP machine running in split clock cycles under the Win7 Host) as long as I don't need to transfer output out of the virtual machine into a Host application aren't you?
The problem is you're running an abstraction layer rather than a direct physical interface. Even so, depending on the VM host you should be able to redirect an interface in the guest and at least spool the data out.