reading a paper about vmware's attempt to develop the first x86 virtualization engine and i stumbled across this glowing review of windows 95
...and what business does any modern operating system have making BIOS calls
idk how a hypervisor deals with that except by emulating the BIOS functions (but luckily i don't have to understand really)
hypervisor: i am here to look over your machine instructions and make sure you don't do anything stupid
windows 95: ok. i am calling INT 13h now
hypervisor:
@saladeredux
Windows 95 was created for IBM PCs. That was Microsoft's original captive market, right from the beginning.
They had no intentions or need to support other processors because their market was the home computers IBM was selling - and these used the 80x86 architecture.
In other words, IBM and Microsoft worked in a symbiotic relationship; they had no incentive to break backward-compatibility and Microsoft would gain no money by supporting additional architectures.
@saladeredux i only work indirectly with folks who handle windows and virtualization and ... i don't know if they even *know* how cursed their stuff is relative to everything else
@saladeredux Sounds like an interesting paper ... can I have a citation?
@elb of course! here is a link: https://dl.acm.org/doi/pdf/10.1145/2382553.2382554
@saladeredux For posterity's sake, along with a public PDF (not behind the ACM paywall):
@elb ah my bad! forgot about that
@saladeredux No problem! I actually have access to the ACM DL, but I thought others who followed this thread might not. :-)
i have absolutely no trouble believing that the way windows 95 addresses memory is the stuff of nightmares. i see the word "segmentation" and i'm ready to close the tab