Hacker News new | ask | show | jobs
by jstepien 5228 days ago
I'm intrigued by the 0.8 GB difference between the x64 and the x86 build. It's even more interesting given that this difference must be caused solely by different machine code, which, by its nature, shouldn't pose a big challenge for compression algorithms. Assuming that 32 extra bits per pointer and a different instruction set aren't enough to generate such such an increase in size, I'm wondering what might be the reason. Extra binaries and portability-related code for for 32-bit compatibility? Any ideas?
2 comments

64-bit Windows contains both 64-bit and 32-bit versions of all the userland libraries -- things like user32.dll and comctl32.dll -- as well as both the 64-bit and 32-bit versions of .NET.
Extra binaries for 32bit compatibility probably take most of the space. On my 64bit win 7 install, C:\windows\SYSWOW64 takes up 1.4 gigs.
> On my 64bit win 7 install, C:\windows\SYSWOW64 takes up 1.4 gigs.

For the uninformed: on a 64bit Windows, SysWOW64 (i.e Windows On Windows 64) holds 32bit binaries while System32 holds 64 bit binaries (whoever pulled that April's fool prank must have laughed out himself to death by now)