RFP: UPX (Was Re: reducing binary distribution size with UPX)

[Christopher Faylor]

On Fri, Mar 15, 2002 at 04:42:27PM +0300, egor duda wrote:
>Hi!
>
>Friday, 15 March, 2002 Robert Collins robert.collins@itdomain.com.au wrote:
>
>RC> I vote for including UPX... and Lapo makes two. Do we need a third? And
>RC> are there any objections?
>
>Does anybody ever tried to measure if upx impose any performance
>penalties? If i understand things correctly, upx compress executable
>file and attach a small "decompressor" stub to it. Then, when
>executable starts, this stub decompresses original executable image.
>This will totally defeat the features that most modern OSes have,
>mapping pages from executable and loading them on demand, sharing
>common read-only pages between different instances of one application,
>etc. I really don't understand what's the point in saving disk storage
>worth several cents (1byte == $1e-7), while increasing memory
>footprint and reducing speed. Hey, just read upx docs, they contain
>all these points already.
>
>Not that i'm against inclusion of upx to cygwin distro -- it's a
>normal package like many others after all, but i really don't
>understand why somebody would want to use such a program.

Excellent points.  This is, IMO, an argument against using upx for
all (any?) cygwin binaries.

cgf