Re: BDS2006 exe sizes

Dunny wrote:
J French wrote:
On Wed, 22 Feb 2006 22:31:02 -0600, Rob Kennedy <me3@xxxxxxxxxxx>
wrote:
Ian wrote:
Can someone tell me what the size of a simple "hello world"
program's .exe file size is in BDS2006? I'm curious to see how much
different they are from D7.
Why is a "hello world" program a meaningful benchmark?
It gives one an idea of 'compiled in' run time support
- probably the first test one does is a console mode 'Hullo World'
- then a MessageBox 'Hullo World'
- then a ShowMessage 'Hullo World'
- then a Form showing 'Hullo World'

Not a bad test of compiler (or rather linker) efficiency

Right. Not compiler efficiency at all, since you're not giving the compiler any significant amount of code to process. All the work is being done by the linker.

Exactly. Why did none of the others who responded to this figure that one out?

I asked my question because I can't read minds. Can you?

The guy never asked for Hello World to be a good test of the compiler -

In fact, that's *exactly* what he said it would be:

"As soon as I get my hands on a new compiler, the first thing I do is test it out using a simple "Hello World" ... If the compiler produces a large executable from such a simple program as those, then they aren't efficient enough in my book."

A "hello world" program gives the compiler only *one* line to work with. There's not much a compiler can do with that.

he wanted to know what amount of bloat he could expect from each delphi product.

Ian never said anything about bloat.

I'd have thought it was obvious that the greater the revision number of Delphi, the greater the file size.

Why would that be obvious?

--
Rob
.

Relevant Pages

  • Re: BDS2006 exe sizes
    ... program's .exe file size is in BDS2006? ... - probably the first test one does is a console mode 'Hullo World' ... The guy never asked for Hello World to be a good test of the compiler - ... Delphi, ...
    (comp.lang.pascal.delphi.misc)
  • Re: BDS2006 exe sizes
    ... .exe file size is in BDS2006? ... - probably the first test one does is a console mode 'Hullo World' ... Not a bad test of compiler efficiency ...
    (comp.lang.pascal.delphi.misc)
  • Re: Efficiency and the proposed changes
    ... that the efficiency loss is only paid by people that use them. ... Operator overloading is a change that will affect only people ... No efficiency lost since the compiler generates ... If you examine the raw size of the C standard, ...
    (comp.lang.c)
  • Re: OO primitives
    ... Sometimes that means that time efficiency matters far more than space efficiency. ... So we simply instantiate hardware to speed-up those specific operations, and we're instantly running faster than even the best optimizing compiler could ever generate code for. ... It is slower than the standard version and has been criticized for that slowness. ...
    (comp.lang.forth)
  • Re: VB6 Compiler
    ... Yes, with the combination of useful "internals" documentation, and all ... sections, readable and writeable sections, executable code sections, ... way the compiler creates the object code... ... I've also noticed that the EXE file size ...
    (comp.lang.basic.visual.misc)