Re: MASM Expert needed immediately


Betov wrote:
"randyhyde@xxxxxxxxxxxxx" <randyhyde@xxxxxxxxxxxxx> écrivait
news:1148997669.468724.68820@xxxxxxxxxxxxxxxxxxxxxxxxxxxx:

I took a disassembly of RosAsm (done with IDA Pro 4.3) and ran it
through MASM.

Great, ass-hole. Just show us the result. This is quite
easy: You upload this somewhere, and you provide a link.
That way, we will learn how to rebuild a PE, with IDA.

Look back via Google in this very newsgroup.

MASM compiled the output *much* faster than did RosAsm.
Now to be fair, this was *before* the last performance update for
RosAsm, but based on that benchmark, I'd suggest that MASM and RosAsm
are roughly equivalent, maybe with RosAsm having a slight edge, today.

One of the very typical swindling of our prefered ass-hole.

Yep. Anything that contradicts whatever you say is a "swindling", eh?
So go ahead, post your *own* benchmark numbers, as people are asking.
And if you haven't figured it out yet, allow me to give you a clue --
people don't consider FASM compiling FRESH to be comparable to RosAsm
self-compiling itself. They want to see both assemblers compiling the
*same* code. Compiling two different programs is not a benchmark, it's
a swindling.



In case somebody could be tempted by giving any credit to
such a joke, it is useful to explain that flat naked Sources
are naturaly way faster to compile, given the suppression
of all of the Equates, Macros, and such HLL-able features.

Uh, remember the part about running RosAsm through it's own
disassembler?
Surely MASM processing IDA Pro disassembled code can be considered
equivalent to RosAsm processing disassembled code, eh?


The Macros alone represent usually half of a Compile time.

There were no macros in the disassembled code.



OTOH, on the basis of other benchmarks I've run, FASM is about 2.5x
faster than RosAsm. And this *is* comparing the latest version of
RosAsm against FASM.

You never had any competency for any benchmarking,

Spoken by the guy who claims that FASM compiling FRESH is comparable to
RosAsm self-compiling itself. As you said, thanks for the great laugh.


the first
one of which is intellectual honnesty. As you never had any,
the lurkers will keep happy with doing self-compliations of
FASM-Fresh and of RosAsm, and with reading the Stats, that
say all there is to say.

And you have the nerve to claim others are incompetent at benchmarking
:-)




Of course, expect Rene to start hollering about how incompetent I am
for creating benchmarks. After all, they disprove his claims.

No, ass-hole, they confirm them.

Well, we'll just let your comments about FASM/FRESH and RosAsm/RosAsm
speak for themselves and we'll let everyone out there judge who is the
incompetent one at benchmarking :-)




And when
you ask *him* for proof of his claims, he'll compare RosAsm compiling
itself against FASM compiling a completely different program. Yeah,
right; that's a competent benchmark.

Yes, ass-hole. FASM-Fresh and RosAsm are very similar sources,

No, they are not.

doing similar things,

FRESH is an IDE; that's about all it has in common with RosAsm.


similar ways.

Oh how you insult John Fine.

Now, as long as having under
the hand, two identical Sources is something that does not exist,
there is no way but comparing what can be compared.

Disassemble RosAsm with your disassembler. Disassemble Rosasm with IDA
Pro. Now you've got two source files that do pretty much the same
thing, have the same mix of instructions (sans disassembly errors, but
they can generally ignored in a program this large). Compile the two.
That's a relatively fair comparision. True, it doesn't test macro
processing and a few other details, but it's as close as you're going
to get.



Though, it is partialy true that FASM-Fresh Sources and RosAsm
Source are somewhat different, when talking of Speed comparison:

They are completely different. You need the same instruction mix, the
same program organization, the same use of features. FRESH and RosAsm
have almost nothing in common when you make such a comparision.


That is that RosAsm Source are evidently more HLL-Style than
FASM-Fresh ones. So the comparison is not completely fair...
at the advantage of... FASM, which does not need to spend as
much efforts on the Macros resolving, as RosAsm does.

We've been through this before and the FRESH crew *has* pointed out the
use of macros in FRESH. But that's irrelevant, because you're still
comparing apples and oranges. The programs have to be *identical*, or
pretty damn close, for the benchmark to have any meaning whatsoever at
all.



:)

And, as for comparing with MASM... this does not deserve any
answer.

Of course not. You're embarassed by the fact that MASM almost matches
the performance of your product after years of claiming that MASM is
10x, 20x, whatever times slower than RosAsm.
Cheers,
Randy Hyde

.

Relevant Pages

  • Re: Benchmarking RosAsm vs. FASM
    ... comparing the compile time of RosAsm compiling RosAsm against ... > FASM compiling FRESH *is* of value. ... FASM just did it faster than RosAsm. ...
    (alt.lang.asm)
  • Re: MASM Expert needed immediately
    ... RosAsm, but based on that benchmark, I'd suggest that MASM and RosAsm ... One of the very typical swindling of our prefered ass-hole. ... there is no way but comparing what can be compared. ...
    (alt.lang.asm)
  • Re: ///RosAsm - exports issue with some DLLs
    ... Oh hell, I'm kinda getting bored and this is *kinda* getting ... have done for RosASM. ... Comparing this efficient simplicity with the hell of the ... NASM C sleeping devs is nothing but laughable. ...
    (alt.lang.asm)
  • Re: HLA v2.x and / or LASM suggestion: Win32 Resources
    ... > I took the EQUATES.EQU file from the RosASM distribution, ... then assembled it using MASM. ... compiling dummy Files do not mean a thing, and that when, ... and that, when comparing to ...
    (alt.lang.asm)
  • Re: The Randall Hyde lies go on
    ... And what you find is that RosAsm crashes when you do this. ... differing only in the syntax required by each assembler. ... Comparing the assembly of RosAsm ... Compiling RosAsm under FASM, or compiling FRESH/FASM under RosAsm would ...
    (alt.lang.asm)