Re: Two Click disassembly/reassembly

On 2 Feb 2006 16:32:28 -0800
"randyhyde@xxxxxxxxxxxxx" <randyhyde@xxxxxxxxxxxxx> wrote:

:Inferring that a
:program can be written to do the same thing as a talented team of
:people is a mistake on your part.

Or at least it would be, if I had ever made such an inference, and an even
bigger mistake if I deliberately implied such a thing. But in fact, you
missed the point. My umbrage was taken at your claim that having written
several compilers and assemblers, and having done some unspecified
research, made you an expert on porting computer systems from one platform
to another.

To begin with, is there anyone with a CS degree who HASN'T done what you
did? For my senior project [in EE, not CS], I designed a (trivial)
computer, invented a (trivial) language, and wrote a (trivial) compiler.
Big deal, NOT!

In addition, writing an assembler brings nothing to the table, as this can
be done without any ability to program the target architecture. Writing
the code generation portion of a compiler is a bit more relevant, but not
as valuable as actually writing assembly language programs for that
architecture.

:I, too, have been involved with lots of porting
:projects over the years. I have my share of "war stories".

If you wish to share them, describing what went right, what went wrong,
and what lessons you learned from them, then we may have a basis for
rational discussion, perhaps even as equals.

:Now, I'd be willing to bet *big* money that your ports to the IBM were
:*not* automatic.

Some portions were. Some portions used pre-existing tools. Some news tools
were written. Some entire application suites were replaced by existing IBM
offerings. Some applications were converted. Others were replaced by new
applications, written from scratch. In short, I have been in so many "The
difficult we do immediately, the impossible takes a little longer."
situations, that I have an almost intuitive feel for what can, and cannot,
be done.

Which brings us back to what I actually claimed:

1. Given a sufficient quantity of existing x86 code to make it worthwhile,
I (at least) could write a tool such that it would take me less time to
write the tool and port the code, than it would take me to port the code
without the tool.

2. That I postulated the tool as processing the source line by line, and
emitting one or more [e.g. PPC] instructions per x86 instruction, or
flagging the line as needing attention from a human programmer.

3. That MicroAPL already has products which go beyond what I postulated,
in that, like a compiler, they optimize the output code, based on
information derived from multiple lines of source code.

It was, and still is, my impression, that you denied all three of these
assertions. Would you like to take this opportunity to clarify my
understanding?

-- Chuck

.

Relevant Pages

  • Re: How printf() works???????
    ... Getting something totally backwards is,of course, a mistake. ... is no evil intent in my correction. ... What he said to you about your compiler is beside the point and did not ... Therefore Robbie Hatley took your statement as correcting a ...
    (comp.lang.c)
  • Re: #define and (brackets)
    ... If I make a mistake, ... if the compiler knows enough to tell me ... due to some implicit type conversion say, then I define that conversion as ... That's automatic bug detection at compile time. ...
    (microsoft.public.vc.language)
  • Re: How printf() works???????
    ... Getting something totally backwards is,of course, a mistake. ... is no evil intent in my correction. ... What he said to you about your compiler is beside the point and did ... "Hey Robbie, you are an idiot to have said that my compiler is not good ...
    (comp.lang.c)
  • Re: How printf() works???????
    ... Getting something totally backwards is,of course, a mistake. ... What he said to you about your compiler is beside the point ... RHs whole reply was based on explaining the OPs code. ... fuss over *my* reply instead of answering directly to Robbie Hatley, ...
    (comp.lang.c)
  • Re: XV 3.10a build on Alpha comments
    ... This organization isn't always feasible and/or some in-line conditionalization from a previous port may already be in place, but keeping the platform-specific stuff separate from the base code for the package can make on-going maintenance rather easier. ... If you use the function name that is different than the real function, the compiler and linker will only map it into the namespace when it's needed and that's when the define and conditional sequences select the code. ... there has been more work spent on removing existing workarounds than work on the port itself. ...
    (comp.os.vms)