Re: An unknown language with a known implementation ?

Joachim Durchholz <jo@durchholz.org>
Mon, 22 Oct 2007 07:32:13 +0200

          From comp.compilers

Related articles
An unknown language with a known implementation ? marten.cassel@gmail.com (2007-10-13)
Re: An unknown language with a known implementation ? mailbox@dmitry-kazakov.de (Dmitry A. Kazakov) (2007-10-14)
Re: An unknown language with a known implementation ? DrDiettrich1@aol.com (Hans-Peter Diettrich) (2007-10-14)
Re: An unknown language with a known implementation ? tommy.thorn@gmail.com (Tommy Thorn) (2007-10-19)
Re: An unknown language with a known implementation ? jo@durchholz.org (Joachim Durchholz) (2007-10-22)
| List of all articles for this month |

From: Joachim Durchholz <jo@durchholz.org>
Newsgroups: comp.compilers
Date: Mon, 22 Oct 2007 07:32:13 +0200
Organization: 1&1 Internet AG
References: 07-10-043 07-10-062
Keywords: code, practice
Posted-Date: 22 Oct 2007 12:20:20 EDT

Tommy Thorn schrieb:
> x86 is a virtual machine (sure why not?). You can write an interpreter
> for that (hard but it's possible). All interesting programming
> languages have an x86 implementation, but an x86 interpreter will tell
> you nothing about the languages that were compiled to it.


Looking at the object code will still give you a lot of information.
If the interpreter was designed specifically for that language, it
will also be helpful.


I think the most important question is whether it's worth the
effort. If I were charged with that task, I'd spend a day or two
looking for and at debugging symbols in the binaries and google for
the terms I found; I might also try to find strings that the programs
output to the screen and look how they were embedded in the binaries,
trying to find out something about the structure of the bytecode (or
whatever they are using). And I'd take a look at the run-time library
- most contain a copyright notice, which can narrow down the field
considerably through date and company name.


If that turned up anything useful, I'd continue, otherwise I'd declare
that analyzing this thing could easily take several man-years, and
shouldn't we rewrite the thing from scratch?


Regards,
Jo


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.