[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: more evil firmwares found

Don Armstrong <don@donarmstrong.com> schrieb/wrote:
> Secondly, if the original assembly contains any comments at all, then
> the disassembled binary is no longer the prefered form for
> modification.

How do we know there were comments? How do we know the comments were
useful? If the author does not tell us there were no comments, we can't

On the other hand, it is possible to find out how difficult it is to
make sense of the code without comments. Of course, the more complex the
code the more necessary the comments and the less can a commentless
version be considered the source code[1].

> What the GPL asks for is not that controversial. You need to supply
> the prefered form for modification, which is (generally) the exact
> same form that you used to create the work in the first place.

No, there may be different forms of the work that are equally suited for
modification. It's not that obvious with programs written in a high-
level language (such as C) but that's only due to the lack of good

Take images, for example. Is the preferred format PNG, TIFF, XCF, or
even PSD?

With binary firmware, some people may prefer the binary form and a hex
editor to make modifications. Other people may prefer Assembler. So
what? As long as the disassembler works[3] and the code is simple enough
not to need comments, that's not a problem. Both formats can be
considered the "source code".

Of course, we _have_ to find out whether these two conditions are met.
The size of the blob can be a first indication.

[1] And that's even true if the original author did not write any
    comments at all; i.e. the comments only exist in her/his head.
    This *is* a problem as soon as the code is combined with other
    GPL'd code.
[2] If I convert a C program to PASCAL -- and GCC and GPC produce the
    same code, which one is the source code?
[3] Note that the availibility of a free toolchain is not a condition.

Reply to: