waldo kitty wrote:
i also used this technique to store registration data and options
settings directly in the executable instead of having a separate and
external configuration file...
i can post those old sources if anyone is interested... they only need
ask... the sources are dated May 1993 and i likely saved them from the
Fidonet PASCAL echo (message area)... they might have even ended up in
the SWAG archives but i don't know ;) -=B-)
I did this sort of thing back in the days of MZ files, where there was a
simple length field in the header. I believe that NE files similarly had
an accessible overall length, but they also had a checksum field even
though this was rarely (if ever) used. Later Windows formats (PE etc.)
might use signing/blessing/branding/checksumming to varying extents, and
I also think there was discussion of unix-style signing in this (or a
related) list a few weeks ago.
Which leaves me thinking that the safest way of doing it would be to
look for an "official" program, i.e. from MS or from unix binutils,
which- if a program was already signed- might request some sort of key
before it would change anything.
Of course, an even safer way would be to leave the executable alone and
to put an early check in the startup code that a subsidiary key file
existed, and for that key to include something that identified the
machine or site on which the program was entitled to run.
--
Mark Morgan Lloyd
markMLl .AT. telemetry.co .DOT. uk
[Opinions above are the author's, not those of his employers or colleagues]
_______________________________________________
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
