Consider Thompson's seminal paper on trusting trust. It is possible to conceive of embedding a backdoor in the software that translates your high-level UART design into layed out gates, P and N junctions, connections, etc., with the addition of a hardware backdoor that activates on a sequence of bits traversing the circuit. With a circuit of high complexity it might plausibly go unnoticed. Possibly it would be within the capabilities of nation states. Easier to envision for certain classes of hardware, but certainly a very hard problem. If you have the skills to write a recognizer for a hardware UART, that can add a backdoor, could you please also also write an agent that could Do What I Mean as I code my programs?:)
The Computer From Pascal to von Neumann
Herman H. Goldstine
ISBN 0-691-08104-2 (Hard cover)
ISBN 0-691-02367-0 (Paperback)
This book covers the WWII era development in detail. Well written. I think the author is the Naval officer responsible for getting the Eniac funded when all the other experts said it wasn't worth developing.
Eniac - The triumphs and Tragedies of the Worlds First Computer
Scott McCartney
ISBN 0-8027-1348-3
A detailed look inside the development of the Eniac. Extensively researched, a great read.
In the Beginning: Recollections of Software Pioneers
Robert L. Glass
ISBN 0-8186-7999-9
This book comprises a series of anecdotes and histories, by 15 software pioneers. It's
unevenly written, but tremendous fun reading.
Technical Design Labs published "TDL Z-80 BASIC" in 1977. I remember hearing that MS had 'appropriated' TDL's BASIC -- after inspecting the source and declining to license it. When MS BASIC appeared on the market, I compared it with TDL's. Imagine my surprise when undocumented errors in TDL's BASIC had identical counterparts in the MS product. Draw your own conclusions.
Slashdot Mirror
Consider Thompson's seminal paper on trusting trust. It is possible to conceive of embedding a backdoor in the software that translates your high-level UART design into layed out gates, P and N junctions, connections, etc., with the addition of a hardware backdoor that activates on a sequence of bits traversing the circuit. With a circuit of high complexity it might plausibly go unnoticed. Possibly it would be within the capabilities of nation states. Easier to envision for certain classes of hardware, but certainly a very hard problem. If you have the skills to write a recognizer for a hardware UART, that can add a backdoor, could you please also also write an agent that could Do What I Mean as I code my programs? :)
Thanks. I recently implemented denyhosts... I love it.
Any chance you can share what attack vector they used to root your system?
Why not just use zgrep?
These will make great reading for your dad:
The Computer From Pascal to von Neumann
Herman H. Goldstine
ISBN 0-691-08104-2 (Hard cover) ISBN 0-691-02367-0 (Paperback)
This book covers the WWII era development in detail. Well written. I think the author is the Naval officer responsible for getting the Eniac funded when all the other experts said it wasn't worth developing.
Eniac - The triumphs and Tragedies of the Worlds First Computer
Scott McCartney
ISBN 0-8027-1348-3
A detailed look inside the development of the Eniac. Extensively researched, a great read.
In the Beginning: Recollections of Software Pioneers
Robert L. Glass
ISBN 0-8186-7999-9
This book comprises a series of anecdotes and histories, by 15 software pioneers. It's unevenly written, but tremendous fun reading.
Hope that helps!
Technical Design Labs published "TDL Z-80 BASIC" in 1977. I remember hearing that MS had 'appropriated' TDL's BASIC -- after inspecting the source and declining to license it. When MS BASIC appeared on the market, I compared it with TDL's. Imagine my surprise when undocumented errors in TDL's BASIC had identical counterparts in the MS product. Draw your own conclusions.