Ah. Ok.

Would it be easy, Kerm, within the routine you posted to have it add the 'prgm' token to every entry in the appvar, and increase the prepended # by one

So instead of reading 03414141, it would read 04(%)414141, where (%) represents the 'prgm' token. **And the prgm token is 1 byte, right?**

And, what is a checksum (what does it check). Can one be written that checksums the OS and exports the result to an appvar called "BChkSav"

Yup, token 'prgm' is 95, or 0x5F, so that would be easy to insert. Might I ask why? A checksum is a small number used to represent a much larger piece or segment of data, in order to ensure that it is intact. A naive checksum implementation takes the sum of all the bytes in the data modulo some number; the .8xp checksum, for example, is [sigma(i=0,i<size,byte[i])]%(2^16). There are more complicated forms, like a CRC (cyclic redundancy check, used for floppy disks and CDs and some network transmissions), and MD5 and SHA "fingerprints". MD5 and SHA and similar methods are more robust to changes, whereas a naive checksum will yield the same value if you increment one byte and decrement one other byte. For your purposes, though, I think a simple checksum is fine.

Edit: On this very website I found that the OS has MD5 routines already built-in, but they only work on data in RAM:
http://www.cemetech.net/forum/viewtopic.php?p=58501#58501

As for the token, the reason is because when I cut the data, I would then have to implement another routine to annex the program name to the program token in order to use it as a name string in Axe.

As for the checksum, I wanted to incorporate it into the scanner. I was going to have it calculate a reference checksum during installation, then store it into an appvar. Then, I was going to have an option in the scanner to checksum the OS, and if it returns a different value, it outputs that something may be wrong.

The word you were looking for is *append, not annex.

Indeed, I figured that's probably what it whooaaa did you just say you want to checksum the *OS*?

Yes. A simple one? Is it possible?

Why bother? How much harder would it be for someone to do a quick RAM clear after modifying the OS if they really want to be malicious? I thought you were going to want to checksum a program.

How so, if the appvar is moved to archive?

Currently completed:

-Data
-Installer (waiting on mal hex search strings)

Left to Do:

-Get hex mal search strings
-Get program list generator routine
-Get hook installation routine
-Get OS checksum routine (maybe)
>Write main part of program
-Write Scanner
-Write Setting display and Settings change
-Write Update Definitions routine.

> means currently underway.

OK, then update the OS, RAM clear, and delete the AppVar.

Ok, so then forget the checksumming.

Yeah, I think it's not worth the trouble in the long run.

I'll take your work for it. You know better than me.

I'm currently getting the meat of the program and settings done. For the rest, I need to wait on the asm routines.

Oh and Kerm, I forgot to ask you: is it possible to somehow test the calc to see if a hook is installed? And how?

Ok. Now comes this. Who in this community would like to be a BETA TESTER for this project?

Which hook? There are bits to check at (iy+NN) for all of the hooks.

The RunProg Hook discussed above.

This is a place to start:

http://wikiti.brandonw.net/index.php?title=83Plus:Hooks:9BAC

Ok, so this "runprog" hook is really just a parser hook? (Haven't read through the topic yet)

There's no such thing as a RunProg hook per se, just a Parser Hook that detects program execution from the Homescreen.

And *only* the homescreen? Or will this still work for in programs?