Denial

Mike Taylor who wrote Magic Mirror and Nosferatu for Terminal Soft recently posted me two tapes to have a go at archiving for him.

Preserved from one tape is the 3rd in the series that was never released called "Heaven and Hell" (8k expansion required), which you can grab from here:

http://www.gtw64.co.uk/extras/vicbits.zip

There are also some interesting extra files on the TAP images, including I think a 3D Maze game that Mike seemed to be developing (Two versions on the tape) and some title screens of some kind - possibly done for fun.

Unfortunately the second tape that Mike sent is corrupted (Hitting the same problems as I think Mayhem did, as I think he has had a go at this one?), and i'm getting in touch with Luigi Di Fraia to see if he can help salvage the 4th and 5th adventures in the series that never got released.

In the ZIP above, i've added the 5th adventure (16k) which seems to mostly load in - but has an error when running. You can list the program, so i'm wondering if it might be that a line or two are just corrupted slightly in the load - and this one could be saved. If anyone fancies taking a look, feel free

On tape, BASIC stores the data twice, original and copy.

A '?LOAD ERROR' can occur, because the copy was clipped by accident. In that case, the VIC-20 is not able to compare successfully the program already loaded to memory with the copy on tape.

In most cases (and most probably here, too,) the original still has been loaded error-free into memory, but after a '?LOAD ERROR' one important pointer has not been set, the one to the start of variables (45/46, aka 'end of program').

Even if the program was loaded without defects into memory: when you blindingly RUN it with the preconditions stated above, the variables now created by the BASIC interpreter will overwrite the program from its start address, causing all sorts of unrelated errors like '?UNDEF'D STATEMENT' or '?OUT OF DATA'. Effectively, the program destroys itself.

...

It is possible to restore the pointer in 45/46 when LIST reveals a program that is seemingly in order - that action must be done, *before* it is RUN: ... after which it also can be saved again, and the new copy won't require that pointer restauration anymore.

Greetings,

Michael

Cheers for that Frank, any idea which year "Heaven and Hell" is? I assume either 1983 or 1984. I was going to try what Mike suggested even if Mike hadn't suggested it, and it appears to have worked. The game loads up seemingly fine and plays okay (having tried a few commands and walked about).

Are all the raw dumps available?
Did you analyse the dumps to see what the problems are?

Just an update, but Luigi has taken a look and I think we're close to a working version of "The Final Challenge" (5th adventure), but the cleaned tape is still producing the same error. "Stairway" is in a bad way (possible moisture damage), but Luigi is potentially being sent the tape to extract from (Just confirming with Mike now - as it may mean ripping the tape open)

Thanks Mike for the great advice!.... based on your notes and also Luigi's cleaned version, here is a D64 with the game listing:

http://www.gtw64.co.uk/extras/challenge.zip

However, i'm still getting the ?Out of data error. I'm not sure if anything can be salvaged from this.

@Mayhem - Fifth Challenge seems to be dated January 1984 (Though mostly developed through 1983. That means that "Stairway" will be 1983. "Heaven and Hell" i'm not sure. If the 2nd game was 1983, then it will be 1983 too (Otherwise this could be 1982 - Mike seems to be unsure on his website what year exactly).

@Bokvamme - I have all the raw dumps here if anyone wants to take a look themselves. I wouldn't know where to look regarding problems, hence why I sent to Luigi to examine.

fgasking wrote:However, i'm still getting the ?Out of data error. I'm not sure if anything can be salvaged from this.

Tested the linked game, I see you are getting the OUT OF DATA error on start. As I mentioned above I loaded the TAP you originally provided of Fifth Challenge, used the POKEs and resaved, and the game loads up fine and seemingly runs fine. I wandered around a few locations, picked up a few items, tried to use a few of them... no errors whatsoever.

So you can compare, I've uploaded it here so perhaps you can have a look... although I can see in your version, line 30420 has a rogue character on it. I'm guessing there are others that are interfering with the reading of the DATA statements.

Ah, my apologies Mat - I tried the pokes with the fixed version that Luigi did and not the original port.

That seems to be fully working then! Obviously it will need fully testing, but i'll pass this over to Mike Taylor to take a look. Thanks Mat!

So sounding like there is just the one to go now

fgasking wrote: @Bokvamme - I have all the raw dumps here if anyone wants to take a look themselves. I wouldn't know where to look regarding problems, hence why I sent to Luigi to examine.

It's in the best of hands then...