C119!

[wte]

Ok, as airship don't want...

Now my most favorite question:

Why is a C 128 out of the "first batch" basically only a C 119?

WTE

[xlar54]

Wild stab at this one... since I couldnt find anything about a C119 commodore computer, my guess is that the first batch didnt actually have a full 128k, maybe due to a bug or something?

[StyleCHM]

Broken MMU chip?

[airship]

9K? I could see maybe how you could lose 8K, but 9K?!?

[hydrophilic]

I guess because writes to the upper memory (8K) in bank 1 would actually go to bank 0 (bug in the MMU) and because the common ram (1k) could not be disabled (common RAM configuration register not implemented).

[wte]

All answers wrong!

The "first batch" is related to the first kernal version (official kernal! I'm not talking about any prototypes)

WTE

[Andrew Wiskow]

Hmmm...  Where's Bil Herd when you need him? 

-Andrew

[hydrophilic]

Original KERNAL?  With the Caps-Lock "Q" bug?  I have that ROM... now to figure this out...

[StyleCHM]

Im not sure I understand..... who cares about the kernal anyway?

The kernal is something, like BASIC, that you switch out at your earliest convenience

[wte]

Im not sure I understand..... who cares about the kernal anyway?
The kernal is something, like BASIC, that you switch out at your earliest convenience

You shouldn't do so if you want to win this competition.

Read my question like this:

Why is a C 128 out of the "first batch" basically only a C 119?

WTE

PS: The capabilities of BASIC 7.0 (compiled!) and the VDC together were the biggest advantages compared to the C64. The most C64 programmers failed to program the C128 (it was to sophisticated for them).

[StyleCHM]

Bil really shouldve gone the whole hog and had 4Mhz ram and 2Mhz VIC-II.

But its a nice machine anyway

[hydrophilic]

Now I get it!

The answer is because BASIC reports 122365 bytes free.  That equals 119 K (rounded to the nearest K).

[airship]

But why does it?

[wte]

Now I get it!
The answer is because BASIC reports 122365 bytes free.  That equals 119 K (rounded to the nearest K).

Ha, ha! No, No! (good idea)
But, where is the difference between the first kernal and the later one?
All kernals are reporting the same value!

We are surching for a real bug - and because I didn't have an international C128 (but a German C128D with 16k VDC), I only can assume that also the first kernal of the international flat C128 has the same bug.

WTE

[StyleCHM]

It reports correctly but the BASIC pointers are set incorrectly, shorting you 9k?

[smf]

But why does it?

Space is reserved for VIC bitmap mode.

[wte]

No and No!

I'll give you a hint how I "calculated" the lost 9k (128-119=9).
First the C128 has 128k RAM (that's why he is called a C128) 

Using Basic you can't use all memory for your program code.
Some is used by the interpreter and kernal, some is used as memory for the variables, some is used for VIC text/graphics.
But this use is usual!

But this is not usal: The bug steels approx. 9k of memory from my code bank (bank 0).
And as I know now the bug was available in all early kernals and Commodore fixed it in later versions.

I discovered the bug in 1986 after some strange program crashes based on destructed program code.
I called Commodore in Frankfurt/Main but the expert was not reachable (I tried it several times).
So I wrote an article for the German RUN magazine but this mag was discontinued (1988) before publishing my findings.
Late 1988 the 64er published an article from an other dude dealing with "my" discovery. 

WTE

[BilHerd]

I LOVED the capslock Q bug, it showed how bad the QA department was and how useful the usergroups would have been if we had engaged them sooner.

For a time the C128 only had 64k due to a mixup in the films at MOS, they introduced a design artifact by mixing up layers belong to various revs that had all of bank 0 bleed into bank 1.  We almost didn't see it as Terry stored program from the top and variables form the bottom (or vice versa) and we got into a program big enough to actually see it.  after that it could be tested with a 3 line basic program... QA was still learning Basic at the time tho....

Bil

[StyleCHM]

I LOVED the capslock Q bug, it showed how bad the QA department was

And funnily (or sadly) enough, I dont think QA has improved in the last 23 years Bil. Ours is next to useless.

[wte]

Ok - some last hints:
Think on VDC and MMU
or use Google.

WTE

[airship]

Google's cheating. Even though I had to use Amazon to get the answer to my last 'win'.

[wte]

Isn't there anybody out there who knows about this bug?
Maybe you all are VICE lamers and don't have a real C128? 
Or maybe you all are lucky fellows and have only a "new" C128 and not a real old (first kernal, 16k VDC) version?

I'll lift the enigma tomorrow!  ;/

WTE

[nikoniko]

So... no one here has thought to check wte's website?

[Blacklord]

So... no one here has thought to check wte's website?

Cheating!

[hydrophilic]

I've got a 16k VDC machine with original kernal and still don't know!  Never stumbled across the bug appearantly.  Of course I usualy take over the whole machine and use the kernal only for keyboard/screen/disk I/O... in other words, if the bug affects RAMTAS, PHOENIX, cassette, RS-232, etc then I'd never know.

I checked wte's website and couldn't find the answer.  Of course I can read German about as well as I can walk on the Sun.  If they spoke French or Spanish over there I'd have better luck.  I've got a PDF of the German version of the C64 PRG which I like to look at occassionaly.  Sometimes I can actually figure out what they're talking about due to context.

I can't wait for the answer to be revealed...