Talk:Commodore 128/Archive 1

This is an archive of past discussions about Commodore 128. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 1

Hackaday post so wrong

Regarding the link (hackaday.com: Guest Post: The Real Story of Hacking Together the Commodore C128, by: Bil Herd)...I mean. Wow. Don't get me wrong, I have a lot of respect for Bil Herd and I adore the C128 (I own several including the 128DCR). But that entry was just full of errors. The C128 was not the last 8-bit computer produced. The SAM Coupé came after the C128. Yes, the C128 sold millions while the SAM Coupé sold thousands but where do we draw the line on "mass produced"?

Second, the C128 was NOT the first computer to break the 64K barrier. The Apple IIe was commonly sold with 128K expansion and the IIc and IIc+ had the expansion built in.

Third, the C128 was not the first computer to use an MMU. Again, the IIe and IIc had built in MMU's.

I see this all the time. People remember things from LONG ago and they are just inaccurate. All humans suffer from this. Our memories are just not as good as we think they are. For example, I'm a Java developer and I could "invent" some Java app. Years later I could claim I was the first. Well, unless I was very intimate in the .NET/C++/Pascal/etc. world, I wouldn't know if my invention was already there. I would "remember" it being the first. I doubt Herd was all that intimate with Apple, Sinclair, etc. And using them a few times does not count as being intimate with the technology.

Anyway...end of RANT....long live the C128!

--cbmeeks cbmeeks 22:01, 27 November 2017 (UTC)

C128D release dates

I know the 128D was released in 1987 in the United States; I have the back issues of Run Magazine and Compute!'s Gazette that talked about the release. I believe I remember reading that the 128D didn't pass the FCC initially, so it was released earlier in Europe than it was in the United States.

Can anyone confirm the date of the 128D's European release? We need to note the discrepancy. I know that 1985 date isn't correct for the United States. In 1985 and 1986, all that existed in the States was the "flat" 128. --Dave Farquhar 13:08, 29 Jul 2004 (UTC)

I just checked one of my Gazette issues, and by their report from some winterly trade show the D model was definitely released in Europe in '85. One of my childhood pals actually had one. I'll try to find the months of release for both models -- I guess it's somewhere in the pile of mags.

Actually I had one which was released 1984. --80.98.242.95 (talk) 23:01, 10 June 2009 (UTC)

--Wernher 22:18, 9 Feb 2005 (UTC)

C128 trivia

I wrote the original part of the C128 trivia section (now in parantheses). I couldn't remember the exact memory address so I left it unfilled. The later, more specific addition (which is not in parantheses) seems to be talking about the exact same thing, but I am not fully sure. Could anyone confirm this? --Anonymous

But notice: one example says: [Write] 0xFF (255) to memory address 0xD02F (53295)[.] [...] On the 64 this memory location would always contain the value 0xFF no matter what was written to it[.]
while the other one goes: [M]onitor the contents of memory address TBC1 (<decimal value>) for a while. A real 64 used this address as the internal hardware register TBC2 (<decimal value>), whose contents changed hundreds of times per second.

In other words, the two methods given as examples are somewhat opposite of each other. I'll try to find the address you mention in the latter example, as I sit on piles of CBM docs. Some day (or night) I might just bump into it. --Wernher 04:25, 9 Feb 2005 (UTC)

I am the "Anonymous" who wrote above. Thing is, I didn't remember the exact details about the way to distinguish the two computers, only that it involved writing something to a certain memory address and reading it back after a second or so. On a C64, you didn't get back what you wrote, but on a C128, you did. What you did get back on a C64 and why this was so had only become a blur. So it's entirely possible the second example is wrong. 85.76.152.179 20:23, 9 Feb 2005 (UTC)

Well, I'd say there's a good chance you're on to something that really exists, as "my" example does not at all call for a long delay between the write and read (more than a couple of instruction cycles, that is). But I'll put "your" example in wkp source comments (thus hiding it from the displayed article text while still keeping it in there) until we or someone else find the pertinent memory address. --Wernher 22:18, 9 Feb 2005 (UTC)

In this section is a bit about "unrealised multitasking". The article isn't entirely correct because those features (Multiple zero page and CPU stack locations) are not strictly needed for multitasking to be written for the C128 or even the C64. Pre-emptive multitasking was easily coded just by using the NMI and a timer interrupt, this would pop the current stack into memory, store the PC, restore the other stack context by pushing from a memory backup then finish by setting the PC. Multiple thread contexts could be created just by having more backup memory. Fnag aton 23:12, 5 June 2007 (UTC)

I agree with Fnagaton. There are several preemptive multitasking operating systems that support the Commodore PET and Commodore 64, including GeckOS, LUnix and Contiki. As such, multitasking is not unique to 8722 MMU based 650x systems like the C128.

At most, the MOS 8722 buys you speed. This is because the MOS 6502 series has a fixed stack location at page 1 (addr $0100-$01FF). Each time you have to perform a context switch, you have to copy the contents of the outgoing task's stack elsewhere, then restore the incoming task's stack back to page 1. That's a very slow process. The MOS 8722, however, allows you to repoint segment 1 (addr $0000-$4000) elsewhere in memory. So there is no need to copy the contents of the stack around in memory.

The problem with the 8722 is that it only works in chunks of 16KB using 16KB boundaries. That means that any multitasking OS out there that uses the MMU would probably require a minimum program memory footprint of 16KB. For a system that only has 128KB, that's somewhat wasteful.

A much better solution would have been to use a Western Design Center 65816 derived processor in the C128. The 65816 can relocate the stack to any location within the first 64KB of memory with no alignment restrictions. So instead of 16KB, a task might only require a 512 byte footprint (half for its own zero/direct page, half for its stack). If you kept your custom chips and ROM outside of bank 0, that could allow for up to 128 programs using such a scheme.

I'll clean this section up sometime this weekend. Any comments are welcome. Dinjiin (talk) 01:47, 4 July 2009 (UTC)

VIC-II in C128 mode

The VIC-II is available in native mode, isn't it? Mirror Vax 00:36, 6 March 2006 (UTC)

Indeed. All the C128 users who had no RGBI monitor would be in a bad fix if not. :-) Incidentally, a cool feature that not many users had the opportunity to check out was the dual display capability of the 128's native mode: simultanously using the 40-col and the 80-col display---e.g. for showing graphics on the former and text (menus, src code, or whatever) on the latter. --Wernher 04:30, 8 March 2006 (UTC)

there were a couple 2-player adventure games published in ahoy magazine that used dual screens. —Preceding unsigned comment added by A plague of rainbows (talk • contribs) 16:54, 17 December 2007 (UTC)

Depending or not depending on x ?

From the Trivia section: "The Commodore 128's BASIC V7 [...] could be crashed or cause the computer to reboot by executing PRINT""+-[x] (where x is any integer), depending on the number entered for x."

This might seem somewhat ambigous: "x is any integer" and, at the same time, the outcome of the direct program line is "depending on [...] x". Is the real meaning behind this confusing statement something like the following: some value(s) of x will make the machine crash (i.e., freeze/lock up) while some other value(s) will make it reboot? If so, it would also be nice to know of some "working" :-) example values to put in the article. --Wernher 17:19, 17 March 2006 (UTC)

I think the entire statement should probably be removed (but I'm not going to do it without discussing it here first). In order for that crash to work, you have to deliberately cause a TYPE MISMATCH error (you can't add a number to a string). No programmer would ever make that mistake, except by accident. I have no idea why it crashes if you try to add a negative integer, but gives you a TYPE MISMATCH error if you try to add a positive integer. Also, it doesn't just affect Commodore computers... it affects the Apple II series as well (though instead of crashing the computer, it dumps you to the Apple's built-in machine language debugger.) 67.61.121.158 (talk) 21:25, 5 October 2013 (UTC)

Bil Herd's comment

Tidied this up as a quote rather than a comment directly from the man himself. —Preceding unsigned comment added by 212.24.80.93 (talk • contribs) 17:45, 19 June 2006

There's a really odd line in the article that says something like "Bil Herd comments on the C128 article..." That is not very Wikipedia-like.JettaMann (talk) 14:22, 27 April 2010 (UTC)

Numeric Keypad

The main article would be improved if an inaccuracy were corrected. It implies that an unnamed suite of business software would not work because it couldn't find the numeric keypad. I don't remember if Commodore Business Machines actually offered its own "suite" of business software (CBM generally didn't offer very good software but relied on independent programmers to do that sort of thing for them, usually outshining the giant by miles) but there is a numeric keypad. If you bother to take a look, you will see that the numeric keypad exists just right of the keyboard. (If I recall correctly, you can distinguish keypresses on the keypad from those on the regular keyboard by testing a new line of I/O that was fortuitously mapped to address one of the bits in address $01.)198.177.27.21 07:47, 26 May 2007 (UTC)

The keypad sent different keycodes too. They were totally different keys that happened to be programmed have the same effect - lots of sw didn't work with it. —Preceding unsigned comment added by 69.125.110.223 (talk) 16:50, 19 December 2007 (UTC)

The keypad never "sent" keycodes. Keycodes were created in the Commodore Kernal SCANKEY routine where specific lines of I/O are brought high (by writing a '1'), and then read (i.e., 'tested') for a 0 (i.e., a ground state) again and again (in very fast sequence, so there appears to be simultaneity), and wherever the lines are zeroes (i.e., grounded), it is safe to say that a user was pressing a key. The keyboards were neither intelligent nor detachable like modern PC keyboards. Transactor magazine had an article about N-key rollover, which might be helpful and instructive to consult at this point. Dexter Nextnumber (talk) 22:14, 15 December 2009 (UTC)

Fixing another inaccuracy

The C-128 works just fine in 2 mhz mode. The main article says you have to turn off your 40 column monitor to get it to work in 2 mhz mode. This simply isn't true. You just have to turn on an 80 column monitor. You are never required to turn off the 40 column monitor. To prove this is true, hook two monitors up to your C-128, one of the monitors being a 40 column monitor and the other being an 80 column monitor. Now you can switch back and forth from SLOW to FAST mode quite easily without ever turning a monitor off. You can leave both monitors turned on, and even output data to the separate monitors - and switch freely from 1 mhz to 2 mhz. There is never a requirement that you turn off the 40 column monitor. You can leave it turned on all the time that you are in 2 mhz mode. Just remember to plug an extra monitor into your computer so you can enjoy 80 column mode when you switch to fast 2 mhz mode.198.177.27.21 07:54, 26 May 2007 (UTC)

Perhaps the originator of the statement (that the C128 doesn't work in 2 MHz mode) simply thought of the fact that the 40 col monitor output (the VIC-II) didn't produce a legible signal in that mode? A bit of a letdown for those C128 owners that didn't own a (fairly expensive) 80 col monitor back in the day... But, as you say, the machine itself works just fine at 2 Mhz, no problem there. --Wernher 18:39, 14 July 2007 (UTC)

Commodore had the 1802 monitor for just those users. It was a composite monitor that produced a fuzzy but mostly-readable 80 column monochrome display on the 128. —Preceding unsigned comment added by 69.125.110.223 (talk) 18:37, 17 December 2007 (UTC)

Confusing Text About CP/M

I still own one of these machines and don't understand the following:

SPECS: Zilog Z80A @ 4 MHz CP/M Mode: ...Z80 processor ran at an effective speed of only 2 MHz...

The Z80 had to share the address and data buses with the 8502. Therefore, a bus cycle had to be 2 MHz to maintain hardware compatibility. CP/M on the 128 was a bit of a dog, and I, for one, was never able to understand why Commodore even bothered. By the time the 128 was available (late 1985) CP/M was rapidly falling out of favor.

BDD 06:51, 17 December 2007 (UTC)

They did it to give the computer a business-y cachet. They could say it ran wordstar, dbase etc never mind how many people would actually use that.69.125.110.223 (talk) 18:40, 17 December 2007 (UTC)

Added more HI-RES (NON YELLOWED) photos

Also added the original box photo. Too bad I don't have a better one but the computer itself looks good. And no yellowing! These computers are SO hard to find that are not yellowed.

cbmeeks 13:36, 11 September 2007 (UTC)cbmeeks

C128D

I reworded the comment about using PEEKs and POKEs to control the VDC. That was a very unreliable method, owing to the fact that those two BASIC verbs use LDA(xx),Y and STA(xx),Y addressing (where xx is a zero page pointer to the VDC registers). MOS Technology warned in the VDC documentation to not use indirect addressing to talk to the chip. The only reliable way to talk to the VDC from BASIC was by calling the screen editor ROM subs at $CDCC (write) and $CDDA (read).

BDD 06:59, 17 December 2007 (UTC)

No in-text citations

There are no in-text citations (that I can find) in this article.

It's a shame that someone obviously spent a lot of time writing this & failed to include a single cite/source. Especially a potentially contentious assertion such as the one about the Commodore 64's power supply "blowing up". (I had a '64 for years, learned to type on it, as a matter of fact; my parents got it for my brother and I--we were one of the "first adopters", tho' I'm not sure that term even existed in 1982--and I can assure it never "blew up" nor did I know anyone who had one ever have it explode, including a friend who taught himself BASIC programming on it.)

I was 11 and my parents would not have bought it if there'd been any (probable) chance of the thing exploding!

If '64s were still available such an unsourced assertion might cause legal issues.

The works listed at the very bottom of the article are not "references" they are, in fact, bibliography.

In any event, the article needs a good deal of sourcing.

PainMan (talk) 12:31, 27 November 2008 (UTC)

The power supplies didn't "blow up", but they were prone to failure due to overheating due to the fact that Commodore didn't ventilate their power supplies. The 'black brick' style in particular was a solid block of epoxy around a circuit board, with a plastic case around that. The best way to prolong their life was to either keep them cool somehow, or to always unplug them when the computer wasn't turned on, but eventually the linear capacitor would fail, which would cause the voltage regulator to fail, which would cause a power surge that would cause the computer to crash (or even destroy the DRAM chips). Most of the people I know who still have a working Commodore 64 either made their own power supply (all you need is a 5V DC wallwart and a 9V AC converter), modified a Commodore 128 power supply by making an adapter or cutting off the end that plugged into the computer and replacing it with a Commodore 64 7-pin DIN plug, or bought a third-party power supply. 67.61.121.158 (talk) 21:32, 5 October 2013 (UTC)

Alternatively, you can also repair the original PSU – did a few dozen in the 80s. Just replace the regulator and fit it with a decent heat sink. Zac67 (talk) 13:24, 6 October 2013 (UTC)

nonsensical sentence about subdirectories

The C128 Mode includes:

With these three drives, more complex drive data arrangements were also made available to Commodore users in the nature of "track and sector" oriented subdirectories, a feature not available to PC users, who instead had to convolute their file allocation tables to do the same thing.

Huh? What are "track and sector oriented subdirectories" (and why is "track and sector" in quotes)? And in what way do subdirectories on a "PC" (presumably meaning MS-DOS and MS-Windows) involve any more convolution of the file allocation table than anything else?

This nonsensical sentence fails to explain anything, and leaves me with the impression that the writer doesn't quite understand it either (or at least doesn't have a handle on the nomenclature).
—überRegenbogen (talk) 00:12, 25 February 2009 (UTC)

Many users accessed their files by plugging track & sector links into the job queue (found in all Commodore compatible disk drives). A track & sector link is a pair of bytes that are found adjacent to each other. I know of no software written for the Commodore 64 or Commodore 128 that recognizes a "file allocation table" (whatever that is). To this day, I've never exactly understood a FAT. Expecting a 1581 directory to look like a PC directory is usually at the heart of most PC users' difficulty in using a 1541, 1571, or 1581. 216.99.201.57 (talk) 20:52, 4 May 2009 (UTC)

Exactly. You just can't do the same thing with a PC "file allocation table". Some people even went so far as to write little machine language routines that generate (i.e., deposit) a series of track & sector links in a known place on the disk, and after that is done, you have no use for the rest of the directory. Some people over-wrote their Block Allocation Maps with T&S chains. Truly elegant programming that you just can't do with a PC. Dexter Nextnumber (talk) 03:47, 29 November 2009 (UTC)

Apparently someone would like a separate article on "track and sector oriented subdirectories." It's too bad PC users don't have that sort of thing, and can't understand it. 216.99.201.148 (talk) 03:50, 8 October 2010 (UTC)

Commodore 128D Photograph

I uploaded a new photo of the 128D to WikiCommons. Just to quell any suspicion: the Wikipedia logo was not photoshopped onto the monitor. The 128D is actually displaying B&W bitmap of the logo in PBM format. —Preceding unsigned comment added by Jcassara (talk • contribs) 00:43, 25 March 2009 (UTC)

Successor

The Commodore 128 really never had a successor. It was basically a dead-end within the Commodore 8-bit series. The unreleased Commodore 65, while more powerful than the C128, was really a successor for the C64.

Claiming that the Amiga 1000 was a successor really seems wrong. Both the C128 and A1000 were released at roughly the same time. If you do consider the Amiga series as a successor, then I'd think that the Amiga 500 really came next. Many C128 business users who were too cheap to buy an IBM PC never would have paid for an Amiga 2000. -- Dinjiin (talk) —Preceding undated comment added 23:06, 1 July 2009 (UTC).

If nobody objects I'll fix this. Xorxos (talk) 21:35, 27 August 2010 (UTC)

Historical and technical inaccurracies

With these three drives, more complex drive data arrangements were also made available to Commodore users in the nature of "track and sector" oriented subdirectories, a feature not available to PC users, who instead had to convolute their file allocation tables to do the same thing.

Who put that in this article? Only the 1581 had a "subdirectory"" feature in its DOS, and it wasn't a real subdirectory. Subdirectories were available on PCs starting with version 2.0 of PC-DOS, requiring no munging with the FAT. Nearly all Commodore eight bit floppy drives had direct access by track and sector available by writing to the command channel (the U0 and U1 commands, or their older BR and BW equivalents).

That is incorrect. The 1541, 1571, and 1581 could all have subdirectories. It's just that you had to learn how to M-W and M-E your own code to put them where you wanted them, and the same thing to access them. The firmware inside these machines was a little bit on the limited side, that's why you had to resort to M-W and M-E to implement a subdirectory. Dexter Nextnumber (talk) 22:32, 29 November 2009 (UTC)

...externally expandable to 640 KB...

No Commodore 8 bit computer was "externally expandable" to anything. The so-called RAM expanders were not mapped into MPU address space and thus could not be seen as RAM. Storage to or retrieval from a "RAM expander" was accomplished by machine code that talked to the DMA controller in the "expander." BASIC 7.0's FETCH and STASH ultimately programmed the DMA controller to do the actual grunt work.

Assembly language programmers never expected an REU to extend the 6510 addressing space in a linear fashion. They knew ahead of time that only 64K was addressable from the factory, and this remained a physical limit at the home. Maybe BASIC programmers expected the addressing range to be greater than 64, but assembly language programmers knew ahead of time that 64K was the physical limit. To rise over this limit, programmers expected to have the REU provide them with "direct memory access" to bytes (and byte ranges) in the REU, and this is exactly what the REU did. It provided the machine language programmer of fiddling with the REU at $FF00 (which timid programmers found easier to access at its mirrored location, $D500). Dexter Nextnumber (talk) 21:53, 15 December 2009 (UTC)

In the autumn of 1985, Commodore released to the European market an updated version of the C128 with a redesigned chassis.

It was very late in 1985 before any D model was released. Since the flat 128 itself wasn't widely available prior to September 1985, the above statement is historically inaccurate. BTW, general availability of the 128DCR in North America didn't come until the fall of 1986. I purchased one in October of 1986, right after the machine started shipping. It was final assembled in Canada.

Bigdumbdinosaur (talk) 18:06, 4 September 2009 (UTC)

Pantergraph deletions

Please refrain from just deleting the Trivia section yet another time - the content should be moved to the other sections. If it's just deleted nobody has a chance to move it any more. If you have problems with this, please discuss it RIGHT HERE. -- Zac67 (talk) 10:12, 21 January 2011 (UTC)

Relevant information was moved. Stop re-adding trivia section.

I've once again replaced the deleted Trivia items. Feel free to move them to a better location but please refrain from deleting material. And please sign your talk contributions. -- Zac67 (talk) 12:05, 29 January 2011 (UTC)

Wikipedia:Manual of Style (trivia sections) deprecates trivia sections in an article. This section had been flagged as a trivia section for many months. If these facts are verifiable and somehow important to a better understanding of the Commodore 128, rewrite them into proper form in an organized inclusion into the article. --Wtshymanski (talk) 22:43, 29 January 2011 (UTC)

This is not the way it works. Moving items to somewhat suitable places is appreciated but you don't just delete what you can't work out and wait for someone else to re-add it. Trivia sections could be removed by a bot in the blink of an eye - it is not desired to lose that information. If you consider details not notable I challenge you to explain your reasons.-- Zac67 (talk) 23:22, 29 January 2011 (UTC)

If someone could provide a reference for the material, it might usefully be restored to the article. Seems unlikely in the extreme that every version of BASIC had the same bug in it. How do we know the C128 is unique in having so many different types of memory? It's certainly not unique for having more than one type of processor; the machine you're typing on almost certainly has a half-dozen different microprocessors in it. --Wtshymanski (talk) 15:49, 30 January 2011 (UTC)

The article isn't talking about every version of BASIC, just the Commodore 8-bit ones. I believe it to be correct, so KEEP the current text until references are added or sufficient doubt is raised. -- Zac67 (talk) 07:13, 16 February 2011 (UTC)

{{outdent} I challenge the deleted text. It is unlikely on the face to be correct, and is a trivia section which is deprecated by WP:TRIVIA. "I believe" is never sufficient on the Wikipedia. If you have citations that demonstrate the accuracy of the disputed statements, please share them with us. --Wtshymanski (talk) 14:31, 16 February 2011 (UTC)

Fine. So it's 1 vote for delete, 1 for keep. Any more opinions? -- Zac67 (talk) 19:51, 16 February 2011 (UTC)

update: references added, dispute settled. -- Zac67 (talk) 20:07, 16 February 2011 (UTC)

The citations don't support the claims. The BASIC bug has been put in-line with the text and revised to match the claim in the source. A source that quotes Wikipedia is not a valid reference. --Wtshymanski (talk) 21:46, 16 February 2011 (UTC)

"The Commodore 128 family of computers are very unique - having more than one main CPU gives them the ability to run three different operating system" / "Commodore 128-serien was (and still is) quite unique. With its two processors you have the opportunity to run three different OS." - the D models have an additional 6502 for the 1571 part. For those able to read this should be sufficient reference for the - pretty obvious - specialty of the 128D. There is no circular reference as no source cites Wikipedia. -- Zac67 (talk) 07:34, 17 February 2011 (UTC)

The citations don't say the C128 was unique in using more than one processor chip - the Commodore 64 had a Z80 cartridge, the BBC Micro had a range of add-on processors, the Commodore SuperPET had a 6809 as well as a 6502, and the Coleco ADAM had a fleet of processors. (If you start counting I/O processors, every PCish machine has multiple processors - keyboard, hard drives, video, USB devices, and so on.) It's not "very unique", it's not even unique in the Commodore product line. None of the citations offered say that the different kinds of RAM were unique to the C128 and this seems to be a pretty common concept in home computers - the Amiga folk are always talking about "Chip RAM" and "fast RAM", and every computer has different types of RAM. The deleted citation http://tripatlas.com/Commodore_128 is a copy of the Wikipedia Web page. --Wtshymanski (talk) 14:09, 17 February 2011 (UTC)

Splitting hairs? The article is talking about CPUs and obviously it refers to a stock machine. And there are three CPUs sitting right next to each other on the very same mainboard. Hard to believe? Oh well,... -- Zac67 (talk) 16:21, 17 February 2011 (UTC)

The references say some thing like "The Commodore 128 is unique." Hurray for the Commodore 128. They don't say "The Commodore C128 is unique because it is the only computer in the history of the world to have two processors and five kinds of RAM". It's not even unique in the Commodore line, the C64 had a Z80 cartridge and the SuperPet had multiple processor chips. It's still unsupported. --Wtshymanski (talk) 19:42, 17 February 2011 (UTC)

Response to third opinion request:

I am responding to a third opinion request for this page. I have made no previous edits on Commodore 128 and have no known association with the editors involved in this discussion. The third opinion process is informal and I have no special powers or authority apart from being a fresh pair of eyes.

It seems you've resolved many of the issues from the conversation above already. Regarding WP:TRIVIA, the guideline does mention that 'if information is otherwise suitable, it is better that it be poorly presented than not presented at all', and so this shouldn't be used exclusively as the reason to remove such a section. The sourcing issue is valid, though.

Skimming the history, I see a few sources were offered, one of which is [1]. This is a self-published personal page and doesn't qualify as a reliable source - if this is indeed a true statement about a weakness in the Commodore 64's BASIC implementation, there should be better sources out there. You've already resolved the Wikipedia mirror source.

The sources for the 'unique' CPU claim both seem weak but usable. They both make assertions that the Commodore 128 is unique with the connected reasoning that its multiple CPUs allow it to run three different operating systems:

The Commodore 128 family of computers are very unique - having more than one main CPU gives them the ability to run three different operating system [2]
Commodore 128-serien was (and still is) quite unique. With its two processors you have the opportunity to run three different OS. [3]

This is a connected fact. The sources are saying that the Commodore 128 is unique not because it has multiple CPUs, but because it has multiple CPUs that allow it to run three different operating systems. Any other representation of this would be contradictory to the sources. The sentence in the current version of the article is misleading and attempts to state more than what the source does. Even if the fact itself isn't true, I think Wikipedia's principle of verifiability over truth would apply here, unless contradictory sources can be found.

On cursory inspection I didn't see any source for the 'multiple RAM' assertion, but the history suggests that this is no longer a point of contention between you? This seems like a fact that a reader is likely to question, and as such an inline source should be provided at the statement. If it's mentioned in another source elsewhere, make another cite to the same source there. Note that, as in the paragraph above, the sources do not state that multiple RAM types is a unique feature of the Commodore 128, nor multiple CPUs, nor different video chips.

My suggestion for a less ambiguous sentence would be:

The C128 is considered unique due to its multiple-CPU design, which allows it to run three different operating systems, each with their own differing performance and memory characteristics.[4][5]

Hopefully this covers the issues you've been having. I've removed the request from the 3OR page - if you're still having problems on the same issues, feel free to contact me via my talk page for my opinion. If you need more community input, you might like to make use of WP:RfC for a broader range of help and suggestions.—TechnoSymbiosis (talk) 02:19, 18 February 2011 (UTC)

Original research

The article has been tagged as containing Original research for nearly 2 years. Since the person applying the tag failed to provide any clue as to what part(s) of the article they believed to be original research, it is hardly surprising that nothing has happened. I have deleted the tag but if anyone genuinely believes that the article does contain original research, then by all means restore the improvement tag. Also, provide a clue by either individually tagging the contentious point(s) with an {{Original research?}} tag at the appropriate place or raising the issue on this talk page. 109.145.22.224 (talk) 16:08, 2 May 2012 (UTC)

More C128 history

Here are some useful links for C128 history:

http://hackaday.com/2013/12/09/guest-post-the-real-story-of-hacking-together-the-commodore-c128/
http://c128.com/ (Bill Herd's website)