Talk:KERNAL

0xF000 - 0xFFFF is not 8 KiB

Rather 4 KiB. Did the writer mean that the kernel occupies the range 0xE000 - 0xFFFF, which would be 8 KiB, or that it is 4 KiB long as indicated by the range in the most recent version? 88.112.2.159 15:28, 11 June 2006 (UTC)

Yes your right. The KERNAL occupies the last 8k of the 64k of the Commodore 64. The range is 0XE000 - 0XFFFF or 57334 to 65535 (http://www.tkk.fi/Misc/cbm/docs/c64-ram.html). StateOfTheUnion 00:35, 27 August 2006 (UTC)

His right, but yours is wrong. Hexadecimal E000 is 57344 in decimal, not that it matters, but don't be so careless. --93.106.43.117 (talk) 08:11, 13 September 2009 (UTC)

ALL CAPS

WHY IS THE TITLE OF THIS ARTICLE WRITTEN IN ALL UPPERCASE LETTERS?

cause capz r k00l!1! Actually, I think the KERNAL was spelled with caps, as it is spelled that way throughout the article. ST47 22:25, 23 August 2006 (UTC)

That is right. The KERNAL is written in all caps in the Commodore 64 Programmers Rererence manual which details all the routines in the KERNAL. StateOfTheUnion 00:32, 27 August 2006 (UTC)

Yeah, but the programmers reference guide put a lot of stuff in all caps like SPRITE etc. Kernal doesn't stand for anything.

The article explains KERNAL, and that it was probably a Bachronym.

---

John Feagans added the following comments on July 26, 2007 -----

KERNAL is the original name as I spec'd it. The spec section in the original article has the correct acronym. I'll let someone else edit how Bob Metcalf (ethernet PARC and 3Com founder) was a consultant at the Moorpark R&D facility for Commodore in the summer of 1980. He was a great inspiration.

The "N" was for networking. I provided hooks in the PET KERNAL ROM so that the "Teacher's PET" could be attached to a master-slave network. In the Commodore 64 there was a cartridge with the networking software and an interface card to the 6522 serial port. We had two test installations in New Brighton Middle School in Capitola, Ca. and the Watsonville Middle School in Watsonville, Ca. Both had a PET with 30 C-64's. I developed all the hardware and did the fabrication myself. Later a school in Wayne, PA was set up with a similar installation.

Another misconception is the origin of the Vic-20. The real creator of the concept was Jim Kennedy who was the Commodore engineering manager after Chuck Peddle left in August, 1978 to work at Apple. Jack Tramiel asked him to visit MOS and he talked to Bob Yannes where he got the specs for the VIC chip. He and I discussed a prototype VIC-20 -- but unfortunately Kennedy was canned in February 1980 when CHuck was re-hired by Commodore.

John Feagans Ele'ele, Hi.

Footnotes?

I see 4 footnotes, with only 2 references to them.

Same way in the year 2009. And there should be some mention of the 1581 Kernal, which is different from the C-128 Kernal, which is different from the C-64 Kernal. And I suppose the C-64 Kernal is a little different from the Vic-20 and PET Kernals. Dexter Nextnumber (talk) 03:25, 26 December 2009 (UTC)

Why the name

I think it was a play on the word Colonel, which is often pronounced Kernal. Here is an article on the history of the rank Colonel which gave me idea (exact same spelling - "Kernal"): [1]. So since the name "Commodore" is also a military rank it makes sense "Kernal" could also be. Unfortunately, this is an original research, so I cannot put this "find" to a WP article. :( It would be nice if someone could dig out some proof if this was really the case, in the meanwhile i hope my theory can help even here on the comment page. --Arny 00:46, 14 February 2007 (UTC)

That's a nice theory, but more likely Robert Russell is just a typical person who cannot spell. Stupid Americans. --93.106.43.117 (talk) 08:11, 13 September 2009 (UTC)

It's a misspelling of "kernel", which is a common OS term. — Preceding unsigned comment added by 107.220.114.180 (talk) 20:28, 23 June 2015 (UTC)

Did Commodore Calculators have a Kernal?

Does anybody know if the Commodore calculators of the early 1970s had Kernals? Dexter Nextnumber (talk) 03:27, 26 December 2009 (UTC)

No, they had a ROM though. A kernel only makes sense if you have layers of software, and a calculator is too simple to be built in layers. For Commodore computers, BASIC was the upper layer and the Kernal was the hardware interface routines. This way the command for PRINT could be be maintained separately from the routine that writes characters to the screen. — Preceding unsigned comment added by 107.220.114.180 (talk) 20:29, 23 June 2015 (UTC)

Not surprising

"Surprisingly, the KERNAL implemented a device-independent I/O API"

This is not surprising. Lots of machines of the era did something similar. kernal treated everything as a GPIB, Atari treated everything as a CIO. There's little difference in the big picture. Maury Markowitz (talk) 19:14, 25 September 2019 (UTC)

Bad examples of KERNAL jump table modification

"The jump table can be modified to point to user-written routines, for example rewriting the screen display routines to display animated graphics or copying the character set into RAM."

Those examples are not good, as they neither require nor benefit from KERNAL jump table modification. The KERNAL is of little help in implementing them, but also wouldn't interfere or be required to know about them. For instance, redefined characters would look different, but the KERNAL would not need to treat them any differently, not would it interfere with the video chip registers / memory regions involved in the redefinition. Examples I've actually seen: 1) Implementing disk and cassette "turbo loaders", where a changed jump table made the standard SAVE and LOAD commands use your replacement routines instead of the ones in the KERNAL. 2) Implementing an 80-column mode on the C64 using half-width characters, revectoring the KERNAL output routine to one that plots those characters in bitmap mode rather than using character mode. I'll update the page accordingly, but wanted to give more justification than what fits in a changelog entry... 92.35.31.146 (talk) 00:21, 23 August 2023 (UTC)