Talk:LGP-30

Who designed and produced this? Kd5mdk 21:51, 24 Apr 2005 (UTC)

• Agreed - should be merged. Mathew Graham 16:33, 17 Jan 2006 (AEST)

ACT-III

In what sense is this "Algol-like"?! —The preceding unsigned comment was added by Qu1j0t3 (talk • contribs) 22:21, 14 April 2007 (UTC).

It's obviously not Algol-like in any way. However, we can't just remove that, because of WP:NOR. Anyway, the source has two problems: it's primary and the statement of an individual, based on his opinion and memory. It's invalid as a source. The second problem is that it's a dead link. At this point someone should be WP:BOLD and just remove the incorrect claim. David Spector 02:39, 10 March 2010 (UTC)

Fractional multiplication

The article says, "...acting as if the binary point was on the left side of the word as opposed to the right side as most other computers assume." This sounds to me like fractional multiplication (retaining the most significant bits of the result), as opposed to integer multiplication (retaining the least significant bits of the result). I can't add this, because it is WP:OR. I'm not sure how to find a reference. David Spector 01:11, 3 March 2010 (UTC)

Mel Kaye

Mel Kaye actually wrote his legendary program for the RPC-4000. (from the Jargon file: http://catb.org/jargon/html/story-of-mel.html):

"The RPC-4000 computer had a really modern facility'
called an index register."

Mel Kaye's trick used the index bit, a binary digit at the end of the 32-bit instruction word which (basically) enabled the index register's function. This was only possible in the RPC-4000, as the LGP-30 did not have an index register, or an index bit. In the LGP-30, that bit was used as a "spacer" between sectors on the drum memory.

In fact:

"Mel's job was to re-write
the blackjack program for the RPC-4000."

There is some confusion, because Naher's account states that:

"The vital clue came when I noticed
the index register bit,
the bit that lay between the address
and the operation code in the instruction word,
was turned on —"

This is clearly impossible, as the LGP-30 had two bits between the opcode and the operand address (and it did not have a Next Address bit sequence after this). In order for the carry bit to overflow all the way to the opcode (and turn it into a jump instruction) there would need to be two bits set between the opcode and the address (that's probably how he did it in the LGP-30). If either of the intervening bits was set to low, the carry overflow would simply fall into the zero. Then the instruction would execute with the operand in address 0000 0000 0000.

Here's an examle of an instruction word from the LGP-30 manual:

    0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0
                           (-------)   (----------)(------------)
                           order bits   track bits     sector bits
                             = bring       = 20           = 00
                                       ( address = 2000 )

Here's the manual of the LGP-30: http://ed-thelen.org/comp-hist/lgp-30-man.html

And here's the one for the RPC-4000: http://www.bitsavers.org/pdf/royalPrecision/RPC-4000/RPC-4000_Programming_Manual.pdf

(the link opens a pdf, obviously).

So I changed the article to reflect this. Please free to propose a better wording. Stassa (talk) 20:08, 6 June 2010 (UTC)

Actually source says that Mel Kaye wrote his program (blakjack) in hexadecimal for both the LGP-30 and RPC-4000 (his rewrite to RPC-400 triggered that story). --89.25.210.104 (talk) 15:06, 18 September 2018 (UTC)

Hex is just a shorthand for binary. Stassa