On Monday, October 25, 2010, Christian Liendo wrote: Neat. There's a video of my 11/780 from VCF/MW 4 on the page. Pat -- Purdue University Research Computing --- http://www.rcac.purdue.edu/ The Computer Refuge --- http://computer-refuge.org

First of all if you have a problem with the article, please just say so and don't call it "s**t. If you find something inaccurate, say so don't be demeaning. --- On Mon, 10/25/10, William Donzelli <wdonzelli at gmail.com> wrote:

On Oct 25, 2010, at 9:08 AM, Christian Liendo wrote: Also, the sentence starting "The VAX instruction set well revered would later on influence" needs restructured. The paragraph after the bullet points has more of the same error.

On Mon, Oct 25, 2010 at 10:53 AM, William Donzelli <wdonzelli at gmail.com>wrote: ah yes, Will can be so honest ;) I think the missing point that hasn't been so well enunciated is the VAX is a >minicomputer< and the IBM 360 is the >mainframe< - quite big difference - perhaps should be specified in the blog. I hope this doesn't break any hearts in the DEC community ;) As far as mini's go, SEL had a 32bitter in '69 http://www.vintagecomputer.net/ragooman/computers_mini_products.html Nice job though, Chris =Dan --http://www.vintagecomputer.net/ragooman/

If you were the author, then I do apologize for being heavy handed. However, there are some basic facts wrong that could have been very easily corrected before publication by simple fact checks. To not do so is inviting people like me to comment. We, as a whole, flame reporters for getting basic facts wrong, so the rules of this game are pretty well known. -- Will

On 10/25/10, Christian Liendo <christian_liendo at yahoo.com> wrote: Besides the typos and the misplaced smart quotes, this stood out for me: o The VAX instruction set well revered would later on influence Motorola engineers when designing the 68000 CPU, Despite the obvious (to those that know all three relevant architectures and instruction sets) differences between the VAX and 68000 and the similarities between the *PDP-11* and the 68000, it's unlikely there was time for Motorola engineers to be influenced by the VAX. I have held an XC68000 (still owned by my former boss) with a manufacture date in late 1979. I also have preliminary (so marked on each page) data sheets for the 68000 from the same time frame that both point to the design history of the 68000 to be well before the release of the VAX. A quick google of the history of the 68000 reveals that its design origins began around 1976, which is consistent with the product development cycles of the era - three years from inception to shipping silicon. The 68000 is similar enough in instruction set and register architecture (yes, I know there are plenty of differences) that at that same company where my boss had the XC68000, they were able to expedite the porting of the entire PDP-11 codebase to the 68000 with a TECO macro (and the code shows it). They munged the code (literally) through a filter and only had to do some cleanup work around interrupts and such. This is something that is non-trivial for, say, turning 8080 code into 6502 code - it _was_ trivial for PDP-11 -> 68000. So there's my issue with point #2 of the "three important reasons" in the article. Disclaimer: I wasn't "there" in 1977. My first professional exposure to the PDP-11, the VAX, and the 68000 all came in mid-1984, several years after the invention of each, but I did spend quite a bit of time in the late 1970s in my 6502 world, drooling over the 68000 datasheet. The mere idea of moving from three 8-bit registers and zero page indirect references to eight 32-bit data registers and eight 32-bit address registers made my head spin with the possibilities. Six months after my first hands-on experience with the 68000, I bought an Amiga 1000 (the one I brought to VCFmw with the Rejuvenator). -ethan

Thanks I will update... When I researched it there was something that stated it was the first... So if my mistake I will update. --- On Mon, 10/25/10, Mr Ian Primus <ian_primus at yahoo.com> wrote:

On Mon, 25 Oct 2010, Christian Liendo wrote: Almost ANY use of the word "FIRST" in a history is challengable.

--- On Mon, 10/25/10, Dave McGuire <mcguire at neurotica.com> wrote: Those mainframes weren't 32 bit though. They were *much more* than 32 bit. What was it, two 64 bit registers or something like that? I don't remember, but I'm pretty sure it's not just a straight 32 bit machine :D Now, I AM sure that IBM had a mainframe that could do a million instructions per second before the VAX came out. DEC might have been the first one to turn it into a statistic with a cute name though. The VAX *may* have been the first 1 MIPS _MINI_ computer. I know that Prime didn't have one until like '79 or '80. -Ian

--- On Mon, 10/25/10, Dave McGuire <mcguire at neurotica.com> wrote: Hmm. Must have missed that flame war then. Was not aware of that - my knowledge of IBM mainframes is limited, having only read things like marketing literature and overviews. Now, Prime, on the other hand. That I know. I know for a fact that the Prime 200, released in 1972, was a 32 bit machine. Slow as hell, but 32 bits nonetheless. :) -Ian

On 10/25/10 11:24 AM, Sridhar Ayengar wrote: Ahh, but by what definition? You're obviously talking about the addressing capability. I was talking about data paths, and by that definition they're all 32-bit. -Dave -- Dave McGuire Port Charlotte, FL

On 25 Oct 2010 at 15:41, schoedel at kw.igs.net wrote: More to the point, what sense did a 32-bit machine make prior to 8- bit character codes (e.g. EBCDIC)? 24 36, 48 and 60 bit machines made much more sense with 6 bit character codes. All of the above word lengths are divisble by 6 and 4, so BCD isn't an issue either. --Chuck

On 25 Oct 2010 at 17:14, William Donzelli wrote: Is the study available for perusal somewhere? (e.g. FJCC) I'd like to see their logic. 32 bit was a terrible choice when your floating point word was 8 bit exponent and 24 bit mantissa, normalized to a 4 bit boundary, however. --Chuck

Dave McGuire wrote: Point taken. Peace... Sridhar

Fred Cisin wrote: I'm reading this from my bunker, because this whole sub-thread could get ugly, fast ;-)

On 25 Oct 2010 at 11:31, Dave McGuire wrote: Addressing width is a red herring. The S/360 addressed 8-bit bytes with a 24 bit address; the CDC 6600 addressed 60-bit words with an 18 bit (actually 17 until late in the series) address. The CDC STAR addressed at the bit level and had a 48 bit (virtual) address. Addressing bus width is pointless unless you know what was being addressed. --Chuck

Interdata as well. Likely a few more, too. -- Will

From: Richard Sent: Monday, October 25, 2010 10:13 AM In article <AANLkTikxjjuNTMWAvHX3jmXP3vcZ+OeKLirw5YyHZosT at mail.gmail.com>, William Donzelli <wdonzelli at gmail.com> writes: Yes. Up to that time, DEC's mainframes were 36-bit, and their minis were 12, 18, or 16. They phased out the 18-bit line in 1975, and demoted the 12-bit line to word processors not long after. The VAX-11/780 was DEC's entry into the so-called "supermini" market. Rich Alderson Vintage Computing Sr. Server Engineer Vulcan, Inc. 505 5th Avenue S, Suite 900 Seattle, WA 98104 mailto:RichA at vulcan.com mailto:RichA at LivingComputerMuseum.org http://www.PDPplanet.org/ http://www.LivingComputerMuseum.org/