The good ole days...

General discussions related to the Altair 8800 Clone

The good ole days...

Postby KenF » September 11th, 2022, 12:29 pm

I just stumbled across (again) the original announcement for the Altair in Pop Electronics, 1975. And again, I laughed at the ridiculous copy that someone used for the announcement. Without doubt, this was written by some ad employee following input from a techie, and understanding none of it. Not sure what the copyright status is, although I have seen it posted on many old websites for years, so I will just give a link to the image.

https://vintagecomputer.net/images/Popt ... 1_pg33.jpg

To start...
"...full-blown computer that can hold its own against sophisticated minicomputers now on the market." A box with no keyboard, no display, input only by binary switches and output by binary lights is competition for a PDP-11, or an IBM System 36?

"...can be economically expanded for 65000 words." (bytes, he meant) Economically? In 1976, an 8k kit was $165. Times 8 is $1344. In 1976 money. Or about $7500 bucks today. That would have definitely been the rich kid on the block with a 64k Altair. And that is not even considering the need for a full motherboard and the 16 to 25 amps needed just for the memory cards. (Actually, when the Altair was announced, I think only 4k cards were available to the average hobbyist.)

However, if you had the amount of money equivalent to the cost of a Corvette in those days, you could have "...up to 65000 routines going at the same time." This proves that the writer was not technical. I assume he got subroutines mixed up with ram size. You could definitely have 64k of subroutines, as long as they just consisted of a RET statement. Not very productive and certainly not all "going" at the same time.

In the sidebar is a few more defugalties. "Number of auxiliary registers: 8 plus stack pointer, program counter and accumulator. Actually, there are 6 registers, definitely not auxiliary, besides the two index regs and the acc.

"Number of subroutine levels: 65000." Again the nonsense statement that is getting confused with ram or something.

Any how, it is fun to look back to the old days when the hobby was brand new (and loads of fun), and in this case, with the icing on the technical cake being the phony cardboard Altair that was shown in the magazine.

Ken
(My first Altair clone was built in late 1975 and cost $625, a number that I will never forget since it was a good part of a months wages. 8080, 24k, Radio Shack cassette, horrible keyboard and Sears portable TV as monitor. By the time we began to get interested, MITS was swamped with orders and the wait time for new customers was about a century. Thus, if you wanted a system, you built it.)
KenF
 
Posts: 30
Joined: April 25th, 2022, 11:13 am

Re: The good ole days...

Postby toml_12953 » September 11th, 2022, 1:13 pm

KenF wrote:I just stumbled across (again) the original announcement for the Altair in Pop Electronics, 1975. And again, I laughed at the ridiculous copy that someone used for the announcement. Without doubt, this was written by some ad employee following input from a techie, and understanding none of it. Not sure what the copyright status is, although I have seen it posted on many old websites for years, so I will just give a link to the image.

https://vintagecomputer.net/images/Popt ... 1_pg33.jpg

To start...
"...full-blown computer that can hold its own against sophisticated minicomputers now on the market." A box with no keyboard, no display, input only by binary switches and output by binary lights is competition for a PDP-11, or an IBM System 36?

"...can be economically expanded for 65000 words." (bytes, he meant) Economically? In 1976, an 8k kit was $165. Times 8 is $1344. In 1976 money. Or about $7500 bucks today. That would have definitely been the rich kid on the block with a 64k Altair. And that is not even considering the need for a full motherboard and the 16 to 25 amps needed just for the memory cards. (Actually, when the Altair was announced, I think only 4k cards were available to the average hobbyist.)

However, if you had the amount of money equivalent to the cost of a Corvette in those days, you could have "...up to 65000 routines going at the same time." This proves that the writer was not technical. I assume he got subroutines mixed up with ram size. You could definitely have 64k of subroutines, as long as they just consisted of a RET statement. Not very productive and certainly not all "going" at the same time.

In the sidebar is a few more defugalties. "Number of auxiliary registers: 8 plus stack pointer, program counter and accumulator. Actually, there are 6 registers, definitely not auxiliary, besides the two index regs and the acc.

"Number of subroutine levels: 65000." Again the nonsense statement that is getting confused with ram or something.

Any how, it is fun to look back to the old days when the hobby was brand new (and loads of fun), and in this case, with the icing on the technical cake being the phony cardboard Altair that was shown in the magazine.

Ken
(My first Altair clone was built in late 1975 and cost $625, a number that I will never forget since it was a good part of a months wages. 8080, 24k, Radio Shack cassette, horrible keyboard and Sears portable TV as monitor. By the time we began to get interested, MITS was swamped with orders and the wait time for new customers was about a century. Thus, if you wanted a system, you built it.)


To clarify some points you've raised:

Once expanded, the Altair could hold its own against many of the minis available at the time. The Wang 3300, for example, used 2 4-bit ALUs (SN74181) to create an 8-bit word for processing. It was very slow but could service 4 users as shipped and up to 16 users when fully expanded.

He meant words. The word size of the 8080a is 8 bits so a word is the same as a byte on that machine. We referred to bytes as words in those days.

He did make a mistake about the number of subroutines. Each subroutine takes a minimum of four bytes: One byte for the CALL opcode, two bytes for the address and one byte for the RET opcode. If you want to actually call a subroutine, you need an additional two bytes on the stack for the return address to be saved. You could have 16K subroutines but no stack to push return addresses onto.

There are registers B, C, H, L, D, E, W, Z, SP, PC, PSW (Program Status Word or Flags register), there are no separate index registers. H and L were used for that purpose. Registers W and Z are not accessible by the programmer.

The reason for the phony cardboard mockup on the cover was because the only working Altair was shipped to PE's offices but was lost in transit when the shipping company drivers went out on strike. The article had to be finished by a certain date
and there wasn't enough time to build and ship another one. The lost original Altair was never found.

It was an exciting time. If you had a problem with Altair BASIC, you could actually call Bill Gates or Paul Allen to help you debug it.
toml_12953
 
Posts: 297
Joined: June 7th, 2013, 12:54 pm

Re: The good ole days...

Postby BillO » September 11th, 2022, 2:13 pm

toml_12953 wrote:He did make a mistake about the number of subroutines. Each subroutine takes a minimum of four bytes: One byte for the CALL opcode, two bytes for the address and one byte for the RET opcode. If you want to actually call a subroutine, you need an additional two bytes on the stack for the return address to be saved. You could have 16K subroutines but no stack to push return addresses onto.
You do have to count the return addresses as the call instruction will push the return address onto the stack whether you like it or not reducing the maximum to 10,922.

He also made an error in reporting the add subtract times. The add/sub instructions take 4 or 7 machine cycles so that adding or subtracting will take 8us or 14us, not 2us. That's assuming in-line calculations and not adding two numbers from memory and storing the result.
BillO
 
Posts: 136
Joined: November 11th, 2020, 6:29 am

Re: The good ole days...

Postby KenF » September 11th, 2022, 3:24 pm

I have lots of old original books and mags from the 8080 era, and several have different tales about the cardboard Altair. Mostly about being lost in shipping caused by the strike (the Wikipedia version) or supposedly that was the week that Railway Express Agency went out of business and the freight cars languished on a siding for a year and locked up by the bankruptcy court, or it was logged into the docks at Mits, but nothing was there, etc.

It is like the tale of why IBM abandoned talks with Gary Kidall for the use of CP/M. Missed appointment, airplane flying, refusal to sign non-disclosure, etc. We will never know for sure.

But I would sure like to suddenly find that serial number 0000001 Altair in some dank and abandoned warehouse. That would complete my collection and then some. (Although, in reality, I guess it would still belong to someone else.)

Ken
KenF
 
Posts: 30
Joined: April 25th, 2022, 11:13 am

Re: The good ole days...

Postby KenF » September 11th, 2022, 3:41 pm

toml_12953 wrote:
He did make a mistake about the number of subroutines. Each subroutine takes a minimum of four bytes: One byte for the CALL opcode, two bytes for the address and one byte for the RET opcode. If you want to actually call a subroutine, you need an additional two bytes on the stack for the return address to be saved. You could have 16K subroutines but no stack to push return addresses onto.


If you will allow me to use self-modifying code, then I can call 64k of RET statements with a tiny nucleus and stack. (The shade of my old instructor grasps his chest with a hand and uses the other against the desk to keep himself from collapsing at my heresy of changing code on the fly.)

What I couldn't do back in the day is afford 64k, or find a place to plug in that many cards, or feed that much amperage down the MB traces without blowing them off the board. And the fans required to keep the silicon from becoming liquid would have made the thing fly.

Ken
KenF
 
Posts: 30
Joined: April 25th, 2022, 11:13 am


Return to General Discussions

Who is online

Users browsing this forum: No registered users and 19 guests

cron