One of the first major computers I ever used was a TI 960B. To get the thing up and running you would have to enter a bootstrap manually into the front panel. It in turn would boot a card deck which would boot a hard drive.

As I recall other computers of that era required a manual bootstrap. The Data General and DEC series to name a couple.
So why the need to a manual bootstrap through the front panel? I used much simpler computers of the ERA like Southwest Products, KIM and PET and a couple of others that I can't recall. They all booted on power up.

So why the need for a manual bootstrap on some of the "higher" machines.

 

Many of those old computers were designed at a time before there was any such thing as ROM or EPROM. Without some form of nonvolatile electronic storage, the only way to get the computer "up and running" was, in effect, to "poke" instruction codes into core through front panel switches. That (hopefully short) sequence of instructions would execute and cause a more elaborate set of instructions to be loaded from some device such as a card reader, paper tape or mag tape. That would load the OS, and away you went.

 

Many of those old computers were designed at a time before there was any such thing as ROM or EPROM. Without some form of nonvolatile electronic storage, the only way to get the computer "up and running" was, in effect, to "poke" instruction codes into core through front panel switches. That (hopefully short) sequence of instructions would execute and cause a more elaborate set of instructions to be loaded from some device such as a card reader, paper tape or mag tape. That would load the OS, and away you went.

Sort of figure it was a ROM issue.

So you think that even though ROM was on the scene at the time, it just was not economically feasible to replace these older computers?

But PROM was invented in 1956. This was well after that. Early 80s. Hard to belive a computer would have be around before the pre PROM era.

 

I think it was largely a matter of cost. I remember some of the early fusible-link PROMS from the early 1970's; they were tiny (64 bits or something like that) and very expensive.

 

I think it was largely a matter of cost. I remember some of the early fusible-link PROMS from the early 1970's; they were tiny (64 bits or something like that) and very expensive.

For a 960B?? I can imagine how expensive they already were, especially after you added the hard drive. And to add a card reader? What could one of those cost?

 

I've no idea. I've told you what I remember, which is that the units that were available at the time were small, and expensive.

 

A card reader for the IBM 7090 was in excess of $14,000.00 IIRC. Many DEC computers of that era (ca.1965) did not even use integrated circuits. A fuse link PROM would have been out of the question on a Flip-Chip backplane. On the other hand most peripheral devices such as the paper tape reader and the DEC Tape had very sophisticated controllers that used a handful of instructions, maybe a dozen or so, to implement a boot loader that would fill up the entire core memory of say 4-32K words. My friend Jim Russo knew the RIM (Read In Mode) Loader by heart for several different DEC machines.

 

Yes, I knew the DEC PDP-8, PDP-15, and DG NOVA RIM Loaders by heart.
It wasn't a big deal to enter and in most times you didn't have to reenter the codes. The code was stored in magnetic core memories and was therefore non-volatile. When you turned the power on, the code was still present in memory.

At a later stage in the 1980's I built memory boards for DG NOVA 2. The boards had both SRAM and UV-EPROMS on board. With a flip of a switch you were able to select and run the Editor, Assembler, Debugger, BASIC, or any custom program.

 

Yes, I knew the DEC PDP-8, PDP-15, and DG NOVA RIM Loaders by heart.
It wasn't a big deal to enter and in most times you didn't have to reenter the codes. The code was stored in magnetic core memories and was therefore non-volatile. When you turned the power on, the code was still present in memory.

At a later stage in the 1980's I built memory boards for DG NOVA 2. The boards had both SRAM and UV-EPROMS on board. With a flip of a switch you were able to select and run the Editor, Assembler, Debugger, BASIC, or any custom program.

It was a time of considerable ferment and innovation. The Apple ][ and the PET were actually steps backwards compared to what other areas of industry were doing, albeit for a price that would shock most individuals. Nobody could visualize the cost reduction required to sell a mass market personal computer until several somebodies kicked the door down and did it.

 

I think it was largely a matter of cost. I remember some of the early fusible-link PROMS from the early 1970's; they were tiny (64 bits or something like that) and very expensive.

The in-between step was to replace manually entering the boot program was with a diode ROM. Lots of diodes.

 

It was a time when people were cheap and hardware was expensive -- comparatively speaking.

 

It was a time when people were cheap and hardware was expensive -- comparatively speaking.

In the summer of 1966 I went to work for a time sharing company for $1.80/hr. Best summer job I ever had. The computer I worked on (PDP-6 with 1 MWord Bryant Drum) cost several Megabucks.

 

Yes, I knew the DEC PDP-8, PDP-15, and DG NOVA RIM Loaders by heart.
It wasn't a big deal to enter and in most times you didn't have to reenter the codes. The code was stored in magnetic core memories and was therefore non-volatile. When you turned the power on, the code was still present in memory.

At a later stage in the 1980's I built memory boards for DG NOVA 2. The boards had both SRAM and UV-EPROMS on board. With a flip of a switch you were able to select and run the Editor, Assembler, Debugger, BASIC, or any custom program.

The PDP11-02 had a boot ROM. This was probably back in the early to mid 80s. The ROM would boot an assembler. There was a tape loader at a certain address. You called that address and if you were real lucky the code and database would be loaded into memory and the code would run. But often you would get a crash and start the whole process over again.

I had one system that had magnetic core. I knew the starting address of the code. It it crashed or there was a power failure, all I had to do was call that address and the system started back up. I loved that particular installation.

 

I have been fortunate enough to work with an repair PDP8 computer. Wihiisone of th oldies you describe Mostly I would just tak it about andcarfully clean it out I remember Alcatel buying an emulate (ran off a 386) to take advantage of the hard disk an networking features Our IT tried to block the purchase say they didm't support 386's, Nut the CEO of the plant informed them otherwise.

 

I’m thinking a lot had to do with the target market. A lot of those early machines were directed at developers and hobbyists who were exploring the machine as opposed to running applications. My first machine was a Rockwell AIM, and I plugged whatever socket rom in, depending on that days project. Even today, I have my main machine setup for multi boot, for working on ‘computors’. If I want to run a specific program, I go to another general machine where the ‘boot’ is irrelevant.