Old programmers, does this seem right?

[Leros]

I'm reading a paper from the early 70s. As an example it describes a system that accepts an ordered set of lines. The system will circularly shift the words in each line, printing out all possible lines in alphabetical order. Pretty simple.

It then says that a good programmer with good supporting software could produce such a system in a few weeks. Would it really take that long to produce this system back then?

I imagine I could easily write this in assembly in a day. I can't understand why it would take so long.

 

[EagleKeeper]

What you are doing is producing an sorted display for every line combination.

Good luck in doing it in a week in assembly

Also realize, that this is not a PC that you would be working on for coding entry, but a instruction punch card or a TTY if lucky

 

[Leros]

What you are doing is producing an sorted display for every line combination.

Good luck in doing it in a week in assembly

Also realize, that this is not a PC that you would be working on for coding entry, but a instruction punch card or a TTY if lucky

True, that changes everything. Not being able to develop interactively would really slow down development and debugging.

 

[DaveSimmons]

Besides the lack of an IDE and interactive debugger, there were also serious constraints on memory and file I/O.

You might need to re-read each line every time you need it rather than buffering them all in memory, and who knows what schemes you'd need to use for storing intermediate results and sorting the output.

 

[Markbnj]

Why am I not surprised to find CC the first respondent on a thread directed specifically at 'Old Programmers'?

 

[EagleKeeper]

Why am I not surprised to find CC the first respondent on a thread directed specifically at 'Old Programmers'?

Pushing your luck there young whippersnapper:hmm:

 

[Markbnj]

Hey I'm 50 in two weeks. I should at least have earned the right to be referred to as "middle aged whipperflopper."

 

[degibson]

Hey I'm 50 in two weeks. I should at least have earned the right to be referred to as "middle aged whipperflopper."

50? 6 bits. You're not old until you hit 7.

 

[Modelworks]

The issue isn't that the code is too complex its the resources you have to do it with that make it complex. In the 70's the majority of cpu were 8 bit and you would be using something like a z80 or 8080 cpu. The OS would have been CP/M and you would probably be limited to 8-16KB ram for the entire computer. The architecture of the systems were also very different from what we have now where nothing was free and every instruction cost you, if you want to do two instructions, it could take 4 clock cycles to do it.

I feel really old when I think about my first class in digital logic. We learned it by implementing a circuit to decode the pulses from a rotary dial phone by using circuits that were pure logic gates, flip flops, counters to output the number dialed. There wasn't even a microprocessor involved. Just lots of the 74xxx parts.

 

[Leros]

The issue isn't that the code is too complex its the resources you have to do it with that make it complex. In the 70's the majority of cpu were 8 bit and you would be using something like a z80 or 8080 cpu. The OS would have been CP/M and you would probably be limited to 8-16KB ram for the entire computer. The architecture of the systems were also very different from what we have now where nothing was free and every instruction cost you, if you want to do two instructions, it could take 4 clock cycles to do it.

I feel really old when I think about my first class in digital logic. We learned it by implementing a circuit to decode the pulses from a rotary dial phone by using circuits that were pure logic gates, flip flops, counters to output the number dialed. There wasn't even a microprocessor involved. Just lots of the 74xxx parts.

I code in assembly on simple 8-bit embedded processors, so I have some idea what it is like.

 

[wirelessenabled]

True, that changes everything. Not being able to develop interactively would really slow down development and debugging.

Also keep in mind that you had no actual access to the computer. You turned your card deck in at the window. Results would come back in 2-3 days most of the time, but would stretch out to a week or so as the end of term got near and people were trying to complete all their assignments.

Big improvement when I was a senior was access to a "hot card reader". Limited to 100 cards, ie 100 lines of code including control cards etc. You put your own card deck in the reader, printer in the next room would output the results usually within an hour or two. Again time stretched out towards the end of the term to about 20-30 hours. It was always fun trying to find your printout in the room.

Always was a bummer when you didn't put the right control cards in the right sequence and got a null result.D:

 

[EagleKeeper]

Also keep in mind that you had no actual access to the computer. You turned your card deck in at the window. Results would come back in 2-3 days most of the time, but would stretch out to a week or so as the end of term got near and people were trying to complete all their assignments.

Big improvement when I was a senior was access to a "hot card reader". Limited to 100 cards, ie 100 lines of code including control cards etc. You put your own card deck in the reader, printer in the next room would output the results usually within an hour or two. Again time stretched out towards the end of the term to about 20-30 hours. It was always fun trying to find your printout in the room.

Always was a bummer when you didn't put the right control cards in the right sequence and got a null result.D:

\
I was lucky - worked at a bank while in school, so had access to a card punch.
Pick up the last output run from the school data center before heading to work and could then code/punch at work. Drop the deck off in the AM - one extra run per day as a result :thumbsup:

 

[Markbnj]

In 1975 they opened up an accessible computing center at our school with four time share terminals connected to the HP3000. We had two teletypes, two CRTs, a line printer, and an acoustic-coupling modem, though I don't know what the hell anyone used it for back then. We used to rush over there after our last class, and it was a sad day when the door was closed with a sign taped up reading "Sorry! The Princess sleeps."

 

[Cogman]

Also keep in mind that you had no actual access to the computer. You turned your card deck in at the window. Results would come back in 2-3 days most of the time, but would stretch out to a week or so as the end of term got near and people were trying to complete all their assignments.

Big improvement when I was a senior was access to a "hot card reader". Limited to 100 cards, ie 100 lines of code including control cards etc. You put your own card deck in the reader, printer in the next room would output the results usually within an hour or two. Again time stretched out towards the end of the term to about 20-30 hours. It was always fun trying to find your printout in the room.

Always was a bummer when you didn't put the right control cards in the right sequence and got a null result.D:

I have to say, I don't know if I would be into computers if I was born a bit earlier. I've always liked the thought knowing what was goes on in the black box, but when it is THAT hard to get anything done... I think I would have concluded that it was a waste of time (Because AT is a much better use of my time )

 

[Markbnj]

I have to say, I don't know if I would be into computers if I was born a bit earlier. I've always liked the thought knowing what was goes on in the black box, but when it is THAT hard to get anything done... I think I would have concluded that it was a waste of time (Because AT is a much better use of my time )

There was no AT back then, which is why we were into computers. It was a total thrill to dial into some BBS with my 1200-baud Hayes modem and see a list of messages from real people. It was like some mysterious dimension that only a select few knew how to enter. Now the online world is, if anything, worse than the real one.

 

[EagleKeeper]

There was no AT back then, which is why we were into computers. It was a total thrill to dial into some BBS with my 1200-baud Hayes modem and see a list of messages from real people. It was like some mysterious dimension that only a select few knew how to enter. Now the online world is, if anything, worse than the real one.

Now, people that are determined to be idiots and PITAs get on just for the sake of being one. does not require any skill.

Back then; it was not "cool" - one had to be a nerd.

Different worlds that did not collide.

1200Bauds were a luxury

 

[Markbnj]

Now, people that are determined to be idiots and PITAs get on just for the sake of being one. does not require any skill.

Back then; it was not "cool" - one had to be a nerd.

Different worlds that did not collide.

1200Bauds were a luxury

I remember when I upgraded to a 9600 baud Intel modem for Compuserve access. The modem cost $700 and access was $48/hr.

 

[KentState]

In 1975 they opened up an accessible computing center at our school with four time share terminals connected to the HP3000. We had two teletypes, two CRTs, a line printer, and an acoustic-coupling modem, though I don't know what the hell anyone used it for back then. We used to rush over there after our last class, and it was a sad day when the door was closed with a sign taped up reading "Sorry! The Princess sleeps."

We just moved off our HP3000 here at work. Felt sorry for the one old-timer that couldn't transition to the Windows world.

 

[Modelworks]

Now with the internet it is so much easier to learn things. I remember trying to learn assembly language from a 8 page manual that came with the assembler that was stored on a cassette tape. There was nobody to ask and nobody I could call. Most people didn't even have a computer or cared to learn about it. No books in the library on it and so I spent countless hours staring at a CRT and keyboard typing in things to try to figure it out on my own. oops, that locked up the computer, reset it, rewind the tape and load the software again, repeat, . It wasn't until a few years later that I finally found a BBS local and could ask questions.

 

[RockitMan]

The thing with main frames was that first you write pseudo code, (on paper), then you write the entire program (Every single line on paper) They you type each line on an individual card (80 chr per line NO BACKSPACE!!!). The you run your stack of cards through the reader, wait, wait, wait... Result !!! Error on line 2 !!! END. Yes it only reported the line number of the first error, not what it was or any thing else. A few weeks sounds reasonable.

 

[Leros]

The thing with main frames was that first you write pseudo code, (on paper), then you write the entire program (Every single line on paper) They you type each line on an individual card (80 chr per line NO BACKSPACE!!!). The you run your stack of cards through the reader, wait, wait, wait... Result !!! Error on line 2 !!! END. Yes it only reported the line number of the first error, not what it was or any thing else. A few weeks sounds reasonable.

*hugs his interactive debugger*

 

[bendixG15]

You all brought back memories how great the "good old days" really were.