About the Project => The Core Memory Project => Topic started by: fred on October 06, 2016, 01:18:21 am

Title: Just wondering
Post by: fred on October 06, 2016, 01:18:21 am
Is anybody listening in here or contributing? It seems like all the messages on the forum are quite old. Is everybody dying off or what?

I just found the site but had no way to get in on it because registration had been locked. An email to the webmaster gave me access.

My interest to look came about because my daughter asked me to write a little history of our business. I was one of the founders of SBUG which stood for Service Bureau Users Group where all of us used NCR computers. Of course I threw just about everything out from those days in the early 70s but she wanted a picture of our mainframe. Who keeps such things. Luckily this site has quite a bit of memorabilia available thanks to people who saved stuff.

I have one question that has been bugging me. When I was a Sr. Analyst I had several customer sites under my wing. Most folks were using or trying to use existing NCR software packages but my main client expected me to write most of their software custom and being that NCR included about a year of support, I was the lucky guy to do it. Problem was there was no way to get to data except for random and sequential access. I had to write my own indexed sequential access system and always wondered how did other programmers handle access. Any thoughts?


Title: Re: Just wondering
Post by: heikobuss on October 08, 2016, 07:58:56 pm
Hello Fred,

>> It seems like all the messages on the forum are quite old.

You are right - unfortunately .....

>> Is everybody dying off or what?

No, not all - I'm still alive - Thank God!

I was a Field Engineer for the Century series in Germany.

After 5 years of work I switched to the other side of the desk - from Hard- to Software and from a NCR-employee to a NCR-customer-employee.

There I hat to write a whole lot of software - mostly in NEAT/3 and NEAT/VS.

And you are right - in the pre-Terminal-ages the changes to all the files where made in sequential acces - via the "Father-Son-technique" (I hope, this is the correct english word for that).

Of course I wrote my own indexed-access-routines - I guess, everybody does this. But with the restricted space on a 655-Disk (4,5 MB!) one had not too much of possibilities!

Then came the Terminal-Dialog-Ages. In germany whe had  from NCR Augsburg a software package named "TECOS" - you could write dialog programs with "Send/Reveice"-Commands. And a package named "MIDAS" - this was a small clone of the well known Database-Package "TOTAL" from Cincom Systems. With "MIDAS" you had a set of random commands to retrieve/change/insert data - very comfortable for that times.

Later on on the Criterion Mainframe there were indexed acces via the "CAM"-Method of access.

Good memories - yes, times have changed ....

All the best

Title: Re: Just wondering
Post by: JJW on October 31, 2016, 06:14:48 pm
I'm sorry I didn't discover this web site much earlier. I had done some searches for old NCR information but for some reason this site never popped up until now. An introduction:

I worked at the NCR Electronics Division in Hawthorn, CA and later in Rancho Bernardo, CA from 1964 until 1984 where the 315, Century and Centurion computers were designed and built. I was a programmer on several projects. Most notably, I designed and built the B2 operating system (with one other programmer). worked closely with the designers of the B3 and B4 operating systems. I was also the chief technical architect for the VRX operating system.

-- JJ Whelan

Title: Re: Just wondering
Post by: JJW on November 01, 2016, 06:53:35 am
Heiko, It wasn't until VRX (Centurion not Century though they had almost identical instruction sets and API's) that an indexed sequential access method was provided as part of the standard software. That was available on disks, not CRAM which was pretty much discontinued with the Century series.

Title: Re: Just wondering
Post by: JimT on December 06, 2016, 02:16:27 am
JJ - I do believe that we worked together.  I was in hardware from 1964-1986, then I transferred to Manufacturing.  I worked on the CRAM 2, Century 50/100/150, Criterion series (V and I), Sumitomo memories (Century and Criterion), I9300/9400, and the I10000 series.  B/T/W, they did continue the CRAM into the Century series as the 653.  I don't recall when it was discontinued as I was no longer working on it.

Title: Re: Just wondering
Post by: heikobuss on August 26, 2018, 02:50:20 pm

Hello JJW,

i can't remember any possibility on the Century under B1/B2/B3 to have index-sequential access to data files using NEAT/3. How could that work? Can you explain?
When using "Cobol74" there was a in-built index-sequential access.
As written, on VRX with NEAT/VS you can use "CAM"-Files.

Hello JimT,

yes, the CRAM was available on the Century. I remember during the C-200 education at the ITEC London (North Circular Road) in 1972 glancing stunned at this strange peripheral. I've never seen a CRAM on a Century in the field -  thanks God!


Title: Re: Just wondering
Post by: Herb Fish on November 02, 2018, 12:13:06 am
Hi JJ,

I remember talking to you several times in R.B.  I think the first time was when I ran a DynaProbe on a large Century 200 site running a B2 partition under B3 at the United Bank in Denver in the early '70s.  My analysis showed that a large portion of the processor time was spent in the B2 OCD.  You became involved and made significant efficiency improvements in the OCD based on that analysis as I recall.  I also spent quite a bit of time working with Berry Gilmore's Systems Services team in Rancho Bernardo.

Herb Fish