6 Aug
2020
6 Aug
'20
2:01 p.m.
On 8/6/2020 1:07 PM, Stan Sieler via cctalk wrote:
much of the slowdown was due to the GUI nature of
the PC
program they switched to, and they no longer had the luxury of having
relatively purpose-related hardware on the L9000.
Most people have never seen real
heads down data entry in action. We had
a lady who could do it, and I wrote a custom program for her to get data
off of inventory cards.
She did two runs of all of the inventory of Western Digital, at the time
had progressed to the early disk controller chip time of its history,
about 5000 tags.? Took her about an hour to do two passes entering via
keypad on a Microdata Reality.
4 fields / card, numeric.? It wasn't even a big deal, she had worked in
shops with 8 hours shifts doing such.
The inventory program they used via timeshare remote on our system had
an estimate of 5 people working the entire day (a Saturday) after
pulling the cards.? As it was they paid I think $200 for the job, and
had it done in half a day with the edit pass.? That wasn't possible time
wise with a data entry screen.
Systems on mainframes with 3270s could do the data entry on gui screens,
and some systems on Reality I saw had gotten the interfaces fast enough
to allow such.
When the PC came out we held on as long as we did because such as 123
and whatever input sucked so bad that people would have none of it.
CP/M had some spreadsheets and MRP systems and the like but they didn't
scale well.
thanks
Jim