The following message is a courtesy copy of an article that has been posted to bit.listserv.ibm-main,alt.folklore.computers as well.
[email protected] (Shane) writes: > So long as it wasn't an object deck. > Very early on I got into the habit of diagonally marking (the edge of) > decks with a texta. > Just in case. re: http://www.garlic.com/~lynn/2009h.html#41 Book on Poughkeepsie http://www.garlic.com/~lynn/2009h.html#42 Book on Poughkeepsie http://www.garlic.com/~lynn/2009h.html#44 Book on Poughkeepsie when science center people came out in Jan68 to install cp67 at the univ. ... all assemblies were still done on os/360 ... the card decks punched ... output "binary" from assembler placed in order in card tray with bps loader in front ... and cp67 kernel built by ipling/loading that deck from the 2540 (which would then write kernel image to disk). not all that different from punching up "stage1" os/360 sysgen deck ... assembling it under the "starter system" ... which then punched out "stage2" sysgen card deck ... which turned around and read/ran from reader. patching/fixing cp67 required re-assembly of the specific module (under os/360) and replacing the specific module "binary" in the card tray and rebooting. since there were 60+ "modules" in the cp67 kernel ... got into the habit of diagonally marking each module "binary" along with writing module name across the top edge ... similar to what is shown in picture of box of cards: http://en.wikipedia.org/wiki/Computer_programming_in_the_punch_card_era except the full cp67 "binary" was more than box of cards (2000) but less than single card tray (3000 cards). later in '68, cp67 assembly had been moved under cms ... and the cp67 kernel build was done by punching the different decks to virtual punch ... which was routed to virtual reader ... and the cp67 cards "ipl'ed" from the virtual reader. I had also started process of doing highly optimized os/360 sysgens, building os/360 2314 system packs with datasets and PDS members carefully ordered on the disk to optimize arm seek operation. Part of this was fiddling the production system and the starter system pack ... so I could run stage1 sysgen under production system. I would then run the stage2 punch card output through interpreter http://www.columbia.edu/acis/history/interpreter.html which read the punch holes and printed the corresponding alphabetic across the top. A slight problem was compared to 026/029, the print was larger than the column size ... only printing 60(?) characters across the width of card ... so full 80 characters had to be printed on two lines. I would then carefully reorder the cards in the stage2 sysgen ... so when the new system disks were built (under production system, instead of starter system) ... the order of the datasets and PDS members were optimized for seek operation. at the aug68 SHARE meeting in boston ... i gave a presentation on some of the work I had done rewritting cp67 to reduce pathlength as well as running os/360 in virtual machine ... and some comments about the optimization done for os/360 (whether running on bare iron or in virtual machine). misc posts with pieces of that share presentation: http://www.garlic.com/~lynn/94.html#18 CP/67 & OS MFT14 http://www.garlic.com/~lynn/97.html#22 Pre S/360 IBM Operating Systems? http://www.garlic.com/~lynn/97.html#28 IA64 Self Virtualizable? http://www.garlic.com/~lynn/98.html#21 Reviving the OS/360 thread (Questions about OS/360) http://www.garlic.com/~lynn/99.html#93 MVS vs HASP vs JES (was 2821) http://www.garlic.com/~lynn/2001h.html#12 checking some myths. http://www.garlic.com/~lynn/2005m.html#16 CPU time and system load One of the os/360 problems I had was periodic PTFs would replace PDS members ... which would mess up my carefully placement and PTF activity over six month period noticeably degrade thruput (I would have to carefully rebuild PDS with member ordering to restore performance). -- 40+yrs virtualization experience (since Jan68), online at home since Mar1970 ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
