[apt4ssx8] Re: update to my Google site
- From: "Kent A. Reed" <kentallanreed@xxxxxxxxx>
- To: apt4ssx8@xxxxxxxxxxxxx
- Date: Wed, 06 Jun 2012 10:40:03 -0400
On 6/6/2012 5:20 AM, Dave Caroline wrote:
On Wed, Jun 6, 2012 at 5:54 AM, Kent A. Reed<kentallanreed@xxxxxxxxx> wrote:
Gentle persons:
Nothing earth shattering.
I created a subtopic on APT vocabularies under APT Resources and uploaded
three text files to it. They contain the sorted lists of unique vocabulary
(aka reserved) words I extracted from the three sources identified there
along with a few summary statistics. I don't know why the sidebar directory
doesn't automatically show this subtopic when you land on the home page but
you know how to use the clicky thing. Maybe I'll get around to organizing my
work better, but don't hold your breath.
What it looks like with a visual diif tool called meld
http://www.archivist.info/cnc/meld_apt_vocab.png
Thanks, Dave. You know all the cool tools:-) This is much better than
the tabular exercise I had in mind.
I used the traditional Unix command-line utilities to strip blanks,
sort, delete duplicates, and compare. Apparently, you can take the man
out of the 20th Century but you can't take the 20th Century out of the man!
CL seems to become CIRLIN in later versions
I wonder if stuff like that is documented anywhere
<...>
As to the APT360 source code, I found a backdoor in it that allows for
tricky use of the IBM System/360 Linkage Editor to preload variables using
NAMELIST statements "in the FORTRAN Library" (this phrase came from a
comment in the code). Normally this backdoor is shut. Obviously Brent's port
of the code to C works without it being open, which is a good thing because
this feature isn't documented well, er, at all. The issue here isn't the use
of NAMELIST, which is a well-known IBM FORTRAN IV feature. The issue is that
we have no idea what variables might have been predefined and stored away
"in the FORTRAN Library" and at what stage in the proceedings the backdoor
would be opened.
Would they have stuffed anything in with assembler?
Dave Caroline
Well, the backdoor itself is a combination of FORTRAN and assembler
routines. I haven't noticed any predefined values in passing, but I
expect to start looking for them in earnest this week.
Regards,
Kent
Other related posts:
