Re: interesting use of NEXT SENTENCE vs. CONTINUE

Sorry about that; messed up in the editing. Here's a second try:

"Pete Dashwood" <dashwood@xxxxxxxxxxxxxx> wrote in message
news:3gmlfjFd8f0qU1@xxxxxxxxxxxxxxxxx
>
> "Chuck Stevens" <charles.stevens@xxxxxxxxxx> wrote in message
> news:d84fb0$1qbd$1@xxxxxxxxxxxxxxxxxxxxxxx
> >
> > "Pete Dashwood" <dashwood@xxxxxxxxxxxxxx> wrote in message
> > news:3gk48iFcotqsU1@xxxxxxxxxxxxxxxxx
> > > It isn't. And it has been dropped from the 02 standard (not that I
> really
> > > care about that...)
> >
> > Well... not exactly. Its use is *disparaged by* the '02 standard, but
> > support for it in compliant compilers is still required by that same
> > standard.
>
> An important but subtle distinction. Thanks for clarifying that Chuck. I
> understood it had been dropped form a previous post in this thread.
>

To clarify further: the 2002 standard puts NEXT SENTENCE in the "archaic"
list. The proposed 2008 COBOL standard makes no change in that status. It
can stay "archaic" through any number of revision cycles to the standard.

However, for it to disappear altogether, it has to move from "archaic" to
"obsolete" for one full revision d cycle, and only then can it be dropped
from the standard.

Presuming that the process of developing and approving a COBOL standard
seems to take around six years with no "hitches", and presuming that process
doesn't get streamlined further, the *earliest* it could move from "archaic"
to "obsolete" would be late 2014, and from "obsolete" into the sunset
sometime after 2020.

> To digress for a moment, in Odene Dayse I worked on sites where ALTER was
> used as a matter of course. It was extremely efficient and, in machines
> with limited memory, was MUCH better than using IF for a first time
switch,
> for example. With IF, the first time condition is tested for every
iteration
> of the code; with ALTER, once only...(and IF required other machine code
> overheads, which ALTER didn't.)

I've heard that it was very effective on certain architectures, though I
would not necessarily agree that limited memory was a significant criterion
in all architectures. On certain limited-memory machines ALTER is much
slower than a first-time switch.

One of the worst programs I've ever had the misfortune to attempt to debug
was a cross-reference utility for COBOL, written in COBOL by a Fortran
programmer who had just discovered the ALTER statement.

> We had conventions for using it and it wasn't just sprinkled willy nilly
> through our code. When it fell into disfavour, and, eventually, disuse,
none
> of us could figure why... We had never had a problem with understanding,
> using, or debugging it.

The use of ALTER implies the use of GO TO, and those who disparage GO TO
like ALTER even less for that reason alone. It's a bit tough to tell,
particularly in poorly-written COBOL code, exactly where "GO TO." actually
ended up GOing TO.

Put more succinctly, in two words: Blame Dijkstra!

> Installation standards should encourage a common style that all agree with
> for axiomatic reasons. When standards are used in this way they actually
> help programmers, rather than stifling their thought and creativity.

I wholeheartedly agree. One issue, though is what is "axiomatic" as a Good
Idea in one operating environment may be a performance catastrophe in
another. One implication of "axiomatic" is "taken for granted", and it
would be appropriate if the *reasons* for the rule in that particular
environment were explicitly available alongside the rule itself; "everybody
knows" is not really an appropriate justification!

-Chuck Stevens


.

Relevant Pages

  • Re: interesting use of NEXT SENTENCE vs. CONTINUE
    ... The proposed 2008 COBOL standard makes no change in that status. ... >> To digress for a moment, in Odene Dayse I worked on sites where ALTER ... branch being slower than a compare on ANY architecture; ...
    (comp.lang.cobol)
  • Re: Fiend reuntied
    ... Austin Shackles said: ... > besides, *all* standard ... You have to alter 'em. ... Prev by Date: ...
    (uk.rec.sheds)
  • Re: How to work with Remote Files on Web Host?
    ... The problem is that I created the site with Visual Basic.NET standard ... used Visual Basic.Net Standard 2003, and now I'm using the pro version, ... which means I can't alter the original files.. ... one at a time until the web site is ready to go again. ...
    (microsoft.public.dotnet.general)
  • Re: RosAsm - right click
    ... > AA> manager' program is that MWRLOAD doesn't alter the standard ... > should anything mess with the standard MBR layout, ... alter the standard lay-out of the MBR. ... or you risk losing the ability to boot from the disk. ...
    (alt.lang.asm)
  • Re: Method to force keeping of source
    ... It appears to me that your wish is that the COBOL standard should prevent ... environment for the compiler or in the execution environment. ... > implementor defined is one of the problems as I look at the situation as ...
    (comp.lang.cobol)