Re: CoBOL moved to OO
From: Judson McClendon (judmc_at_sunvaley0.com)
Date: 01/02/04
- Next message: Howard Brazee: "Re: (not entirely...) OT: OPINION... chicken entrails, runic stones, and crystal balls... WAS CoBOL moved to OO"
- Previous message: Howard Brazee: "Re: (not entirely...) OT: OPINION... chicken entrails, runic stones, and crystal balls... WAS CoBOL moved to OO"
- In reply to: Joe Zitzelberger: "Re: CoBOL moved to OO"
- Next in thread: Howard Brazee: "Re: CoBOL moved to OO"
- Reply: Howard Brazee: "Re: CoBOL moved to OO"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Fri, 02 Jan 2004 16:15:24 GMT
"Joe Zitzelberger" <joe_zitzelberger@nospam.com> wrote:
>
> Why do you assume the 50% of new implementations is all "OO"? There are
> plenty of new systems being developed in plain procedural styles even
> yet. Do they not figure into that number?
Yes, the 50% that work. ;-)
I don't say all of those failures are due to OO, but primarily due to the
almost exclusive use of three relatively new technologies for most new
development: GUI, Client/Server, and OO. In all three, complexity is
significantly increased, development costs are much higher, and in the
case of the first two at least, reliability is considerably poorer. Beyond
that, the reasoning goes:
A. Traditional methods applied toward new system development have
historically never shown even remotely such a failure rate industry wide.
B. Virtually all movement in the industry is toward the three technologies
above, nearly all education and media presentations in the last decade are
oriented that way, resulting in a mindset, not only that these are the 'newer'
technologies,but that they 'are the' technologies. The result is a significant
majority of new systems employ at least one or two of those technologies.
C. Any time there is a change, it is clearly obvious that the primary
suspect(s) should be the thing(s) that changed. Given A and B, it is
also clearly obvious that those are the primary suspects here.
D. We resolve the issue by examining system development overhead
(cost, time, etc) for industry wide (or as close as you can get) projects
using them compared to similar projects using traditional technologies.
E. The only such statistics I have seen are negative, that the newer
technologies are costlier, take longer, are more fragile, and more
complex and difficult to apply, in a given situation.
My whole point in these posts is to express my dismay that we (our
industry) have spent billions of dollars and man-hours, over more
than the last decade, on technologies that have serious disadvantages.
It can be argued that GUI and client/server provide capabilities not
available using traditional technologies. But I don't see that as a
significant argument in the case of OO, other than in limited areas.
And for these tools our industry is hell bent on abolishing the older
development tools, character based interfaces, procedural methods
and mainframes.
I'm simply saying that we should use the best tools for the job, be it
old or new, and that a new tool almost never completely obviate the
need for the old tool, it merely reduces the need.
-- Judson McClendon judmc@sunvaley0.com (remove zero) Sun Valley Systems http://sunvaley.com "For God so loved the world that He gave His only begotten Son, that whoever believes in Him should not perish but have everlasting life."
- Next message: Howard Brazee: "Re: (not entirely...) OT: OPINION... chicken entrails, runic stones, and crystal balls... WAS CoBOL moved to OO"
- Previous message: Howard Brazee: "Re: (not entirely...) OT: OPINION... chicken entrails, runic stones, and crystal balls... WAS CoBOL moved to OO"
- In reply to: Joe Zitzelberger: "Re: CoBOL moved to OO"
- Next in thread: Howard Brazee: "Re: CoBOL moved to OO"
- Reply: Howard Brazee: "Re: CoBOL moved to OO"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
|
