From: Anonymous on
In article <751l8494ejaho69uiv503vvqsvk89llncq(a)4ax.com>,
Robert <no(a)e.mail> wrote:
>On Fri, 25 Jul 2008 08:25:09 -0600, Howard Brazee <howard(a)brazee.net> wrote:

[snip]

>>So I guess I'm missing your point here. Do you mean we should spend
>>as much testing a new program design as GM spends testing a new car
>>design?
>
>If GM changed only the ash tray, would it re-test every other system to
>insure the ash
>tray didn't inadvertantly affect something else?

Mr Wagner, answering a question with a question is no answer at all... and
I believe that every year's New Model, no matter what the changes on the
preceding year's might be, gets crash-tested and evaluated.

DD

From: Michael Mattias on
"Pete Dashwood" <dashwood(a)removethis.enternet.co.nz> wrote in message
news:6evb57F8vtfcU1(a)mid.individual.net...
>
>
> "Howard Brazee" <howard(a)brazee.net> wrote in message
> news:7nlj849b8ev27htjlbnbafnonmkr43j1jd(a)4ax.com...
>> On Fri, 25 Jul 2008 11:54:34 +0000 (UTC), docdwarf(a)panix.com () wrote:
>> Sometimes it is better to get what we need today and replace it
>> tomorrow with what fits tomorrow better.
>
> Very wise observation, Howard.

Having been a businessman long before I was a 'geek,' I can agree.. but
with one observation/stipulation ..

Seems to me not a whole lot of users actually get around to making that
replacement... that is, for these users tomorrow never comes.

MCM










From: Anonymous on
In article <kLGik.8932$L_.5250(a)flpi150.ffdc.sbc.com>,
Michael Mattias <mmattias(a)talsystems.com> wrote:
>"Pete Dashwood" <dashwood(a)removethis.enternet.co.nz> wrote in message
>news:6evb57F8vtfcU1(a)mid.individual.net...
>>
>>
>> "Howard Brazee" <howard(a)brazee.net> wrote in message
>> news:7nlj849b8ev27htjlbnbafnonmkr43j1jd(a)4ax.com...
>>> On Fri, 25 Jul 2008 11:54:34 +0000 (UTC), docdwarf(a)panix.com () wrote:
>>> Sometimes it is better to get what we need today and replace it
>>> tomorrow with what fits tomorrow better.
>>
>> Very wise observation, Howard.
>
>Having been a businessman long before I was a 'geek,' I can agree.. but
>with one observation/stipulation ..
>
>Seems to me not a whole lot of users actually get around to making that
>replacement... that is, for these users tomorrow never comes.

'Just for now' may not only limited to how humans deal with business, Mr
Mattias... and there are times when taking a step back to look at a
situation results in a superior view of the solution.

Something about draining swamps and alligators, as I recall.

DD

From: HeyBub on
Pete Dashwood wrote:
>> "You, uh, SORT the master file?" I asked.
>>
>> "Sure," said the manager.
>
> Is that so terrible? :-) It could make sense. We store master data on
> RDBs in any old sequence and order it at the time it is retrieved...
>

Sure. At one time the Reader's Digest subscription list comprised 200 reels
of tape. You don't go around sorting that sort of mess for drill.


From: Pete Dashwood on


<docdwarf(a)panix.com> wrote in message news:g6fa4o$btd$1(a)reader1.panix.com...
> In article <751l8494ejaho69uiv503vvqsvk89llncq(a)4ax.com>,
> Robert <no(a)e.mail> wrote:
>>On Fri, 25 Jul 2008 08:25:09 -0600, Howard Brazee <howard(a)brazee.net>
>>wrote:
>
> [snip]
>
>>>So I guess I'm missing your point here. Do you mean we should spend
>>>as much testing a new program design as GM spends testing a new car
>>>design?
>>
>>If GM changed only the ash tray, would it re-test every other system to
>>insure the ash
>>tray didn't inadvertantly affect something else?
>
> Mr Wagner, answering a question with a question is no answer at all... and
> I believe that every year's New Model, no matter what the changes on the
> preceding year's might be, gets crash-tested and evaluated.
>
> DD
>

Despite your constant repetition of the "answering a question with a
question is no answer at all" mantra, it simply isn't true. Trotting out
cliched boilerplate as if it is Holy Writ, with the only basis for it's
validity being constant repetition, is unworthy of you.

Sometimes, answering a question with a question can evoke a whole series of
considerations which lead to a far better solution than a simple short
answer.

Robert's question is a fair one. It also causes us to realise, in the
context of the base analogy, that systems DON'T always need regression
testing. The ashtray is an isolated object that has no interaction with any
other part of the vehicle other than the fascia. It would be pointless to
rerun engine tests because the ashtray was changed.

In Object Oriented systems the same kind of separation can be attained
(encapsulation does it), and the analogy is a good one.

Pete.
--
"I used to write COBOL...now I can do anything."