Re: Saving CoBOL

<docdwarf@[EMAIL PROTECTED]
> wrote in message
news:fven7l$eb5$1@[EMAIL PROTECTED]
> In article <m4-dndWCMZ2C74fVnZ2dnUVZ_gWdnZ2d@[EMAIL PROTECTED]
>,
> Rick Smith <ricksmith@[EMAIL PROTECTED]
> wrote:
>
> [snip]
>
>><
>>http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=266128

>> >
>
> [snip]
>
>>Then a Cobol programmer looked at the job. "He extracted that into
>>a flat file and used an old-fa****oned sort and was able to process the
>>entire thing within a 20-minute period. I was blown away," Crego says.
>
> This, I would argue, is not an advantage of the language but of learning
> how to see data-flows in different ways.  On my current site I came
across
> a few folks who were amused/worried by another team's difficulty in
> loading data into Oracle tables; it seemed like the daily
> transaction-load-and-update took 32 hours.
>
> I said 'Do you want to see someone look uncomfortable? Just for laughs,
in
> the next meeting ask if they have sorted the input data into their
> table-key sequence, de-duping it for later research... oh, and ask if
> they've turned journalling off and whether they're using a single-row or
> bulk-load utility.'
>
> The response was a kind of blank-faced 'whuh?' but they did it... and
> re****ted back that the database team liason (an Anderoid... errrr...
> Accenturion) got all fluster-faced and took notes and said he'd look
into
> that... and suddenly there weren't day-and-a-half windows needed for
daily
> loads.
>
> DD
>
Good job, Doc.

I haven't worked with Accenturions but I sure had my fill of 
Anderoids...back in their heyday.

I agree entirely with your point about seeing dataflows in different ways.

I would be quick to admit that a background in COBOL is good grounding for

this.

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