Re: Why all the max length constraints?

From: paul c <toledobythesea_at_oohay.ac>
Date: Sun, 28 May 2006 23:16:33 GMT
Message-ID: <lbqeg.202249$7a.133700_at_pd7tw1no>


dawn wrote:
> paul c wrote:
>

>>J M Davitt wrote:
>>
>>>dawn wrote:
>>>
>>>
>>>>J M Davitt wrote:
>>>>
>>>>
>>>>>dawn wrote:
>>>>>
>>>>>
>>>>>>...
>>
>>This is a plea to stop this silly thread.

>
>
> Honest questions might seem silly until you understand. The purpose of
> asking questions is to understand better. If you think it silly, then
> you need not answer.
> ...

it's pretty clear based on the attention this stuff gets that people who are interested in seeing real progress in this field and who let the really irrelevant questions ride are doing a disservice to those who have an honest interest in it.

>

>>Also to humbly point out to
>>those born after the fact that VSAM has as much to do with the various
>>RM's as Christianity does (that OCCURS stuff was in Cobol, not VSAM, in
>>fact VSAM had no notion of fields, just bytes and offsets).

>
>
> I was going to point that out but thought it was nitpicking to do so.
> I simply redirected his point to COBOL instead. Perhaps it was not
> necessary to specify lengths when working with VSAM, but I only saw it
> done that way. ...

well, if you can't remember what you saw, i doubt if anybody else could guess. i could have been a little more accurate about VSAM, not that that would not be likely to do much good in this case. one could specify a record length in bytes which was not always strictly an offset since variable length recs were allowed, in which case the "length" was a maximum length, either 32k-1 or 64k-1, i forget. plus a length in bytes for a "key", regardless of how many fields were externally imputed to that key. regardless, i fail to see what Cobol has got to do with the RM. a comparison with assembler or even C might be apt as enough experience with those might lead one to conclude that data lengths for physical purposes could be suppressed in a UI except for when there was some logical reason for a limit. VSAM had just come out when Codd wrote his first paper but it wasn't really new, just an amalgam of previous access methods and it was also the new underpinning for IMS so i'd bet that Codd had its extremely limited physical perspective in mind when he wrote his stuff.

m Received on Mon May 29 2006 - 01:16:33 CEST

Original text of this message