Re: Relational vs network vs hierarchic databases

From: Tony Douglas <tonyisyourpal_at_netscape.net>
Date: 12 Nov 2004 02:05:01 -0800
Message-ID: <bcb8c360.0411120205.19fee055_at_posting.google.com>


guntermann_at_verizon.net (Dan) wrote in message news:<3e68f717.0411101058.1a27498_at_posting.google.com>...

<snip>  

> The interfaces and layer interactions don't seem to be as clean as the
> reference model, but the functions/services within the protcol stack
> implementation are virtually the same, just not necessarily in the
> same places or called the same things. Again, its a model vs.
> implementation issue, I think.
>

Remembering it's about 10 years since I last cared deeply about this stuff... if I remember correctly, TCP & IP wouldn't have fitted into the originally proposed OSI model pretty much at all, and the final model was fiddled to make it fit somewhat, as TCP & IP were already well and truly off and running.

Were I being truly facetious, I would ponder if this is what happens when an implementation runs away before things have been fully thought through with a resulting negative feedback to the formal spec, and if there were any analogy between this and SQL ? But, that would be facetious ;)

Received on Fri Nov 12 2004 - 11:05:01 CET

Original text of this message