Stack Opens Manually But Script Thinks It's Corrupted?

[Keith Martin]

>So the central question is: Which shoud be the default, the HC model which
>alters the data or the SC model that doesn't?

Personally, I prefer the idea that a request for something without 
added instructions should get everything that's asked for, intact and 
unaltered. Adding additional elements to the read request is what I'd 
prefer to cause 'custom' behaviour in the data acquisition, such as 
massaging to suit certain formats and uses.

>But in spite of my own initial tantrum I've come to appreciate the
>flexibility, and since most of my reads are with text files I love the
>platform-independence granted by having the default option take care of line
>endings without having to think about it.

A good point. But I still feel that simpler requests should get 
simpler results - 'simple' meaning without extra things happening to 
what you're getting. When deeper involvement is needed, ask the 
engine for it. :-)

Thoughts?

k