Colin,

I don't really think you mean what you're saying ;-) Noone is arguing
against the abstraction, what what we are discussing is what the defaults
for that abstraction should be. And here, it seems to me (looking at MCFile
for example) that you do in fact prefer a smart, auto-translating,
default-to-platform handling approach over a dumb, make-line-ends-garbage,
default-to-CR approach ;-)

Cheers,
  - Andreas

> -----Original Message-----
> From: squeak-dev-bounces at lists.squeakfoundation.org 
> [mailto:squeak-dev-bounces at lists.squeakfoundation.org] On 
> Behalf Of Colin Putney
> Sent: Sunday, August 03, 2003 3:12 AM
> To: The general-purpose Squeak developers list
> Subject: Re: Extending FileList with CrLf
> 
> 
> 
> On Saturday, August 2, 2003, at 04:23  PM, Daniel Vainsencher wrote:
> 
> > I have stated before that if I am the only one that 
> believes so, I will
> > yield, but it seems there is still quite a bit of 
> misunderstanding of
> > what the issue is about, so I still have hope that my point of view 
> > will
> > be actually understood. Heck, maybe even accepted. BTW, at least two
> > others have at some point or another expressed some sympathy to my
> > arguments -
> 
> Not wanting to added to the chaos, I've kept out of this 
> thread, but I 
> do want to chime in with support for Daniel's position lest it seem 
> that he is a lone maniac, howling in the wilderness.
> 
> Over the last year or so, Avi, Julian and I have had a whole 
> series of 
> problems dealing with line ending conventions in DVS and Monticello, 
> which have some fairly specific and non-standard requirements. The 
> solution we finally hit upon involved utterly violating the 
> encapsulation of CrLfFileStream in order to make it do what 
> we wanted. 
> The kind of abstractions that Daniel is talking about would 
> have spared 
> us much grief.
> 
> Colin
> 
>