LYNX-DEV Re: chartrans, 'raw mode' saveable

[Klaus Weide]

On Sun, 2 Mar 1997, Hynek Med wrote:
> On Sat, 1 Mar 1997, Klaus Weide wrote:
> 
> > Why I don't make 'raw mode' saveable -
> > 
> > frankly, because I don't know how it works.
> > 
> > Each time I looked at the code in LYCharSets.c and LYOptions.c and
> > LYReadCFG.c etc. I finally gave up with a headache.  Those variables
> > LYRawMode, LYDefaultRawMode, LYUseRawModeDefault are just not compatible
> > with my brain.  They look too similar, I cannot tell them apart.  
> > There you have my confession.
>[...] 
> When I see these, I'm not wondering why you ended up with a headache. :-) 

My variable names are probably not any better to anybody else...

> > I am not sure what the function of 'raw mode' in a fully charset
> > translating Lynx should be.  Maybe it would not be not needed at all.
> > What is it supposed to do?  (Assume I am viewing a document labelled
> > as iso-8859-5, while Display Character Set is KOI8-R.  What's the
> > meaning of toggling 'raw mode' or '@'?)
> 
> As I wrote in my previous message - what you encounter is unmarked
> documents, thus treated as ISO-8859-1, but these documents are in KOI8-R,
> ISO-8859-2, and what-not charsets. When you don't have the raw mode, these
> get faultly translated from ISO-8859-1 to KOI8-R, whan you have it, lynx
> leaves them untouched, which is fine, because your display is set for
> KOI8-R. 
> 
> > The current behaviour (with chartrans enabled) is probably
> > inconsistent, because I tried to mimick the existing code, but I feel
> > the semantics of 'raw' don't translate well when you move from (a) a
> > Lynx which can either interpret characters as iso-8859-1 or leave them
> > untouched to (b) a Lynx which can (theoretically) translate any
> > incoming 8bit charset to any Display Character Set.  If we can agree
> > on a functional definition of '@', I can try to make the code do what
> > we want.
> > 
> Well, for the reasons I wrote, there's a need for something that would let
> us treat unmarked documents as documents of the charset they are in fact
> in, and not as ISO-8859-2. Sorry to say, there are way many of these. I
> have used the -raw or the '@' key for it, because it made lynx make no
> translation. When you have the -assume_charset switch, I agree that -raw
> is no longer necessary for everything-translating lynx. Maybe its easier
> to make saveable the -assume_charset family?

Yes, they probably should be saveable.  It just wasn't a high priority
for me (when I didn't even know whether they worked correctly in typical
cases).  And then we come back to the Options screen size problem.
(Hey there's a totally empty line at the top, above "E)ditor".  Maybe I
should steal that...

  Klaus

;
; To UNSUBSCRIBE:  Send a mail message to address@hidden
;                  with "unsubscribe lynx-dev" (without the
;                  quotation marks) on a line by itself.
;