Re: lynx-dev -anonymous restrictions not set correctly.

[Thomas E. Dickey]

On Sun, 18 Feb 2001, Rob Cameron wrote:

> >
> >> It would also seem that every CAN_ANONYMOUS.. variable in userdefs.h
> >> should be inspected here (or elsewhere) in order that the
> >> configurations specified in userdefs.h take effect.  I can't
> >> figure out how that happens.
> > 
> >that's a slightly different matter - the effect of my change is that a lot
> >of inline assignments are done in a table to make it easier to modify, but
> >there shouldn't be (aside from bugs) any difference in behavior or how
> >userdefs.h is used.  I didn't add any CAN_xxx symbols (but having them
> >in a table does make it easier to see which restrictions had none).          
> >   
> 
> I think the use of a table to regularize these settings is a good
> idea.   With respect to the CAN_ANONYMOUS settings, I thought that
> the setting for "bibp:" should use CAN_ANONYMOUS_GOTO_BIBP from
> userdefs.h.   But I also thought parallel set ups should exist

I either didn't notice CAN_ANONYMOUS_GOTO_BIBP there, or forgot to
go back and check on that (I do recall thinking that it wasn't defined).
My error - I do see it there.

> for other schemes such as finger.  I couldn't figure out how
> the setting
> #define CAN_ANONYMOUS_GOTO_FINGER    TRUE
> is used.   Actually, when I try finger: URLs now, I get an
> unsupported scheme message whether I start lynx with -anonymous
> or not.

I may have broken something (but I should be able to test some parts of
this and repair things - very likely it's an error in the function 
that handles the table).  Most of my attention was on the couple of
problems I noticed with TRST - fixing the restrictions code shouldn't be
that hard.

-- 
T.E.Dickey <address@hidden>
http://dickey.his.com
ftp://dickey.his.com


; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to address@hidden