lynx-dev Re: 'reloading' lynx.cfg

[Michael Warner]

On or about 23 Jul, 1999, Klaus Weide
<address@hidden> wrote:

> On Fri, 23 Jul 1999, Michael Warner wrote:

        [...]

> Actually, IIRC I first saw this after correcting '#ifdef
> PERSISTENT_COOKIES' in reload_read_cfg to '#ifdef
> EXP_PERSISTENT_COOKIES' in some intermediate version a while ago.
> 
> Looking closer at those sections of code now, they seem wrong.
> If it's necessary to temporarily store a way of the cookie file
> name, that should not be done as (currently)
> 
>         char * LYCookieFile_flag = LYCookieFile;
> 
> but something like
> 
>         char * LYCookieFile_flag = NULL;
>         StrAllocCopy(LYCookieFile_flag, LYCookieFile);
> 
> (and a 'FREE(LYCookieFile_flag)' at the end) is needed.
> 
> > Since I know this is happening, and the real cookie file doesn't
> > appear to be damaged, and I (think I) know how to repair the
> > damage, I can still use the reload cautiously and under limited
> > circumstances, but I think it's an undesirable (to put it mildly)
> > state of affairs for the unsuspecting user.
> 
> Could you try with the changes above?

I'll give it a shot, but be forewarned - given my rudimentary C
skills, any increase in the specifity of the instructions will
greatly increase the confidence level of the results :).

        [...]

> > Colors, howeve, are fine :)
> 
> I remember a small change only, I think it was with slang; most COLOR
> pairs were as before, maybe one or two had changed.  It may have been
> only a change in one of the A_{BOLD,UNDERLINE,REVERSE} attributes which
> sometimes get combined with colors (and can have the effect of changing
> the color on the linux console).

If the change was minor, I may just have missed it (I don't use
a very colorful scheme).  Also, might something specific to the
terminal emulator (TeraTerm, in my case) mask the effect?

-- 
Michael Warner
<address@hidden>