Re: lynx-dev problem at DejaNews

[T.E.Dickey]

>       Note that the tlog stuff doesn't work with the W32 binary.  The 
> devel CHANGES log indicates that TD made the CTRACE(message) changes to 
> avoid using redirection, per se (though he hasn't yet dealt with some 
> old code that used perror() instead of TRACE clauses).  How often is the 
> W32 binary updated? 
the first cut of CTRACE changes are actually cosmetic - but do serve
a purpose:

        + they do pick up about 90% of the places where we should
          change the code to restructure it (to make the inline fprintf
          a call to a function that 'does the Right Thing'), but

        + there's a number of places where TRACE is used to wrap a block
          of code - and I didn't modify the fprintf's in those since it
          seemed redundant,

        + and the switches between stderr as stderr vs a logging file
          have to be restructured.

But I didn't do anything on that last week, since I was (for the amount of time
I had available for Lynx) trying to reduce the backlog of patches from other
people (John Davis's patches are in the right direction, but required a fair
amount of work).  But that (CTRACE etc) and tempname are both things that I 
will work
on more.  There's a couple of problems that I spotted that also have to be
resolved:
        + one of the recent changes (probably mine) leaves one of the temp
          files undeleted, and

        + turning on the trace (in all versions) changes the color behavior.

So I'm looking into those (before I get into John Davis' keymap patch).

-- 
Thomas E. Dickey
address@hidden
http://www.clark.net/pub/dickey