Re: Possible bug in finding CR's

[Harry Binswanger hb@xxxxxxxx]

At 08:36 AM 5/27/96 -0400, you wrote:
>Harry:
>
>I can't figure out why you consider *@r etc. a bug. When I ran across
>the same problem I assumed the second two bytes was an EOL, and would
>have been frustrated if it didn't work. I would call the problem you
>describe a deficiency in documentation.
>
Nathan:
  Well, it's an issue of degrees. Best solution: allow as-R as-F to find
what it would be assumed to find (cr,lf). Second best solution: whenever a
user puts that sequence up on the command line, the pgm should automatically
convert the two symbols into the left-arrow, showing the user thereby what
is going on. Third best solution: make the user look it up in the manual's
index. Fourth best solution, have it unindexed in the manual. Fifth best
solution (the actual situation): no solution. So I consider it a bug .
    The line ending *is* cr-lf. You can search for cr and you can search
for lf. Why, then, can't you search for the pair, just as you can search for
the letter combination "ab" by combining "a" and "b."
    I really appreciate it when programs take into account what a user
would rationally expect, and provide same, even when the underlying code has
to do a few tricks to get that result.
    Regards,

     Harry Binswanger
      hb@xxxxxxxx