Not really a bug in MacClade 3.04

David Maddison dmaddsn at CCIT.ARIZONA.EDU
Thu Mar 17 16:03:55 CST 1994


A report was just posted about what was claimed to be a small bug
in MacClade.  The behavior is as follows:

Define state names as follows:

state   name
0       2
1       1
2       0

There are two commands that control the display/interpretation of
these things in MacClade's editor.  If "interpret state names"
is turned on in the Entry Interpretation dialog box (see book p. 180),
then any time MacClade attempts to interpret an entry (which
can happen if, for example, one tabs between cells), it first checks
to see if the entry is a state name.  E.g. if 2 is in the cell, then
MacClade will interpret it as the state named "2", that is, state 0.
This gets messy if Display State Names is turned off, in which case
a 2 displayed in a cell represents state number 2.  In this case,
MacClade will alter the data as you move between cells.

However, this is not really a bug.  In most circumstances (that is,
whenever you don't name your states in this unusual way), this
is the best way for MacClade to behave.  If anything, the problem is
that MacClade doesn't police the user enough to check to see if
they are doing something dangerous.

There are three possible solutions:
   1. don't give names to states that are equivalent to state symbols or
numbers
   2. If "Display State Names" is turned off, make sure that "Interpret
State Names" is also turned off.  (We could change MacClade so that it
polices you about this, and turns things on and off behind the scenes,
but this gets messy in other areas.)
   3. We could change MacClade so that it automatically scans your list
of state names and warns you if a state name is identical to a
state symbol or state number, and it gives a warning whenever there
could be some messy situation (e.g. Display State Names is turned off,
but Interpret State Names is turned on).  This we will do eventually;
it is on the list of things to do.

I apologize for the fact that this aspect of MacClade's behavior was
not clear, and that the book did not provide adequate warning about this.

We would appreciate it if users would contact us about apparent
bugs in MacClade before posting it on the net, as we may be able to
clarify the situation.  The appropriate e-mail address for this is

           clade at arizona.edu

(By the way, when we get our act together, there will be an offical
MacClade gopher site and World Wide Web site, both of which will contain,
among other things, the Offical Bug List, with all known bugs of MacClade
3.0 and later.)





===================================================================
David R. Maddison                PHONE: (602) 621 9781
Department of Entomology         FAX: (602) 621 1150
University of Arizona            Internet: dmaddsn at ccit.arizona.edu
Tucson, AZ   85721
===================================================================




More information about the Taxacom mailing list