[aprssig] spaces in object names

[Dave Skolnick]

I've spent my entire professional career on one side or the other of
specifications and standards. While the intent of the drafters is
interesting and may be useful in dealing with ambiguity that is not
the case here.

If as quoted the APRS spec says "An Object Report has a fixed
9-character Object name, which may consist of any printable ASCII
characters" then there is in fact no significant ambiguity. Since no
reference documents are listed in the APRS spec the definition of
"printable ASCII characters" is dependent on common usage. As an old
Fortran guy there is no question in my mind that a space is a
printable character, but we need not be dependent on my opinion. As
noted earlier the ANSI specification for ASCII is very clear on the
matter. A space is a printable character.

If we can agree on that, at least for purposes of discussion, the next
interesting phrase is "fixed 9-character Object name." To me, that
says the true Object name should be padded with spaces (or with
something) to fill out the nine character specification.

That may not be what the drafters MEANT but it is what they SAID.
Herein lies an important point. A specification is by its nature an
imperfect device intended to communicate effectively. Natural language
is often unclear and ambiguous. That's why people so often end up in
court arguing of contract terms.

As I noted above, the intention of the drafters is interesting but not
really relevant. The spec as written and previously amended has been
depended on by developers and users for a long time. It isn't fair to
those people to arbitrarily change something (haven't most of the 1.1
and 1.2 differences been extensions and clarifications?) that they
could reasonably depend on. Certainly Bob *can* do that but I don't
believe he *should* do that.

The fact that some people have misinterpreted the spec (not supporting
case sensitivity is a good example) doesn't justify a change. Consider
all the effort going to bring station and digipeater configurations
into compliance with the spec - the argument for compliance with the
spec loses credibility if there is a change to the written spec to
align with any one (or more) particular "as-built" application.

If it were me I would add a reference section that includes the ANSI
spec for ASCII and incorporate by reference with the caveat that in
the case of a discrepancy between the APRS spec and a reference
document the APRS spec applies.

With respect to the possible confusion between Rescue 1 (one space)
and Rescue  1 (two spaces), I suggest that it be left up to the
clients to choose how to show Object names to minimize confusion AND
to the various tactical organizations to establish their own policies.
Just because APRS supports Object names with embedded spaces doesn't
mean an organization has to take advantage of the capability.

73 es sail fast, dave KO4MI
S/V Auspicious