Re: [gpsd-dev] AIS Type 24 handling

[Eric S. Raymond]

Roy Barkas <address@hidden>:
> My suggestion is that an option be added to the gpsd command line that
> explicitly requests independent type 24 handling.

But I consider adding command-line options to be sheerest, utter
poison.  I've documented my reasons.  See the Hacking Guide section
"Don't add invocation options!".

> My suggestion is that the gpsd developers accept that (regrettably) not all
> AIS feeds work the same and incorporate a way of handling unmatched type 24
> messages.

Probably the least bad thing would be an option flag in the stream command.
 
> I'm also happy to capture and send to you any other specific packet types
> that may be of interest or are rarely seen in the wild if I can find them
> in my feed.

As it says in AIVDM.txt:

== Improving This Document ==

To avoid copyright difficulties, I rely only on freely redistributable
public documents and paraphrased reports from people who have seen the
relevant proprietary standards.  If you are such a person, please help
by reporting the following to be included in future versions of this
document:

* Sample sentences of types 16-17, 22-23, and 25-27.

* Sample sentences of type 6 and 8 conforming to <<IMO236>> and <<IMO289>>.

For verification purposes, I need the raw sentences together with
decoded dumps of their field values.  Please note that sample
sentences not accompanied by field dumps are *not* useful; I can get
plenty of those.

-- 
                <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>