[CAP] NOAA Implementation of CAP
Matt Hoffman
matthoffman at acm.org
Wed Mar 21 12:02:46 PDT 2007
Bob,
I look forward to seeing the updated feeds. I've done a couple of
implementations of software that pulls CAP feeds from NOAA, and it has
always required a separate parser for NOAA feeds than for other CAP
messaging. The NOAA feed gives the impression of CAP-being-used-as-RSS.
I could give quite a bit of feedback on using the feed from my
experience on the programming level, if there's interest.
Matt
On Wed, 21 Mar 2007 13:36:12 -0400, "Bob Bunge" <Robert.Bunge at noaa.gov>
said:
> Bill and Barry,
>
> After wearing a "CAP education" hat here at NWS for the past several
> years... meaning I've spent anytime I have extra educating a traditional
> (and operational) govt organization who has done business a certain way
> for 50+ years to the benefits of CAP and other modern data formats,
> we've seen some of recent movement.
>
> For several months, I've been awaiting for a window for a programming
> resource become available to finally revisit our experimental CAP feeds.
> We hope to put out for comment some proposed changes in the coming
> weeks. I will be posting a URL for this on this list when we get there.
>
> Since 2004, we've come a little ways in being able to better parse the
> traditional NWS products into CAP fields, but there are some we won't be
> able to do given the current text formats we are working with. CAP 1.1
> solved a couple minor issues we had (specially with the polygons) as
> well.
>
> A couple of partners have provided some real nice examples of what
> they'd like to see. This has been a real benefit for us, else we are
> just guessing, guessing, guessing as to what you want. If you do have
> suggestions, please provide to me directly. Please keep in mind we have
> to work within the paradigm of the existing NWS text products.
>
> Hope that helps.
>
> Bob Bunge
> NWS
>
> Barry Britnell wrote:
> > Bill,
> >
> > I wrote a couple of applications a couple of years ago using the NOAA CAP
> > files and ran into the same issues. I had some brief email conversations
> > with Bob Bunge who works with NOAA in Silver Springs, MD. I have included
> > part of the conversation below. As you can tell, this email thread is a bit
> > dated (August 2004), but it will give you a little more background to the
> > problem you have discussed.
> >
> > The last time I checked, NWS was still using a "non-standard" approach to
> > the CAP implementation, but I could be wrong.
> >
> > Barry Britnell
> >
> > --------------------------------
> >
> > Bob,
> >
> > I would like to echo Art's comments about NWS's use of CAP. As you know I
> > have contacted you a couple of times concerning some idiosyncrasies in your
> > implementation. And, you have always been extremely open and willing to
> > suggestions. Yes, it is "experimental", but we appreciate your efforts. I
> > have been writing a couple of programs that use both the "state" and
> > "national" versions. Anyway...good work (and let us know what we can do to
> > help!)
> >
> > However, I have more of a general question to the whole group...it is my
> > understanding that NWS is using the <info> tags properly. Am I correct in
> > this assumption? There was a comment earlier that seem to indicate that
> > individual alerts should be limited to one CAP file a piece. It seems that
> > multiple alerts can be placed within a single file by using the <info> tags
> > (perhaps we need to think of a better word than "info"?). The XML schema
> > supports this.
> >
> > One other question: What is the general "color scheme" that y'all are using
> > to represent the five different levels of "urgency", "severity"
> > and "certainty"? Is it blue, green, yellow, orange and red?
> >
> > Like all of us, I am still learning exactly the best way to use this and
> > consequently, might not be correct in my assumptions. Good to hear
> > everyone's comments.
> >
> > Barry Britnell
> >
> > -----Original Message-----
> > From: usa-wg-admin at lists.incident.com
> > [mailto:usa-wg-admin at lists.incident.com] On Behalf Of Robert Bunge
> > Sent: Tuesday, July 13, 2004 6:29 PM
> > To: usa-wg at lists.incident.com
> > Subject: Re: [USA-WG] Latitude/Longitude Field(s)
> >
> > Thanks for the encouragement, Art :-)
> >
> > We are impressed that our little NWS CAP experiment is generating some
> > comments, thoughts and ideas!
> >
> > Some thoughts, notes and background for the NWS CAP experiment:
> >
> > * The CAP files were an afterthought add-on. The state files were made to
> > fill a requirement for state emergency management agencies to allow us to
> > shut down and older system. The RSS feed was the real XML target.
> >
> > We "discovered" at the end of the process and said "let's try this and
> >
> > learn." Indeed, at least one state EM is doing something with the CAP
> > files;
> >
> > * We learned. We learned that NWS backend systems need to output CAP in
> >
> > the first place. This is a major paradiem shift for NWS and is a couple
> >
> > of years off at the best;
> >
> > * Since we were at the end of the development cycle, if there had been a
> >
> > requirement that lat/long data not be in the description, we likely would
> > have not deployed CAP at all since it would have taken resources that
> > weren't available;
> >
> > * We did explore parsing the lat/long data into the Polygon field.
> > Asides from the resources to do the programming, there are a few issues,
> >
> > including the true value of the data for some of the products (the lat/lons
> > often just follow county lines), when the geocode data achives
> > a similar result;
> >
> > * As it is, CAP is ideal for a stream data over a serial type device.
> > Perhaps someday, we'll see NWS alerts coming over the NOAAPort Satellite
> >
> > feed in CAP format :-). In a file system environment, managing hundreds
> >
> > (thousands?) of alerts as individual files is not only a challenge, but can
> > result in a huge system impact when (hundreds/thousands of) users request
> > directory listings every few seconds, looking for new files. NWS
> >
> > has had systems die a hard death by this effect. The answer, of course,
> >
> > is a index list. I'd LOVE to see this group tackle an index/xml format and
> > get it worked into the standard;
> >
> > Hate it as you might, the us.cap file is by far the most popular of the cap
> > files. While some would like a county level feed, it appears most want the
> > entire county.
> >
> > This is an experiment and a chance for NWS to learn about XML and CAP.
> > It's also a chance for the user/partner community to let NWS know if XML/CAP
> > a step in a good direction or not.
> >
> > Other notes: NWS has announced two other XML services for forecasts and
> >
> > current observations that can be explored at http://weather.gov/xml/
> >
> > We hope to have an rss feed up at that location this week to announce
> > changes to the NWS xml services. If you are using the NWS CAP files, you
> > may want to subscribe. :-)
> >
> > Cheers,
> >
> > Bob Bunge
> > NWS
> >
> > Art Botterell wrote:
> >
> >> Of course, we don't want to discourage the NWS folks, who deserve a
> > lot
> >> of credit, IMHO, for taking the initiative to experiment with CAP
> > within
> >> their existing program constraints.
> >>
> >> That said, clearly we're identifying some things that would be easier
> > if
> >> the data was encoded directly to XML instead of being "backed in" from
> >
> >> the existing freeform text format. Hopefully the current experimental
> >
> >> application will be deemed a success and NWS will commit more
> > resources
> >> toward refining their feed in the future.
> >>
> >> In the meantime, if anyone needs some test data with the NWS boxes
> >> translated into the <polygon> element, and with one warning per
> > message,
> >> the weather products on California's EDIS feed
> >> (<http://www.edis.ca.gov/cap_1.0/>) are done that way, albeit for
> >> California only.
> >>
> >> - Art
> >> _______________________________________________
> >> USA-WG mailing list
> >> USA-WG at lists.incident.com
> >> http://www.incident.com/mailman/listinfo/usa-wg
> > _______________________________________________
> > USA-WG mailing list
> > USA-WG at lists.incident.com
> > http://www.incident.com/mailman/listinfo/usa-wg
> > _______________________________________________
> > USA-WG mailing list
> > USA-WG at lists.incident.com
> > http://www.incident.com/mailman/listinfo/usa-wg
> > _______________________________________________
> > This list is for public discussion of the Common Alerting Protocol. This list is NOT part of the formal record of the OASIS Emergency Management TC. Comments for the OASIS record should be posted using the form at http://www.oasis-open.org/committees/comments/form.php?wg_abbrev=emergency
> > CAP-list mailing list
> > CAP-list at lists.incident.com
> > http://eastpac.incident.com/mailman/listinfo/cap-list
> >
> > This list is not for announcements, advertising or advocacy of any particular program or product other than the CAP itself.
> _______________________________________________
> This list is for public discussion of the Common Alerting Protocol. This
> list is NOT part of the formal record of the OASIS Emergency Management
> TC. Comments for the OASIS record should be posted using the form at
> http://www.oasis-open.org/committees/comments/form.php?wg_abbrev=emergency
> CAP-list mailing list
> CAP-list at lists.incident.com
> http://eastpac.incident.com/mailman/listinfo/cap-list
>
> This list is not for announcements, advertising or advocacy of any
> particular program or product other than the CAP itself.
--
Matt Hoffman
matthoffman at acm.org
More information about the CAP-list
mailing list