Hi all,
the wcs-branch is fully aligned with trunk release.
Regards,
Alessio.
Hi,
my boss at RSI has resourced me to put some time into adding netCDF coverage support into Geoserver, the reason that it took so long since my initial enquiry is that we had to develop a baseline solution using MapServer.
Please advise on what needs to be done, and how to do it, and I will get on.
Many thanks,
Norman Barker
Hi,
I have posted again since no one replied, I am available to develop the netCDF plugin for the Geoserver WCS is you want, RSI have resourced my time. I have also written a netCDF to geotiff routine if that is of any use.
Please advise on what to do next.
Thanks,
Norman Barker
Hi Norman:
I personally think that netCDF capability for Geoserver will be
fantastic--something that is talked about on a regular basis.
Alex
Hi Norman,
Whilst I'm not the greatest expert on these matters I'll try to give you some response based on the possible use cases for this that i know about.
Others may be able to put us both right 
a) geotools needs to have data source for netCDF - this will be either Feature-centric or Grid-centric : this will depend on what has actually been packed into the netCDF structure. If you want to subset the netCDF to extract feature sets - give me all the readings for sea surface temp within this polygon in Jan 97... then you may wish to use Geoserver to translate between packed arrays and individual mappable objects - a perfectly valid use case as far as I can see.
which of course brings us to the other NetCDF issue - there are many ways of putting data into NetCDF files,
b) what NetDCF conventions are you going to support?
and of course another perfectly valid use case is to pack extracted features from any geotools supported data source into a NetCDF response format.
c) What Geoserver functions are you planning to invoke? WMS, WFS, NetCDF in or out etc - depending on this you will be working on different parts of the code base, geotools or geoserver
Of course, geotools could be an OpenDAP client - getting NetCDF data via the OpenDAP interface, just as it would via WFS
Have I thrown the cat amongst the pigeons enough here? Would help to make your ambitions a bit clearer, and I would suggest this documentation is built up on the Wiki. I'm willing to help refine the requirements and try to understand where in the code base the starting points are, but by narrowing downh exactly what the priorities are we can probably identify the real experts and get some pointers.
Cheers
Rob Atkinson
Norman Barker wrote:
Hi,
I have posted again since no one replied, I am available to develop the netCDF plugin for the Geoserver WCS is you want, RSI have resourced my time. I have also written a netCDF to geotiff routine if that is of any use.
Please advise on what to do next.
Thanks,
Norman Barker
-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, 4
opportunities to win big! Highest score wins.NEC IT Guy Games. Play to
win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
_______________________________________________
Geoserver-devel mailing list
Geoserver-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-devel
>From: Rob Atkinson <rob@anonymised.com>
>To: Norman Barker <normanbarker@anonymised.com>
>CC: geotools-devel <geotools-devel@lists.sourceforge.net>, Geoserver-devel <geoserver-devel@lists.sourceforge.net>, "Woolf, A (Andrew)" <A.Woolf@anonymised.com>
>Subject: Re: [Geoserver-devel] WCS - netCDF resource available
>Date: Tue, 03 May 2005 22:20:31 +1000
>
>Hi Norman,
>
>Whilst I'm not the greatest expert on these matters I'll try to give
>you some response based on the possible use cases for this that i
>know about.
>
>Others may be able to put us both right
>
>a) geotools needs to have data source for netCDF - this will be
>either Feature-centric or Grid-centric : this will depend on what
>has actually been packed into the netCDF structure. If you want to
>subset the netCDF to extract feature sets - give me all the readings
>for sea surface temp within this polygon in Jan 97... then you may
>wish to use Geoserver to translate between packed arrays and
>individual mappable objects - a perfectly valid use case as far as I
>can see.
>
>which of course brings us to the other NetCDF issue - there are many
>ways of putting data into NetCDF files,
>b) what NetDCF conventions are you going to support?
>
>and of course another perfectly valid use case is to pack extracted
>features from any geotools supported data source into a NetCDF
>response format.
>c) What Geoserver functions are you planning to invoke? WMS, WFS,
>NetCDF in or out etc - depending on this you will be working on
>different parts of the code base, geotools or geoserver
>
>Of course, geotools could be an OpenDAP client - getting NetCDF data
>via the OpenDAP interface, just as it would via WFS
>
>Have I thrown the cat amongst the pigeons enough here? Would help to
>make your ambitions a bit clearer, and I would suggest this
>documentation is built up on the Wiki. I'm willing to help refine
>the requirements and try to understand where in the code base the
>starting points are, but by narrowing downh exactly what the
>priorities are we can probably identify the real experts and get
>some pointers.
>
>Cheers
>Rob Atkinson
>
<snip>
Rob,
thanks for the email, I am looking to use the java-netCDF bridge from Unidata, so I will support any netCDF files that they support, in particular these java libraries do support OpenDAP so that is available.
In particular the netCDF files that are causing us a problem are netCDF files on irregular grids, so conversion to geotiff is not much of a goer since you lose this, my ideal would be to allow the user to subset a netCDF file on temporal, spatial, and parameter selection, and ideally retrieve the file in
netCDF, or a browse as an image.
I agree that getting in the use cases, and the points from domain experts is important, but in XP style I am going to get started and incorporate those comments later in a refactoring exercise :-). The Galeon project is also helping with these use cases.
Simone has kindly sent me some coding pointers, and I will look into this.
Many thanks,
Norman Barker
Norman Barker wrote:
Hi,
I have posted again since no one replied, I am available to develop the netCDF plugin for the Geoserver WCS is you want, RSI have resourced my time. I have also written a netCDF to geotiff routine if that is of any use.
Please advise on what to do next.
The next step would be to team up with the other developers madly setting up raster support for GeoServer. I know they are on a short two week deadline right now, but heck the more the merrier.
http://docs.codehaus.org/pages/viewpage.action?pageId=24821
Please read the last Geotools IRC logs (simboss provided a status update and asked for help). And then strike up an email coversation with them to figure out the technical details.
(Near as I can tell that will involve making a GridCoverage our of your netCDF file, I suspect the class you need to write is a NetCDFFormat).
Jody out.
I am so confused. I can't find this original post, but I think my email
also just turned back on. David Blasby has also been having troubles,
and is in the midst of a move. Glad I caught this, it would be tragic
if we missed new developer's resources. A few things:
1) netCDF in WCS would be great.
2) I believe Alessio and Simone are planning on netCDF, but I have no
idea about their progress.
3) They have a deadline in 2 weeks, so may be ignoring email.
I'm not sure if they need netCDF for their coming deadline, but they do
need it soon. So hopefully they'll get back to you and slot you into
action. They can advise a lot better than I, indeed the WCS is all
their work.
If you want to get started without them, then look into the grid
coverage api in GeoTools. I think it's going through a bit of
revision, as they need to get it up to speed for WCS. And what you
will do for netCDF is write a plugin, implementing all the grid
coverage interfaces (if it was a datastore - the vector stuff, I could
point you in a lot more specific directions. Check the wiki, as I
think there are some gc overviews there, but they may be out of date).
Arcgrid might be a good one to look for inspiration from. Then it
should be able to plug it into the WCS, which is on a branch at:
http://svn.refractions.net/geoserver/branches/wcs-branch/ You'll have
to work off of svn - though I would like to try to get a release out
sometime soon.
best regards,
Chris
Quoting Rob Atkinson <rob@anonymised.com>:
Hi Norman,
Whilst I'm not the greatest expert on these matters I'll try to give
you
some response based on the possible use cases for this that i know
about.Others may be able to put us both right
a) geotools needs to have data source for netCDF - this will be
either
Feature-centric or Grid-centric : this will depend on what has
actually
been packed into the netCDF structure. If you want to subset the
netCDF
to extract feature sets - give me all the readings for sea surface
temp
within this polygon in Jan 97... then you may wish to use Geoserver
to
translate between packed arrays and individual mappable objects - a
perfectly valid use case as far as I can see.which of course brings us to the other NetCDF issue - there are many
ways of putting data into NetCDF files,
b) what NetDCF conventions are you going to support?and of course another perfectly valid use case is to pack extracted
features from any geotools supported data source into a NetCDF
response
format.
c) What Geoserver functions are you planning to invoke? WMS, WFS,
NetCDF
in or out etc - depending on this you will be working on different
parts
of the code base, geotools or geoserverOf course, geotools could be an OpenDAP client - getting NetCDF data
via
the OpenDAP interface, just as it would via WFSHave I thrown the cat amongst the pigeons enough here? Would help to
make your ambitions a bit clearer, and I would suggest this
documentation is built up on the Wiki. I'm willing to help refine the
requirements and try to understand where in the code base the
starting
points are, but by narrowing downh exactly what the priorities are we
can probably identify the real experts and get some pointers.Cheers
Rob AtkinsonNorman Barker wrote:
> Hi,
>
> I have posted again since no one replied, I am available to develop
> the netCDF plugin for the Geoserver WCS is you want, RSI have
> resourced my time. I have also written a netCDF to geotiff routine
if
> that is of any use.
>
> Please advise on what to do next.
>
> Thanks,
>
> Norman Barker
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: NEC IT Guy Games.
> Get your fingers limbered up and give it your best shot. 4 great
> events, 4
> opportunities to win big! Highest score wins.NEC IT Guy Games. Play
to
> win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
> _______________________________________________
> Geoserver-devel mailing list
> Geoserver-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/geoserver-devel-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great
events, 4
opportunities to win big! Highest score wins.NEC IT Guy Games. Play
to
win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
_______________________________________________
Geoserver-devel mailing list
Geoserver-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-devel
--
Please make use of this limegroup email address until our openplans
domain returns, apologies for any missed emails.
----------------------------------------------------------
This mail sent through IMP: https://webmail.limegroup.com/
>
> > Hi,
> >
> > I have posted again since no one replied, I am available to develop
> > the netCDF plugin for the Geoserver WCS is you want, RSI have
> > resourced my time. I have also written a netCDF to geotiff routine
> if
> > that is of any use.
> >
> > Please advise on what to do next.
> >
> > Thanks,
> >
> > Norman Barker
{Pounce}!
Ah hah!! I have a number of uses for you, depending on what perks your
interest.
1] Geotiff could use some more user-defined projection support. Most of it
is there but commented out because the underpinnings in Geotools changed.
If you want to learn how to make a CRS in Geotools, this should make a good
framework within which to learn (especially if you already know Geotiff).
http://jira.codehaus.org/browse/GEOT-507 Simone may already be fiddling
with this, so we'll have to get you talking to him.
2] No one has decided which "flavors" of NetCDF Geotools will support.
There are eight "conventions" posted on the website:
http://my.unidata.ucar.edu/content/software/netcdf/conventions.html. Some
are super/subsets of others, some are distinct. I want to support WRF (
http://jira.codehaus.org/browse/GEOT-505), which does not use any of these
conventions. My trouble is, I don't know which applications use which
conventions.
My first inclination is to suggest that you develop a NetCDF variable <->
GridCoverage2D adapter. You will then be able to use this for any of the
supported conventions. Each convention, then, will require that you adapt
the georeferencing information from the file to the Geotools framework.
Let me know how you'd like to proceed. If we (you, me, simone, alessio,
others) come up with a plan of attack we can document
http://jira.codehaus.org/browse/GEOT-127 and flesh out some sub-tasks.
Interested parties can edit/comment the appropriate task. This is probably
the best way to divide up the pie. So far, all interested parties are up
to their eyeballs in work, whether related to Geotools or not.
Bryce
>From: Bryce L Nordgren <bnordgren@anonymised.com>
>To: Chris Holmes <cholmes@anonymised.com>
>CC: alessio.fabiani@anonymised.com, "Woolf, A (Andrew)" <A.Woolf@anonymised.com>, Geoserver-devel <geoserver-devel@lists.sourceforge.net>, geotools-devel <geotools-devel@lists.sourceforge.net>, geotools-devel-admin@lists.sourceforge.net, Norman Barker <normanbarker@anonymised.com>, simboss_ml@anonymised.com
>Subject: Re: [Geotools-devel] Re: [Geoserver-devel] WCS - netCDF resource available
>Date: Thu, 5 May 2005 11:07:15 -0600
>
> > >
> > > > Hi,
> > > >
> > > > I have posted again since no one replied, I am available to develop
> > > > the netCDF plugin for the Geoserver WCS is you want, RSI have
> > > > resourced my time. I have also written a netCDF to geotiff routine
> > > if
> > > > that is of any use.
> > > >
> > > > Please advise on what to do next.
> > > >
> > > > Thanks,
> > > >
> > > > Norman Barker
>
>{Pounce}!
>
>Ah hah!! I have a number of uses for you, depending on what perks your
>interest.
>
>1] Geotiff could use some more user-defined projection support. Most of it
>is there but commented out because the underpinnings in Geotools changed.
>If you want to learn how to make a CRS in Geotools, this should make a good
>framework within which to learn (especially if you already know Geotiff).
>http://jira.codehaus.org/browse/GEOT-507 Simone may already be fiddling
>with this, so we'll have to get you talking to him.
>2] No one has decided which "flavors" of NetCDF Geotools will support.
>There are eight "conventions" posted on the website:
>http://my.unidata.ucar.edu/content/software/netcdf/conventions.html. Some
>are super/subsets of others, some are distinct. I want to support WRF (
>http://jira.codehaus.org/browse/GEOT-505), which does not use any of these
>conventions. My trouble is, I don't know which applications use which
>conventions.
>
>My first inclination is to suggest that you develop a NetCDF variable <->
>GridCoverage2D adapter. You will then be able to use this for any of the
>supported conventions. Each convention, then, will require that you adapt
>the georeferencing information from the file to the Geotools framework.
>
>Let me know how you'd like to proceed. If we (you, me, simone, alessio,
>others) come up with a plan of attack we can document
>http://jira.codehaus.org/browse/GEOT-127 and flesh out some sub-tasks.
>Interested parties can edit/comment the appropriate task. This is probably
>the best way to divide up the pie. So far, all interested parties are up
>to their eyeballs in work, whether related to Geotools or not.
>
>Bryce
>
Thanks for all the mail, I am starting to look at the coverage API in geotools, I think (after discussion with Andrew Woolf some weeks ago) that I will initially implement a netCDF plugin compliant with Climate-Forecast netCDF, I still don't see how we can implement temporal support in geoserver (though the ability to subset on parameter, spatial is cool). The idea being that you have a coverage, and time is a parameter on that coverage that actually relates to several files rather than one, wheras the current model assumes the parameter is contained in one file (correct me if I am wrong on this, I will be glad to hear it!).
The Galeon project http://my.unidata.ucar.edu/content/projects/THREDDS/GALEON/GALEON-Activity-Plan.htm has some initial use cases that I was going to try to implement.
I would like to help with geotiff (since when I used to be a humble Delphi programmer I had to rewrite libgeotiff in pascal as an exercise!), but netCDF comes first in our priorities.
Moving geoserver forward with netCDF coverage would be a big step forward, and thanks for all your help and suggestions.
Norman Barker