Dave was asking how much time it cost to set up the STRUTS configuration that GeoServer has now.
Chris Holmes and I were attempting to answer that question:
Jody Garnett: Dave was asking after the coast of implemnting the struts config.
Jody Garnett: We spent equal time just improving geoserver
Jody Garnett: making use of the container, defining DTO objects for the configuration etc ...
Jody Garnett: I think we only spent 2 months on STRUTS, mostly 1.5 people.
cholmesny: yeah
Most of the cost was in reverse engineering what GeoServer had, writing documentation for it, defining Data Transfer Objects, making sure that startup occurred in a reproducible manner (it used to be a lazy loading nightmare where the first request would time out), etc ...
Basically the STRUTS part was relatively easy. The most annoying thing was not being allowed to modify the running GeoServer application directly (hense the Config / Apply / Save workflow). If configuration changed the running GeoServer you could remove about 30% of the code.
Even though STRUTS was not the major cost, I do find that time has moved on a bit - I would use Spring if I was doing it today. GeoServer makes limited use of the Application Container today (mostly to find the three subsystems Data, WFS, WMS and Validation).
You should find that you can make a GeoServer that is just made up of Data & WMS for example. The trick is to take this idea and run with it, I have not had a look at the WCS work but I would be dissapointed if it was not a separate module discovered via a container lookup.
Dave was asking how much time it cost to set up the STRUTS configuration
that GeoServer has now.
Chris Holmes and I were attempting to answer that question:
Jody Garnett: Dave was asking after the coast of implemnting the
struts config.
Jody Garnett: We spent equal time just improving geoserver
Jody Garnett: making use of the container, defining DTO objects for
the configuration etc …
Jody Garnett: I think we only spent 2 months on STRUTS, mostly 1.5 people.
cholmesny: yeah
Most of the cost was in reverse engineering what GeoServer had, writing
documentation for it, defining Data Transfer Objects, making sure that
startup occurred in a reproducible manner (it used to be a lazy loading
nightmare where the first request would time out), etc …
Basically the STRUTS part was relatively easy. The most annoying thing
was not being allowed to modify the running GeoServer application
directly (hense the Config / Apply / Save workflow). If configuration
changed the running GeoServer you could remove about 30% of the code.
Even though STRUTS was not the major cost, I do find that time has moved
on a bit - I would use Spring if I was doing it today. GeoServer makes
limited use of the Application Container today (mostly to find the three
subsystems Data, WFS, WMS and Validation).
You should find that you can make a GeoServer that is just made up of
Data & WMS for example. The trick is to take this idea and run with it,
I have not had a look at the WCS work but I would be dissapointed if it
was not a separate module discovered via a container lookup.
Cheers,
Jody
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click