Cascade delete Styles
---------------------

                 Key: GEOS-2947
                 URL: http://jira.codehaus.org/browse/GEOS-2947
             Project: GeoServer
          Issue Type: Sub-task
            Reporter: Gabriel Roldán
            Assignee: Gabriel Roldán
             Fix For: 2.0-beta1


{code}
 <aaime> I'd suggest to not cascade a deleation of a style, if a style is the 
default of a layer, don't remove it, if a style is an "extra style", remove it 
from the list
 <aaime> also, polygon/point/line styles should not be allowed to be removed
 <groldan> hummm
 <aaime> we use them when building new layers
 <aaime> or, raster too
 <groldan> I don't think there's any cascading on styles
 <groldan> like in they're first class citizens?
 <aaime> yet they are used
 <aaime> if this was a database you could say the fk cascade deletes
 <aaime> I suggested above a way to handle style deletes
 <aaime> what do you suggest instead and why
 <groldan> reading over your comment...
 <groldan> if you delete a layer, do not delete any style. If you try to delete 
a style that's the default for a layer, do not allow it?
 <groldan> take care and go change the layers style
 <aaime> how?
 <groldan> another option would be to delete the style and update the layer to 
a default style
 <aaime> revert it to the default for the default geometry? could work
 <groldan> its going to be less annoying for the user I guess
 <aaime> agreed
 <groldan> say you actually WANT to delete the style, but have 10 layers using 
it...
 <groldan> ok
 <aaime> yet we have to report which layers will be affected
 <aaime> so I guess that's part of the helper as well?
 <groldan> yes
 <aaime> if the code is about to change 10 layers I want to know, as a user :-)
 <groldan> there seems to be two kind of side effects: cascade delete and 
config change
 <aaime> and I also would like to know if the layer default style is changed, 
or if only an altearnate one is removed, no?
 <aaime> groldan, yes
 <aaime> even removing a layer should, imho, just change the layer groups in 
which the layer is included no?
 <groldan> lets do that so, report back any change that's gonna happen before 
the user applies the deletion
 <aaime> for styles it's also relevant the kind of change is going to happen 
imho
 <aaime> default style or jsut one less alternate style?
 <groldan> yup
 <aaime> so I guess the helper will have to return appropriate beans listing 
the affected parts?
 <aaime> as in, it's not just a flat list of catalog objects, we need to know 
what's deleted, what's changed for "data"
 <aaime> and which layers are affected and how for styles
 <groldan> not sure it has to return the actual beans
 <aaime> separate methods to gather separate results?
 <groldan> but an indication of what's going to happen, say a struct like type, 
id, name, kind of affection
 <aaime> that's a bean to me ;)=
 <aaime> java does not have structs ;)
 <groldan> yeah, I thought you meant the catalog objects
 <groldan> like returning the affected LayerInfo etc
 <aaime> nah, generic pojo
 <groldan> yup, agreed
 <aaime> and the pojo contains a list of LayerINfo and such
 <aaime> no?
 <groldan> not sure
 <aaime> what would it contain otherwise?
 <groldan> not sure :)
 <aaime> let's make an example
 <aaime> say  I'm deleting a datastore
 <aaime> the pojo, imho, could contain:
 <aaime> 1) a list of layers that are going to be removed
 <aaime> 2) a list of LayerGroup that are giong to be changed
 <aaime> if I'm deleting a workspace, then the above, but also a list of 
datastores that are going to be deleted
 <groldan> the pojo in question is gonna be kind of recursive
 <aaime> not sure we want that
 <aaime> I mean, a flat list coudl be enough no?
 <aaime> think also about presentation
 <aaime> do you want to present a tree of what's going to be deleted?
 <aaime> maybe a flat list is enough ;-)
 <aaime> (a paged one)
 <groldan> not a tree
 <groldan> for presentation
 <groldan> more like titles with lists
 <groldan> The following layers are going to be removed:
 <groldan> -
 <groldan> -
 <groldan> -
 <aaime> indeed... so, why do you need the bean to be recursive?
 <groldan> The following layer groups are going to be changed:
 <groldan> -
 <groldan> -
 <groldan> to have a single bean able to hold any kind of affected list of 
objects
 <groldan> a single class I mean
 <groldan> containing a list of the same bean class instances
 <aaime> hmmm... maybe let's do it another way
 <aaime> you propose a design in the jira
 <aaime> and then I look into it?
 <groldan> yeah stop wasting time here
{code}

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: 
http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       

------------------------------------------------------------------------------
Register Now & Save for Velocity, the Web Performance & Operations 
Conference from O'Reilly Media. Velocity features a full day of 
expert-led, hands-on workshops and two days of sessions from industry 
leaders in dedicated Performance & Operations tracks. Use code vel09scf 
and Save an extra 15% before 5/3. http://p.sf.net/sfu/velocityconf
_______________________________________________
Geoserver-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geoserver-devel

Reply via email to