metaTiling problems wrapping the earth...

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

metaTiling problems wrapping the earth...

Andrew Hughes-2
Hi Again,

I've noticed that, although tilecache is probably not the cause of the
problem here I'm encountering some problems with metaTiling....

When requesting BBox = 90,-45,135,0
The corresponding metaTile caclculated is BBox = 45.0,-90.0,*270*.0,90.0
(metaSize=5,5  metaBuffer=0)

Anyway, the problem is that the WMS is not smart enough to wrapthe
coverage around to "270" degrees X.... Should TileCache not be composing
such BBox's?

I'm know the answer is not particularly straight foward, and it gets
down to the fact that the metaSize then needs to divide into coverage...
some testing on the metaTile eta... but anyway it's more about if this
belongs in the project... or if the WMS should be expected to solve this
problem.


--Cheers



_______________________________________________
Tilecache mailing list
[hidden email]
http://openlayers.org/mailman/listinfo/tilecache
Reply | Threaded
Open this post in threaded view
|

Re: metaTiling problems wrapping the earth...

Andrew Hughes-2
Now that I've had sometime to think about this here goes...

Sure it would be nice for TileCache and the WMS to handle such
circumstances... but it would be a significant amount of work.
Another solution is to use metaSize that are to the power of 2. Thus,
1,2,4,8...... all metaTile to the grid without any "remainder" tiles
that overlap past +/-180 or +/-90 degrees.

Im not sure what the default metaSize is? but so that TileCache works
out the box with one less potential error.. can I make a suggestion that
the default be one of these powers of 2.... I would suggest 4,4.... but
you would also neeed to use a metaBuffer=0 as well. Maybe put this in
the readme or at least on the wiki?

Hope this adds some value to TileCache....


--AH



Andrew Hughes wrote:

> Hi Again,
>
> I've noticed that, although tilecache is probably not the cause of the
> problem here I'm encountering some problems with metaTiling....
>
> When requesting BBox = 90,-45,135,0
> The corresponding metaTile caclculated is BBox = 45.0,-90.0,*270*.0,90.0
> (metaSize=5,5  metaBuffer=0)
>
> Anyway, the problem is that the WMS is not smart enough to wrapthe
> coverage around to "270" degrees X.... Should TileCache not be composing
> such BBox's?
>
> I'm know the answer is not particularly straight foward, and it gets
> down to the fact that the metaSize then needs to divide into coverage...
> some testing on the metaTile eta... but anyway it's more about if this
> belongs in the project... or if the WMS should be expected to solve this
> problem.
>
>
> --Cheers
>
>
>
> _______________________________________________
> Tilecache mailing list
> [hidden email]
> http://openlayers.org/mailman/listinfo/tilecache
>
>  


--

Regards,

*Andrew Hughes*
Software Engineer
LISAsoft Pty. Ltd. (Adelaide)

------------------------------------------------------------------------


      LISAsoft Pty. Ltd.

<http://www.lisasoft.com>
*ADELAIDE Office*
<http://terrapages.net/mapbutton/RetrieveButtonServlet?buttonID=4>
38 Greenhill Road
Wayville SA 5034
Australia
*Telephone +61 8 8272 1555*
*Facsimile +61 8 8271 1199*
        *SYDNEY Office*
<http://terrapages.net/mapbutton/RetrieveButtonServlet?buttonID=1>
Suite 112 The Lower Deck
Jones Bay Wharf
19-21 Pirrama Road
Pyrmont NSW 2009 AUS
*Telephone +61 2 8570 5060*
*Facsimile +61 2 8570 5099*
        *MELBOURNE Office*
<http://terrapages.net/mapbutton/RetrieveButtonServlet?buttonID=3>
Level 7 520 Collins Street
Melbourne VIC 3000
Australia
*Telephone +61 3 9629 1799*
*Facsimile +61 3 9629 4955*

------------------------------------------------------------------------
_______________________________________________
Tilecache mailing list
[hidden email]
http://openlayers.org/mailman/listinfo/tilecache