[Tiling] Fwd: [OSGeo-Standards] Irregular tile coverage in TMS?

[Arne Kepp]

  My interpretation would be that 404 is the closest thing to a correct 
response, at least if you want TMS to function the same way when served 
from a directory of static files.

In GeoWebCache you can create b) using request filters, and the requests 
that land on blank tiles result in a transparent PNG. That was mostly 
for WMS-C's sake, where you can't return 404 and instead would have to 
standardize a new OGC exception.

A transparent PNG is only some 200 bytes, which is usually comparable to 
the HTTP overhead. The trouble with redirecting is that if you use a 301 
the browser caches the redirect for as long as the target is valid, and 
if you use a 302 it will recheck the redirect every time. (Only tested a 
few browsers.) I stuck to sending the image each time with the same 
expiration time as regular tiles.

-Arne


On 9/12/10 1:33 PM, Mateusz Loskot wrote:
> Folks,
>
> I have posted a little question about TMS spec to the standards,
> but perhaps I could receive some feedback here as well:
>
>
>
> -------- Original Message --------
> Subject: [OSGeo-Standards] Irregular tile coverage in TMS?
> Date: Sat, 11 Sep 2010 19:46:51 +0200
> From: Mateusz Loskot<mateusz at loskot.net>
> To: standards at lists.osgeo.org
>
> Folks,
>
> I'm reading the TMS spec which says:
>
> "a coverage of regularly sized and spaced images that taken together
> form a complete visual representation of the entire area of coverage"
>
> I understand the size and spacing constraints.
> What I'm missing, however, is what TMS says about "topological"
> regularity of tiles within a coverage.
>
> I attached basic drawing with two example tile coverages:
>
> a) blue - regular coverage, rectangular
>
> b) red - irregular coverage
>
> Is only blue version kosher for TMS or both versions are valid?
>
> Best regards,
>