- From: Dirk Schulze <dschulze@adobe.com>
- Date: Sun, 14 Jul 2013 07:03:04 -0700
- To: "robert@ocallahan.org" <robert@ocallahan.org>
- CC: "Tab Atkins Jr." <jackalmage@gmail.com>, fantasai <fantasai.lists@inkedblade.net>, www-style list <www-style@w3.org>, Simon Fraser <smfr@me.com>
On Jul 12, 2013, at 2:29 PM, Robert O'Callahan <robert@ocallahan.org> wrote: > On Fri, Jul 12, 2013 at 5:49 AM, Dirk Schulze <dschulze@adobe.com> wrote: > On Jul 12, 2013, at 2:28 AM, Robert O'Callahan <robert@ocallahan.org> wrote: > > > I thought we had agreed to solve this by, for url() values for SVG properties like 'mask', if they contain anything other than a plain #ref, perform an image load, check for same-origin and then checking the loaded image to see if the URL refers to a <mask> element an SVG image. > > Yes, this is point 1) on the option list. > > No, it isn't. > > Your option 1 says "if there is an hash, assume it is an SVG resource, otherwise an CSS Image". I thought we had agreed to "if there is an hash *and the referenced element is of the right type* assume it is an SVG resource, otherwise an CSS Image." So for 'mask', the referenced element will only be treated as an SVG resource if it's a <mask>. Not sure if we agreed on that. This would basically mean that we need to fetch the resource first, analyze it and then (if necessary) either fetch the same resource with different policies or (if the implementation is smart enough) allow changing the policies of the resource on fly. The later one is difficult, since the initial fetching decision decides about the fetching that the resource is allowed to do on it's own (SVG image that references further images vs. SVG mask that fetched further images). Example: mask-image: url(test.svg#notThere) 1) fetch the resource as SVG mask resource because we found a hash 2) analyze if the hash exists a) if it does: continue as normal b) if it does not: refetch resource with different policies, since CSS Image is not allowed to fetch further resources, while SVG mask resource was allowed to. This would be contra productive for our initial requirement to decide during parse time which fetch algorithm is used. Greetings, Dirk > > Rob > -- > Jtehsauts tshaei dS,o n" Wohfy Mdaon yhoaus eanuttehrotraiitny eovni le atrhtohu gthot sf oirng iyvoeu rs ihnesa.r"t sS?o Whhei csha iids teoa stiheer :p atroa lsyazye,d 'mYaonu,r "sGients uapr,e tfaokreg iyvoeunr, 'm aotr atnod sgaoy ,h o'mGee.t" uTph eann dt hwea lmka'n? gBoutt uIp waanndt wyeonut thoo mken.o w
Received on Sunday, 14 July 2013 14:03:36 UTC