Issue 304 in webp: Is there a way to get a particul:qar -DWEBP_SWAP_16BIT_CSP by default?

[mtkl… via monorail]

Status: New
Owner: ----
Labels: Type-Defect Priority-Medium

New issue 304 by mtkl...@google.com: Is there a way to get a particul:qar -DWEBP_SWAP_16BIT_CSP by default?
https://bugs.chromium.org/p/webp/issues/detail?id=304

I've just stumbled over -DWEBP_SWAP_16BIT_CSP a bit, and I notice that it's basically defined by everyone I can find:

Skia (and thus Android): https://bugs.chromium.org/p/skia/issues/detail?id=4174
Arch Linux: https://git.archlinux.org/svntogit/packages.git/tree/trunk/PKGBUILD?h=packages/libwebp
Homebrew:

I imagine Chrome/Blink haven't done this only because they don't decode into 16-bit formats generally.

--
You received this message because:
1. The project was configured to send all issue notifications to this address

You may adjust your notification preferences at:
https://bugs.chromium.org/hosting/settings

[mtkl… via monorail]

Comment #1 on issue 304 by mtkl...@google.com: Is there a way to get a particul:qar -DWEBP_SWAP_16BIT_CSP by default?
https://bugs.chromium.org/p/webp/issues/detail?id=304#c1

Err, I filed this accidentally while finding out that this is _not_ generally agreed on. Looks like a lot of places define WEBP_SWAP_16BIT_CSP, and lots don't.

I was about to reword the bug into, is there a way we can always ask for little endian 16-bit, regardless of how WEBP_SWAP_16BIT_CSP is set? I guess we can always run a little canary yuv -> 565 conversion and postprocess with bswap if it comes out backwards?