An interesting feature was proposed in #90, and I suggested that we decline to add it because:
I'm personally not a fan of workarounds for either security controls or standard behaviors. svg4everybody is a polyfill, which means that it is intended to upgrade implementations to support the standard behavior. Because this feature introduces behavior that deviates from the standard, I don't think that it belongs here.
For this issue and more generally, I'd like to suggest criteria for deciding whether to accept features like this in the future:
Does the input SVG work as intended in standard implementations without the feature?
If the answer is no, then I think the feature should be rejected. One counter-argument is that users may pass an option to force the shim, but I would argue that this is bad practice because it penalizes users of modern browsers and effectively prevents JS-free users from seeing the SVGs.
Whatever we decide should go into CONTRIBUTING.md and possibly into an issue or pull request template so that contributors know what to expect when they propose a new feature.
An interesting feature was proposed in #90, and I suggested that we decline to add it because:
For this issue and more generally, I'd like to suggest criteria for deciding whether to accept features like this in the future:
Does the input SVG work as intended in standard implementations without the feature?
If the answer is no, then I think the feature should be rejected. One counter-argument is that users may pass an option to force the shim, but I would argue that this is bad practice because it penalizes users of modern browsers and effectively prevents JS-free users from seeing the SVGs.
Whatever we decide should go into
CONTRIBUTING.md
and possibly into an issue or pull request template so that contributors know what to expect when they propose a new feature.