New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Regression in :before pseudo element rendering (when async: true
)
#361
Comments
async: true
)
@usmonster I wrote #270 under another account; here's the issue as I see it: The PR contained a few loops that ran through all css (via The issue is that the |
The whole lib has been rewritten (merged into master yesterday), and one of the many changes it introduced was async parsing (in a secondary hidden contentWindow). Any old code, including this PR has been replaced. |
@niklasvh Great news! How is the speed on the new release compared to the old async? |
Hasn't been benchmarked yet, still some issues (#421) to be resolved. It won't need to be thread blocking though, so DOM can change while the parsing/rendering is occuring. |
Awesome! So it's always async by default, then? Good stuff. :] |
The rendering is currently async by default, the parsing still needs to be implemented. For reference: |
When setting
async: true
, certain:before
content specified in the CSS does not appear on the rendered canvas. This issue seems to have been introduced in #270 (e115180 specifically), though I haven't yet thoroughly read through the commit to identify the root cause.I first noticed this when trying to capture the main, not-logged-in page of http://www.github.com . The 4 circles below "Why you’ll love GitHub" are blank when using the version cited and later, but content does appear when using older versions. (Unrelated issue that the content itself is not correct.. possibly related to #226 and/or fonts/proxies?)
If it helps, the CSS responsible for the content is of this form:
I haven't had time to test whether this issue also affects
:after
content, or whether the fact that the content is escaped has anything to do with it, but it's a definite regression in any case.Thanks for a great tool, and keep up the amazing work!
The text was updated successfully, but these errors were encountered: