-
Notifications
You must be signed in to change notification settings - Fork 141
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
Switch to Vello (new PR) #635
Conversation
It's really great to see this. Why would a second window require a second We have not put much effort into reducing memory usage yet - some work has started in linebender/vello#606 but has stalled due to a lack of review bandwidth. Italic/bold text should be handled by your text rendering library providing you with the bold/italic font for that run, or else providing the correct emulated variable font parameters. |
Sounds like it doesn't need to. I'll see about fixing that.
👍
Yeah I've probably just messed something up.
I'll see what I can do there. |
@dzhou121 What do you think/what are priorities? |
I think maybe we can still keep vger, and control via feature flags? |
Yeah that would probably be the way to go for now. That would allow us to keep improving the vello integration without affecting lapce. I'll add back vger and then mark this as ready for review and then open issues for the current issues that I've mentioned |
I think this is ready for review/to be merged now that it has vger as the defualt and vello behind a flag |
Anything else need changed? |
Vello uses a lot more memory than vger does.
In debug mode, the widget gallery with vger uses about 90MB of memory, vello uses over 400. (specific numbers depend greatly on window size).
similar in release mode.
Each new window will use another another renderer which will use another 400-600 MB of memory which gets excessive quickly.
Need to fix:
Marking as draft because:
vello doesn't have the font_embolden. (Personally I don't like the look on macOS with font embolden so I am fine with that.)
high memory usage
Need to verify that vello actually performs comparable to vger
I do really want the features (and correctness) vello offers. Even with the tradeoffs I'd much prefer to be working on floem with vello as the renderer.
@DJMcNab