Daniel Rolf

icon indicating a website link https://de.linkedin.com/in/dr-daniel-rolf-8185576 icon indicating an email [email protected]

icon company Meta icon location Berlin icon location XR Maps at Meta, Ex-Technical Fellow at HERE Technologies

Results 51 comments of Daniel Rolf

Using pnpm: Error: Can't find stylesheet to import: `@use '@material/theme/color-palette';`

I know it’s destructive, but my solution was to drop smui and directly use the material web components and css classes

Type.getOwnProperties to get properties without base

I think it’s not possible. I ended up climbing the AST for my use case

Unidirectional A* variants have issues with multiple destination edges

Pretty sure that this is the root cause for https://github.com/valhalla/valhalla/issues/1952

Unidirectional A* variants have issues with multiple destination edges

Thanks @kevinkreiser ! Right now our “flag” is to ride on the fact that destination edge labels have dist = 0.0. However that feels a bit like a hack and...

Unidirectional A* variants have issues with multiple destination edges

@kevinkreiser I am on it. Do you have a good pointer for me to create test cases with multiple start and destination points?

Unidirectional A* variants have issues with multiple destination edges

Guten Abend @kevinkreiser ! So, I added some tests in: https://github.com/derolf/valhalla/commit/d836460a089ab73ac042d722b5f7c6d5f50a070d They pass with `radius=0`, but fail as soon as I raise the `radius` and hence generate multiple origin, destination...

Unidirectional A* variants have issues with multiple destination edges

Interestingly, my test caught a potential bug in the `time_dependent_reverse_a*`. With `radius=0`, I the following examples fail with `type=2` (`time_dependent_reverse_a*`). ``` Expected equality of these values: actual_names Which is: {...

Unidirectional A* variants have issues with multiple destination edges

Fixed (including the bug in https://github.com/valhalla/valhalla/issues/3582#issuecomment-1078800283)

Multiple origin and destination candidates results in strange detours in unidirectional A*

Dear reviewers, the basic idea is to wrap label creation into a lambda (`add_label`) in both `SetOrigin` and `ExpandInner`. `add_label` is always called to at least add the "normal" edge....

Multiple origin and destination candidates results in strange detours in unidirectional A*

The test `Standalone, NotThruPruning` fails. I looked through my changes and can't figure out what's wrong there. Any idea? EDIT: My fault, found the reason...