Using node.id would produce route names like r_transmon_1_transmon_2 instead of r_transmon_transmon$N. The advantage is that the route name would directly identify which specific node pair it connects, not just which component types. This would be more informative for debugging.
Technically, any change to auto-generated names can be breaking if downstream code depends on specific name patterns (e.g., string matching on GDS cell names). In practice:
uniquename() already appends a counter suffix ($N), so names are not stable across sessions anyway.
- The docstring says
name should be unique but does not specify the format.
- Users can always pass
name=... explicitly to override.
Using
node.idwould produce route names liker_transmon_1_transmon_2instead ofr_transmon_transmon$N. The advantage is that the route name would directly identify which specific node pair it connects, not just which component types. This would be more informative for debugging.Technically, any change to auto-generated names can be breaking if downstream code depends on specific name patterns (e.g., string matching on GDS cell names). In practice:
uniquename()already appends a counter suffix ($N), so names are not stable across sessions anyway.nameshould be unique but does not specify the format.name=...explicitly to override.