Proposal
Today the CF/Org/Space filter column (apps wall, other signal-list pages) renders as a single concatenated haystack — cfName / orgName / spaceName — and the text filter does a case-insensitive .includes() match.
Practical effect:
- A single token (e.g.
engineering) matches if it appears in any of the three segments (effectively OR).
- A query containing slashes (e.g.
/open/pro) is treated as a literal substring against the concatenation. It does not scope the query to a specific segment, so the slash is dead weight in user input.
A user might reasonably expect /open/pro to mean "Org contains open AND Space contains pro" — but no such parser exists.
Possible directions
- Per-field syntax — parse
org:open space:pro into a structured multi-clause predicate. Most discoverable, but requires a small parser and grammar choice.
- Implicit AND across whitespace-split tokens — typing
open pro matches if both tokens appear somewhere in the haystack. Cheapest lift (~5 LOC over the current extractor).
- Slash as segment anchor — adopt a convention where
cf/org/space syntactically maps to segment slots (a leading / anchors to org, two slashes to space). Lower discoverability; needs a placeholder hint.
Per-column dropdowns (CF / Org / Space) are already present on the apps wall and were just added to Routes / Services / Users / Events, so this is purely about enriching the text-filter UX. Not blocking any current workflow.
Acceptance / open questions
Proposal
Today the CF/Org/Space filter column (apps wall, other signal-list pages) renders as a single concatenated haystack —
cfName / orgName / spaceName— and the text filter does a case-insensitive.includes()match.Practical effect:
engineering) matches if it appears in any of the three segments (effectively OR)./open/pro) is treated as a literal substring against the concatenation. It does not scope the query to a specific segment, so the slash is dead weight in user input.A user might reasonably expect
/open/proto mean "Org containsopenAND Space containspro" — but no such parser exists.Possible directions
org:open space:prointo a structured multi-clause predicate. Most discoverable, but requires a small parser and grammar choice.open promatches if both tokens appear somewhere in the haystack. Cheapest lift (~5 LOC over the current extractor).cf/org/spacesyntactically maps to segment slots (a leading/anchors to org, two slashes to space). Lower discoverability; needs a placeholder hint.Per-column dropdowns (CF / Org / Space) are already present on the apps wall and were just added to Routes / Services / Users / Events, so this is purely about enriching the text-filter UX. Not blocking any current workflow.
Acceptance / open questions
placeholderso it's discoverable