Repository navigation
10.x: docs: ordering by LowerBound() or UpperBound() puts open-ended and empty rows where PostgreSQL puts NULLs (port of #359) #364
Description
Problem
The same gap as #359, in the release/10.x docs. docs/ranges.md#sort-ranges documents PostgreSQL's NULLS LAST against memory's null-first for a nullable column, but not for an ordering by a bound. There the NULLs come from non-null ranges, so a required column has them too. README.md and docs/efcore.md#read-bounds-merge-ranges-and-aggregate show bookings.OrderBy(b => b.Period.LowerBound()) as their ordering example. The behaviour is the same on 10.1.0, since the translation is identical.
Change (10.2.0)
- Cherry-pick Docs: ordering by LowerBound() or UpperBound() puts open-ended and empty rows where PostgreSQL puts NULLs, also on a required column #359's docs change, with its claim and the test that holds it, with
-x. - A
Docs:line under### Fixedin[10.2.0]of the root and EF Core changelogs.
Done when
The docs on release/10.x say where an ordering by a bound puts open-ended and empty rows. The claim is held there and OracleCatalogueTests passes. Close by hand after the merge.
Activity
CaffeinatedCoder commentedon Oct 24, 2026
The
mainchange is PR #372 (commit 0fc386f, docs plus a test and a seed): the bullet indocs/efcore.md#read-bounds-merge-ranges-and-aggregate, the sentence indocs/ranges.md#sort-ranges, the comments at the four examples (both READMEs,efcore.md,getting-started.md), claim E72,ExecutedQueryTests.OrderingByABound_PutsTheRowsWithoutItWherePostgresPutsNullsand the seedbound-declared-null-only-for-a-null-argument.patch. The test relies on #349's fix, which 10.x has from #367, so the port cherry-picks the commit as it is. ItsDocs:line goes under### Fixedin[10.2.0].CaffeinatedCoder commentedon Oct 24, 2026
Port PR: #374, stacked on #373 (both write
[10.2.0]andclaims.json). It is the-xcherry-pick of 0fc386f (#372), with a### Fixedsection for theDocs:line. The ordering test is green on PostgreSQL 16 and 18 and on EF Core 10.0.13, and its seed is caught on the branch.CaffeinatedCoder commentedon Oct 24, 2026
Fixed on
release/10.xby #374, merged as fb9aeed:docs/efcore.mdsays an ordering byLowerBound()orUpperBound()puts the empty rows and those without that bound where PostgreSQL puts NULLs, also on a required column, with the!= nullkey that sorts them as memory does.[10.2.0]has theDocs:row under### Fixed. Ships in 10.2.0 (#358). Closed by hand, sinceClosesacts only onmain.