Tom Lane d6542f8dfc Doc: remove obsolete, confused <note> about rowtype I/O syntax. ... This <note> was originally written to describe the double levels of de-backslashing encountered when a backslash-aware string literal is used to hold the text representation of a composite value. It still made sense when we switched to mostly using E'...' syntax for that type of literal. However, commit f77de4b0c mangled it completely by changing the example literal to be SQL-standard. The extra pass of de-backslashing described in the text doesn't actually occur with the example as written, unless you happen to be using standard_conforming_strings = off. We could restore this <note> to self-consistency by reverting the change from f77de4b0c, but on the whole I judge that its time has passed. standard_conforming_strings = off is nearly obsolete, and may soon be fully so. But without that, the behavior isn't so complicated as to justify a discursive note. I observe that the nearby section about array I/O syntax has no equivalent text, although that syntax is equally subject to this issue. Author: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/2998401.1767038920@sss.pgh.pa.us Discussion: https://postgr.es/m/3279216.1767072538@sss.pgh.pa.us		2025-12-31 13:19:27 -05:00
..
sgml	Doc: remove obsolete, confused <note> about rowtype I/O syntax.	2025-12-31 13:19:27 -05:00
Makefile	Remove distprep	2023-11-06 15:18:04 +01:00