From: Steinar H. Gunderson <sgunderson@bigfoot.com>
Date: Sat, 11 Nov 2023 15:43:37 +0000 (+0100)
Subject: Better comment about the UCA cache.
X-Git-Url: https://git.sesse.net/?a=commitdiff_plain;h=59418fa3ce05b97bcae028b351259196e4fa843e;p=skvidarsync

Better comment about the UCA cache.
---

diff --git a/bin/sync.pl b/bin/sync.pl
index 14c6a3b..d8866a6 100644
--- a/bin/sync.pl
+++ b/bin/sync.pl
@@ -53,9 +53,8 @@ sub log_timing {
 }
 
 # Unicode::Collate is seemingly slow, so add a cache for each name part
-# (which, of course, only works for equality).
-# Doesn't seem to help the initial one, though; I guess not enough people
-# have the same names.
+# (which, of course, only works for equality). Helps especially in
+# --daemon mode, where even the first request gets a warm cache.
 my %sort_key_cache = ();
 my $sort_key_sp = $uca->getSortKey(' ');