The approach in the patch is different: (4) Add transaction ids, etc to the index tuple (totaling 16 bytes) I would group (1) & (4) together and (2) & (3) together. I think the time and space trade-offs are pretty obvious, so I won't waste time on those. postgresql で、インデックス

ita-wasa のブックマーク 2008/10/21 11:49

<blockquote class="hatena-bookmark-comment"><a class="comment-info" href="https://b.hatena.ne.jp/entry/10490731/comment/ita-wasa" data-user-id="ita-wasa" data-entry-url="https://b.hatena.ne.jp/entry/archive.netbsd.se/?ml=pgsql-hackers&a=2008-06&t=7789335" data-original-href="http://archive.netbsd.se/?ml=pgsql-hackers&a=2008-06&t=7789335" data-entry-favicon="https://cdn-ak2.favicon.st-hatena.com/64?url=http%3A%2F%2Farchive.netbsd.se%2F%3Fml%3Dpgsql-hackers%26a%3D2008-06%26t%3D7789335" data-user-icon="/users/ita-wasa/profile.png">archive.netbsd.se</a><ul class="comment-tag" style="list-style: none; margin: 0px;"><li style="float: left">[<a href="https://b.hatena.ne.jp/q/PostgreSQL">PostgreSQL</a>]</li><li style="float: left">[<a href="https://b.hatena.ne.jp/q/MVCC">MVCC</a>]</li></ul><br><p style="clear: left">The approach in the patch is different: (4) Add transaction ids, etc to the index tuple (totaling 16 bytes) I would group (1) &amp;amp; (4) together and (2) &amp;amp; (3) together. I think the time and space trade-offs are pretty obvious, so I won't waste time on those. postgresql で、インデックス</p><a class="datetime" href="https://b.hatena.ne.jp/ita-wasa/20081021#bookmark-10490731"><span class="datetime-body">2008/10/21 11:49</span></a></blockquote><script src="https://b.st-hatena.com/js/comment-widget.js" charset="utf-8" async></script>

このブックマークにはスターがありません。
最初のスターをつけてみよう！