On 19 Oct 2016, at 16:47, Jonathan S. Katz <jkatz@postgresql.org> wrote:
<snip>
> Because this is my first time doing this, I want to work on it early and often to get into the proper cadence :-) I
tooka try at writing up the data corruption issue as well as adding the EOL language:
>
>
https://git.postgresql.org/gitweb/?p=press.git;a=commitdiff;h=38aadd18e9c64f30423f81920381b001fff8b4c5;hp=1ada1db63eaed94e499c1289e90bdf70a47f9fc0
>
> I understand part of the challenge is taking the detailed commit notes and putting it into language that makes sense
tothe users, which is why I want more eyes on this. I intend to fill out more details once the rest of the release
notesare out, but if there are major issues I find it prudent that we get the language for that correct so users
understandthe ramifications and the need to upgrade.
Some bits after scanning through that. :)
* For this fragment:
... and 9.1.24, where the 9.1 release is the last release for PostgreSQL
version 9.1.
It's kind of unclear about the 9.1.
Would this be correct?
... and 9.1.24, where this 9.1 release will be the last for PostgreSQL
version 9.1.
* Typo bit here:
This update fixes around the WAL-logging of truncated relations, which
now ensure that the ...
Probably nuke both "around", and "that"
That being said, this whole fragment seems super unwieldy:
This update fixes around the WAL-logging of truncated relations, which now
ensure that the FSM is truncated when a TRUNCATE command on a relation is
issued, leading to data corruption. Should the FSM not be truncated, a
PostgreSQL database in recovery mode could return a page that has already
been truncated and return an error such as ...
Probably just needs breaking up into shorter sentences, for easier mental
digestion. :)
+ Justin
--
"My grandfather once told me that there are two kinds of people: those
who work and those who take the credit. He told me to try to be in the
first group; there was less competition there."
- Indira Gandhi
