Re: PostgreSQL 17 Beta 1 release announcement draft

From: Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org>
To: "Jonathan S(dot) Katz" <jkatz(at)postgresql(dot)org>
Cc: PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: PostgreSQL 17 Beta 1 release announcement draft
Date: 2024-05-23 12:00:13
Message-ID: 202405231200.adombcwm6hd4@alvherre.pgsql
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Looks good. Some minor changes:

On 2024-May-22, Jonathan S. Katz wrote:

> ### Query and Operational Performance Improvements
>
> PostgreSQL 17 builds on recent releases and continues to improve performance across the entire system. [Vacuum](https://www.postgresql.org/docs/17/routine-vacuuming.html), the PostgreSQL process responsible for reclaiming storage, has a new internal data structure that has shown up to a 20x memory reduction for vacuum,

This reads funny:
"Vacuum ... has shown a memory reduction for vacuum"
Maybe just removing the "for vacuum" words at the end of the phrase is a
sufficient fix.

> PostgreSQL 17 can now use both planner statistics and the sort order of [common table expressions](https://www.postgresql.org/docs/17/queries-with.html) (aka [`WITH` queries](https://www.postgresql.org/docs/17/queries-with.html)) to further

Is usage of "aka" typical? I would have expected "a.k.a." but maybe I'm
just outdated.

> Finally, PostgreSQL 17 adds more explicit SIMD instructions, including AVX-512 support for the [`bit_count](https://www.postgresql.org/docs/17/functions-bitstring.html) function.

Note the lack of closing backtick in [`bit_count`].

> ### Developer Experience
>
> PostgreSQL 17 continues to build on the SQL/JSON standard, adding support for the `JSON_TABLE` features that can convert JSON to a standard PostgreSQL table, and SQL/JSON constructor (`JSON`, `JSON_SCALAR`, `JSON_SERIALIZE`) and query functions (`JSON_EXISTS`, `JSON_QUERY`, `JSON_VALUE`). Notably, these features were originally planned for the PostgreSQL 15 release but were reverted during the beta period due to design considerations, which is one reason we ask for you to help us test features during beta! Additionally, PostgreSQL 17 adds more functionality to its `jsonpath` implementation, including the ability to convert JSON values to different data types.

I'm not sure it's accurate to say that converting JSON values to
different datatypes is part of the jsonpath implementation; as I
understand, jsonpath is the representation used to search for elements
within JSON values. If you replace "including" with "and", the result
seems reasonable.

> PostgreSQL 17 adds a new connection parameter, `sslnegotation`, which allows PostgreSQL to perform direct TLS handshakes when using [ALPN](https://en.wikipedia.org/wiki/Application-Layer_Protocol_Negotiation), eliminating a network roundtrip. PostgreSQL is registered as `postgresql` in the ALPN directory.

Typo here "sslnegotation" missing an i, sslnegotiation.

> PostgreSQL 17 normalizes the parameters for `CALL` in [`pg_stat_statements`](https://www.postgresql.org/docs/17/pgstatstatements.html), reducing the number of entries for frequently called stored procedures. Additionally, [`VACUUM` progress reporting](https://www.postgresql.org/docs/devel/progress-reporting.html#VACUUM-PROGRESS-REPORTING) now shows the progress of vacuuming indexes. PostgreSQL 17 also introduces a new view, [`pg_wait_events`](https://www.postgresql.org/docs/17/view-pg-wait-events.html), which provides descriptions about wait events and can be combined with `pg_stat_activity` to give more insight into why an active session is waiting. Additionally, some information in the [`pg_stat_bgwriter`](https://www.postgresql.org/docs/17/monitoring-stats.html#MONITORING-PG-STAT-BGWRITER-VIEW) view is now split out into the new [`pg_stat_checkpointer`](https://www.postgresql.org/docs/17/monitoring-stats.html#MONITORING-PG-STAT-CHECKPOINTER-VIEW) view.

Note use of one link to "/devel/" here.

--
Álvaro Herrera 48°01'N 7°57'E — https://www.EnterpriseDB.com/
"Always assume the user will do much worse than the stupidest thing
you can imagine." (Julien PUYDT)

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Jonathan S. Katz 2024-05-23 12:22:49 Re: PostgreSQL 17 Beta 1 release announcement draft
Previous Message Ranier Vilela 2024-05-23 11:54:12 Re: Avoid possible dereference null pointer (src/backend/catalog/pg_depend.c)