Performance: Add links to downsampling tutorials

[amotl]

About

Add two more references to tutorials on the community forum. Thanks, @hlcianfagna and @hammerhead.

• Performance: Add link to "Resampling time-series data with DATE_BIN"
• Performance: Add link to "Advanced downsampling with the LTTB algorithm"

Preview

• https://cratedb-guide--45.org.readthedocs.build/performance/selects.html#downsampling-with-date-bin
• https://cratedb-guide--45.org.readthedocs.build/performance/selects.html#downsampling-with-lttb

/cc @surister, @karynzv, @marijaselakovic, @seut, @wierdvanderhaar

[amotl]

@hammerhead: This SQL statement probably also has a formatting hiccup, like #41 (review)?

[amotl]

Should we finally employ a proper SQL linter / formatter for all the technical writing endeavours?

[amotl]

Reformatted using https://sqlformat.org/, then edited manually a bit. Is it a good choice?

Suggested change

	WITH downsampleddata AS
	( SELECT lttb_with_parallel_arrays(
	array(SELECT n FROM demo ORDER BY n),
	array(SELECT reading FROM demo ORDER BY n)
	,100) AS lttb)
	SELECT unnest(lttb['0']) as n, unnest(lttb['1']) AS reading
	FROM downsampleddata;
	WITH downsampleddata AS
	(SELECT lttb_with_parallel_arrays(
	array(SELECT n FROM demo ORDER BY n),
	array(SELECT reading FROM demo ORDER BY n), 100) AS lttb)
	SELECT unnest(lttb['0']) AS n,
	unnest(lttb['1']) AS reading
	FROM downsampleddata;

[hammerhead]

I would love to have a standard formatting tool that we can agree on. Like (at least in our team) we settled on black for Python, and it's considered the truth™️ that we all stick to. Without the "then edited manually a bit" part you mentioned, as that kind of defeats the purpose.

For community posts, I don't dare to be the one nit-picking on formatting all the time, but as soon as we are talking about the main documentation, there needs to be a consistent style and quality of formatting. With formatting being such a subjective topic, only a tool we all abide to can do that job.

[amotl]

I agree we need to level up. Future improvements, also expanding into cratedb-examples, will/should/may probably include:

• Using ruff across the board.
• Use black[jupyter] for canonicalizing all the Jupyter Notebooks we introduced recently.
• For prose, we are already using Vale and proselint.
• More potential linters: https://markuplint.dev/, https://textlint.github.io/
• For other file formats, https://github.com/super-linter/super-linter has been probed at https://github.com/crate/infrastructure/pull/2688.
• For SQL, the formatter at https://sqlformat.org/ is actually from sqlparse, which we are already using at other places.

=> We probably want to break this out into a different discussion.

[amotl]

I've diverted this into another issue. Thanks.

• https://github.com/crate/tech-writing/issues/396

[amotl]

Thanks for approving. Now that GH-42 is also approved, we only need approval for GH-41, which is at the bottom of the stack, and needs to be processed first. Thanks!