Re: meson documentation build open issues

From: Peter Eisentraut <peter(at)eisentraut(dot)org>
To: Andres Freund <andres(at)anarazel(dot)de>, Christoph Berg <myon(at)debian(dot)org>, Andrew Dunstan <andrew(at)dunslane(dot)net>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: meson documentation build open issues
Date: 2023-11-14 20:16:13
Message-ID: 4ede56ab-4370-4d71-b566-b9f450128343@eisentraut.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Some comments on your patches:

v2-0001-meson-Change-default-of-selinux-feature-option-to.patch

Ok

v2-0002-docs-Document-with-selinux-Dselinux-options-centr.patch

Ok, but "selinux" should be "SELinux" when referring to the product.

v2-0003-meson-docs-Add-doc-html-man-targets.patch

We have make targets "html" and "man", so I suggest we make the meson
targets the same.

v2-0004-meson-Add-world-target.patch

AFAICT, this world target doesn't include the man target. (Again, this
would all work better if we added "man" to "docs".)

v2-0005-docs-meson-Add-documentation-for-important-build-.patch

It's nice to document this, but it's weird that we only document the
meson targets, not the make targets.

v2-0006-meson-Add-help-target-build-docs-from-a-common-so.patch

Here also, making this consistent and uniform with make would be useful.

v2-0007-meson-Add-Dpkglibdir-option.patch

Normally, the pkgFOOdir variables are just FOOdir plus package name. I
don't feel comfortable allowing those to be separately set. We don't
allow that with configure; this just arose from a Debian patch.

The description "location to dynamically loadable modules" is too
narrow. Consider for example, another proposed patch, where we are
doing some preprocessing on postgres.bki at build time. Since that
makes postgres.bki platform dependent, it should really be moved from
share (datadir in configure parlance) to pkglibdir. So then we have
things in there that are not loadable modules. I don't know how that
affects Debian packaging, but this patch might not be the right one.

I suggest we leave this patch for a separate discussion.

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Nathan Bossart 2023-11-14 20:22:52 Re: retire MemoryContextResetAndDeleteChildren backwards compatibility macro
Previous Message Andres Freund 2023-11-14 20:01:47 Re: pgsql: doc: fix wording describing the checkpoint_flush_after GUC