Hi Quentin,

On 11/24/21 7:47 PM, Quentin Schulz wrote:

Hi Michael, Richard,

On Wed, Nov 24, 2021 at 06:10:56PM +0000, Richard Purdie wrote:

On Wed, 2021-11-24 at 18:16 +0100, Michael Opdenacker wrote:

Together with the corresponding Bitbake version, which are no
longer supported.

Signed-off-by: Michael Opdenacker <michael.opdenacker@...>
---
scripts/run-docs-build | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/scripts/run-docs-build b/scripts/run-docs-build
index 3db4a97..5e1d649 100755
--- a/scripts/run-docs-build
+++ b/scripts/run-docs-build
@@ -35,7 +35,7 @@ mkdir $outputdir/bitbake/next
cp -r ./_build/final/* $outputdir/bitbake/next

# stable branches
-for branch in 1.46 1.48 1.50 1.52; do
+for branch in 1.46 1.50 1.52; do
git checkout $branch
make clean
make publish
@@ -68,7 +68,7 @@ mkdir $outputdir/next
cp -r ./_build/final/* $outputdir/next

# stable branches
-for branch in dunfell gatesgarth hardknott honister; do
+for branch in dunfell hardknott honister; do
cd $ypdocs
git checkout $branch
make clean

I'm a bit torn on this. They are no longer officially supported releases now but
it may make sense to rebuild all the sphinx docs in this script rather than some
subset?

I think we want to make sure we have all docs up-to-date, even for the
branches that aren't maintained anymore. Especially since it's not
taking a lot of CPU time to build them, it's fine IMO. We could always
make minor changes to old docs. E.g. the releases.rst might get updates
until we figure something out.

Thanks for casting your vote. It makes sense. I'll send another patch
with this decision in the comments.
Cheers
Michael.

--
Michael Opdenacker, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com