currently, in oe-core packagegroup files, there are a number of
recipe files that use what i consider the "right" approach and use
${PN} to refer to the packagegroup name when defining the
packagegroup components, as in:
RDEPENDS_${PN} = "\
nativesdk-pkgconfig \
nativesdk-qemu \
... snip ...
RDEPENDS_${PN}_darwin = "\
and so on, and so on.
OTOH, there's also a lot of this (spelling out the packagegroup name
in full):
RDEPENDS_packagegroup-core-buildessential = "\
autoconf \
automake \
... snip ...
is it worth doing a quick cleanup of that to shorten all that? on the
one hand, it *sounds* like unnecessary churn, but on the other hand, i
think using ${PN} is a useful visual clue in that, if there is a full
packagegroup name in the recipe file, that should be a hint that
there's a *reason* for it to be there.
however, two issues with this.
first, in packagegroup file packagegroup-base.bb, there is the
"unfortunate" naming of:
PACKAGES = ' \
packagegroup-base \
packagegroup-base-extended \
packagegroup-distro-base \ <---
packagegroup-machine-base \ <---
where those two can't be abbreviated given the name structure. oh,
well.
also, in that same recipe file, i assume that one could abbreviate
all the full names in
PACKAGES = ' \
... snip ..
${@bb.utils.contains("MACHINE_FEATURES", "alsa", "packagegroup-base-alsa",
"", d)} \
as in, the above could be abbreviated as:
${@bb.utils.contains("MACHINE_FEATURES", "alsa", " ${PN}-alsa", "", d)} \
yes?
anyway, is is worth the churn? i will defer.
rday
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#136834):
https://lists.openembedded.org/g/openembedded-core/message/136834
Mute This Topic: https://lists.openembedded.org/mt/72635393/21656
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-
