On 07.12.21 18:26, Martin Jansa wrote:
On Tue, Dec 07, 2021 at 06:13:45PM +0100, Konrad Weihmann wrote:
Hi there,
following up on a discussion we had at [1]...
the correct fix would be
magic.bbclass
---
DEPENDS:append = " magic-dependency"
---
but unfortunately sometimes this is out of our control (as we don't want to
have bbclassappends for totally valid reasons)
So instead of correctly fixing a few bbclasses (which you can still easily
overlay in your layer if their upstream rejects using :append there for some
very strange reason - I still haven't seen any evidence that there
are (m)any such globally interitted bbclasses which do this)
you're suggesting to change DEPENDS in every single recipe (where even
more recipes are out of our control) with _incorrect_ fix to support this
scenario?
No I'm definitely not suggesting that - I just encountered that issue
when working with a vendor layer - so the fault is clearly on their end.
*but* it took me in total almost a day to fix the reason for the
behavior, that's why I want to have a broader discussion on this.
If the outcome is "we are not responsible for wrongly coded bbclasses",
I'm totally fine with that (and also instruct the linter, that was
mentioned in the original issue, to ignore these cases).
And as there is not much of builtin quality checks for these kind of
code injection, I would like to hear opinions on how to/if we should at
all tackle this issue.
Cheers,
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#1374):
https://lists.openembedded.org/g/openembedded-architecture/message/1374
Mute This Topic: https://lists.openembedded.org/mt/87569589/21656
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-architecture/unsub
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-
