-
Notifications
You must be signed in to change notification settings - Fork 79
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Move Aries SPIFly bundle after simpleconfigurator to start it afterwards #1252
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't think this is sufficient. Relying on the implementation to write it in the order of the product (and Equinox reading and starting/install it in that order) is waiting for problems to happen. Beside this, it would not be possible to update spifly as far as I know.
Why not move spyfly to level 2 as its done with ds, event and alike?
Changing the order in the configuration section didn't Change the generated config.ini.
Yes that's the only releiable thing left. We have to hope that then all bundels that need the Mediator are activated in a later level. To migitate that we should switch to a Framework-Extension in the next release, ideally one that does not need asm. :) Since we already missed the deadline we need Project-lead or PMC approval. Please note that this does not affect any bundle. |
Is this a problem with some new feature? |
Not directly a feature but with a new configuration necessary for slf4j-2 and IIRC now also others that use ServiceLoader.
Of course, it was #1195 |
Does the #1195 actually work?
(note: I've set |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If that fixes the error, let's merge.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good.
The
|
It would be good to add this to N&N section. |
Is by accident also not slf4j 1.7 installed (so both?) I think you will see this double output |
.
We use in our product even lvl 3 already for years for this, but I guess it greatly depends on where slf4j is used in more core stuff. What I do wonder why Aries needs asm.. not for stuf where we really use it for right? I have a feeling that this service loader thing should really just be done in the core implementation of osgi.. (so not even needing Aries) |
The Eclipse SDK neither uses SLF4J loggers nor service loaders but the earlier it is started the better, just to be sure.
Aries transforms the class-files of classes that use the ServiceLoader in Order to inject special classloaders that consider the OSGi environment.
Me too. You might be interested in eclipse-equinox/equinox#295 |
SLF4J-simple is not used in all packages, many use logback through m2e.logback, therefore I'm not sure that's necessary respectivly could be confusing. |
But how the users of SLF4J-simple should know about the way to configure SDK or more general: that they can now configure SLF4J-simple in the SDK? Or do I miss something and it was always possible but then I don't understand the change anymore? |
I don't know for sure how it was with SLF4J from Orbit was, but from a quick look at https://download.eclipse.org/eclipse/updates/4.25/R-4.25-202208311800/ there was no binding at all. Since slf4j from Maven there was only the noop Binding/Provider present. So no it was'nt possible. For the SDK it would probably be good yes, but I hestitate a bit because in some EPP packages the logback logging backend supplied by m2e.logback is used, which supplied a logback configuration. |
Otherwise requirements (such as org.objectweb.asm) are not available. See eclipse-platform/eclipse.platform.releng.aggregator#1252
Otherwise requirements (such as org.objectweb.asm) are not available. See eclipse-platform/eclipse.platform.releng.aggregator#1252
Fixes eclipse-equinox/equinox#302 (hopefully).