<@U2JKKPMEE> <https://github.com/InsertKoinIO/koin...
# koin-contributors
b
@arnaud.giuliani https://github.com/InsertKoinIO/koin/pull/692 - First draft. I will add changes to this PR, according to the list.
DSL configuration until a little dirty - later cleaned up the code.
a
@BorzdeG what do you mean by
migrate to Kotlin DSL
?
b
@arnaud.giuliani Moving Gradle configurations from Groovy DSL to Kotlin DSL
Groovy DSL script files use the
.gradle
file name extension. Kotlin DSL script files use the
.gradle.kts
file name extension.
a
ok, what real need behind that? apart of using Kotlin DSL?
is it really required for MP project?
b
No, not necessarily. But I focus on 2 points: 1) Projects “only Kotlin” is better to write entirely on Kotlin. What Kotlin DSL includes for Gradle 2) https://discuss.gradle.org/t/gradle-dsl-groovy-vs-kotlin/25242/5
@arnaud.giuliani and using the plugin https://github.com/jetbrains/gradle-idea-ext-plugin you can configure IDEA in the “basic” version - register the settings required for the project to work correctly in IDEA
Although, it works for Groovy DSL too ...
Kotlin DSL also brings such a feature as static typing
a
yeah, I know. But it's to avoid too much work at once 😕
and focus on making the Kotlin MP first
b
I plan to transfer all modules to MPP over the next week. In PR I corrected the order of tasks. Test transfer to KotlinTest will make a separate PR
👍 1
recreated PR: https://github.com/InsertKoinIO/koin/pull/698 I moved the project to MPP configuration. But there is still a problem with generating documentation through Dokka for android modules Also, transfer the first module to the Common module. I did not edit the tests - I will transfer them to Common in the following iterations
a
Thanks for your time @BorzdeG 👍
a such huge PR can be hard to check, could seperate each topic of your PR in a branch please? 😕
would help to check things
b
👍
a
Also, I'm chatting with @kpgalligan from TouchLab also ... he has made several talks about Kotlin multiplatform & kotlin native threading
can be good to decide how we to approach the thing into Koin
b
I left the source code as is - I just moved the modules to the root directory. The main work was done with the Gradle configuration. Code porting was only in the koin-core module
a
ok
b
I have so far selected two small PRs from the main PR. As soon as they are accepted, I’ll do the rest of the PR Left to do: 1) transfer all Android modules to Kotlin DSL 2) Convert modules to MPP configuration
@arnaud.giuliani See these two PRs?
@arnaud.giuliani See the update?
a
I beginning only to check your PR and one from @kpgalligan
I need to double check your work, and decide on what we can go
did you check also branch from Kevin?
your first PR seems ok
I have to check the 2nd one
sorry, a bit ouf of time those days … but it will be ok 👍
k
If I understand the PR’s, they’re just reformating gradle? I’d say < 5% of our gradle is in kotlin dsl at this point. There’s very little to gain from doing that, from a technical perspective, although I don’t discourage anybody from doing so.
I find libraries are more difficult to manage with kotlin dsl and KMP, mostly due to lack of good examples, especially around publishing.
Those diffs are massive. It looks like all the files are being moved from ‘koin-projects’ to the root? I would definitely do that in multiple steps. Doing that in one big PR while also moving to kotlin dsl seems unnecessary.
b
The publication is implemented in PR. Kotlin DSL has static typing and auto tips in IDEA Moving to the root without moving to the Kotlin DSL was in the previous PR: https://github.com/InsertKoinIO/koin/pull/703
k
I’m going to get back to the branch I was working on soon. I’d like to add the multiple state options for jvm, and then discuss if there’s anything else significant, but in general around thread/state design, I think it’s heading in a good direction.
Yeah, OK, but why?
b
Why what?
k
Why change the project structure so dramatically? Merging back in my branch (or anybody else’s) is going to be a nightmare, and you gain nothing from it.
After that, I am in the middle of a KMP rearchitecture. We should probably have some consensus on what that’s going to look like, or there will be several independent efforts.
To be clear, not trying to be critical, but we should coordinate a bit. In my opinion, moving all the files is not a priority. Just those necessary for the KMP project structure, and only in core.
a
Yeah, merging @kpgalligan’s work is too much conflicting here. Maybe we can first focus on merging your work Kevin and then we can go forward
I’m still a bit tedious with Gradle Kotlin DSL 😕
@BorzdeG let’s put your PRs on hold while we can converge on Koim MP work
there is important design decision choice and we should avoid adding more around that