Fix uncurried stream ops in javadsl

[mdedetrich]

(cherry picked from commit f9ad446)

Since this change has been determined to be completely safe (see #1406 (comment) and thread in general) I think its a good candidate to cherry pick back to the 1.0.x series since if we ever do another release in the 1.0.x series it will mean that the docs/source/syntax completion will be better

[He-Pin]

I think we better keep it in 1.1.x, @pjfanning ,wdyt? I don't have strong option on this.

[pjfanning]

I'd prefer to hold off on this. Can we focus on getting 1.1.0 released?

[mdedetrich]

I'd prefer to hold off on this. Can we focus on getting 1.1.0 released?

Well reviewing/merging this isn't going to have an impact on 1.1.0 one way or another

Regarding 1.1.0, I am not sure if there was a discussion on the release plan (I have been a bit out of the loop as of late), or more specifically is the idea that we are meant to wait for a bit just to be sure that its stable (we are currently using 1.1.0-M1 in production), if so its best to wait to get feedback from a critical amount of users.

[pjfanning]

I'd prefer to hold off on this. Can we focus on getting 1.1.0 released?

Well reviewing/merging this isn't going to have an impact on 1.1.0 one way or another

Regarding 1.1.0, I am not sure if there was a discussion on the release plan (I have been a bit out of the loop as of late), or more specifically is the idea that we are meant to wait for a bit just to be sure that its stable (we are currently using 1.1.0-M1 in production), if so its best to wait to get feedback from a critical amount of users.

My point is that we have been very conservative with 1.0.x changes. I would prefer not to merge this to 1.0.x. We can have the 1.1.0 release discussion on the mailing list. There are a few open issues but we could probably consider releasing 1.1.0-M2 soon or even doing 1.1.0 instead.

[mdedetrich]

My point is that we have been very conservative with 1.0.x changes. I would prefer not to merge this to 1.0.x.

Sure, but conservative doesn't mean "none at all" and at least currently without these changes, for cross Java/Scala projects the effected parts in javadsl is currently broken when it comes to IDE's syntax highlighting and considering that it has zero impact on bytecode (and people are using 1.0.x) I think that this should pass the bar, we have certainly merged things into 1.0.x on a lot less than this.

Lets keep the 1.1.0 release discussion out of this issue, there wasn't any reason to bring it up as it has nothing to do with this.

[He-Pin]

I'm not using Kotlin with pekko-stream, but the current one works ok with Java, even with Java21, do you have a screenshot where is cause problem when work with Kotlin @mdedetrich

[He-Pin]

@mdedetrich should this retarget 2.0.0？

[He-Pin]

@pjfanning @mdedetrich what's the status of this?

[He-Pin]

Cherry-pick of uncurried stream operations to the 1.0.x branch.

1. Backport rationale: As the author notes, this is a safe change that improves docs/source/syntax completion for users. Since it only adds overloads (deprecating old ones), it is backward compatible.

2. Target branch: Correctly targeting 1.0.x. However, note that 1.0.x is a maintenance branch -- is there still planned to be another 1.0.x release? If not, this cherry-pick may not be necessary.

3. Staleness: Open since July 2024 with no reviews. If there are plans for another 1.0.x release, this is a safe and useful improvement. Otherwise, consider closing if 1.0.x is no longer being released.