Hello everyone! I suggested a while back that Arro...

# arrow-contributors

[Youssef Shoaib [MOD]]

Hello everyone!
I suggested a while back that Arrow should likely have some library for context-related utilities. Thus, the arrow-context repo was created under the Arrow github. I'm excited to share the first (very rough) prototype of some of the ideas I've implemented for it. This is still a Draft PR, and a lottttt of things are missing, so suggestions are very welcome :)

[Alejandro Serrano.Mena]

this is amazing work!

[Alejandro Serrano.Mena]

but keep in mind that context receivers are still in flux, so the implementation might need to change to keep up with it

[Youssef Shoaib [MOD]]

Absolutely! I've been keeping up with the GitHub issue, and will add/remove things as necessary.

[CLOVIS]

Is the goal of this repository is to put everything related to context parameters that would be a part of arrow-core if they were stable?

[Youssef Shoaib [MOD]]

Practically, yes. Obviously it's currently just a PR by my of what I think should be in there. I think also theoretically if contexts take a while to stabilise, this could be released as an experimental library. That's all up to the Arrow team though of course!

[CLOVIS]

I was confused because the PR looks a lot like arrow-core, but now that makes perfect sense 😂

[Youssef Shoaib [MOD]]

It relies on arrow-core right now even. It's set up so that 1) it mirrors the basic structure of core, so that it's simple to add some of the things right into there, and 2) so that it could be released as a separate library if arrow sees fit

[raulraja]

My personal opinion is that if context receivers become non experimental and stable everything related to core with context receivers should live in arrow core. @Youssef Shoaib [MOD] is that what you were thinking or you wanted a separate module for this ?

[Youssef Shoaib [MOD]]

Yeah that's exactly what I'm thinking @raulraja. The goal of this repo is to be a draft for that design, but also in case context receivers take even longer, this could be released as an experimental library if you guys see fit, and then of course when contexts are stable it'd go away

[Javier]

Correct me @Alejandro Serrano.Mena if I am wrong, but as context receivers/params design is changing, it should be normal they keep in experimental mode for more time so I would expect this library to be published “asap” (first moment it is ready).

[Alejandro Serrano.Mena]

it may not neccessarily be that way: definitely the new design won't be there until 2.1 (since 2.0 is already in Beta), and in theory it could "graduate" to stable in 2.2

[Javier]

2.2 would be mid or end of this year?

[Alejandro Serrano.Mena]

honestly this is something I don't know

[Youssef Shoaib [MOD]]

Kotlin is on a 6-month release cycle, no? So it would be end of year presumably.

[Alejandro Serrano.Mena]

yes, but 2.0 is kind of a special release because of the move to K2, so it may affect deadlines

[simon.vergauwen]

in theory it could

I'll take it 😄

[Javier]

Even if 2.0 is special and would be available soon, 2.1 would be on summer more or less so 2.2 should be end of the year, no? And that would mean three releases in 1 year, so even better than thinking in 6 months release