This diff will wait for a decision for all the services in ENG-1703.

@ashoat will be added as a final reviewer after passing current blocking reviewers.

In D4947#143312, @tomek wrote:

It doesn't make too much sense for me in context of the previous diffs. We wanted to have the initialization in another thread and we were afraid that a short lived thread shouldn't be the one that initializes, and now we're blocking until the initialization is done. I guess this whole approach needs some rethinking, because it doesn't feel consistent.
A quick guess is that maybe we should call waitUntilReady inside connect. But still, we need to find a design that is more consistent.

In D4861#143177, @ashoat wrote:

Sorry @karol but I would prefer to wait until D4805 lands (hopefully tomorrow), then remove all of the CMake from here before landing. Given past experience I suspect we will end up forking things... there will be some Backlog task to merge the CMakes and it will keep getting punted...

ESLint Failure is not related to these changes:

fetch-pack: unexpected disconnect while reading sideband packet

Fixing Folly.

Adding D4949 to the stack.

In D4805#141578, @tomek wrote:

Going to resign for now because I'm not a good person to review this diff. Feel free to add me again when this diff is accepted by other reviewers to give a final review.

Failure to build in a Docker will be fixed by copying the cxx-corrosion cmake file in a follow-up update of this diff.

Rebase on parent changes and corrosion-cxx.cmake directory changes.

Moving cmake file to the /shared/cmake folder after we have a consensus in a chat thread about the shared directory structure.

Rebase on master changes, move threads joining before checks.

In D4861#142965, @karol wrote:

@max you've been blocking this for almost a week now, could you please do a review?

Minor fix with removing the ini file from the commit.

Rebase and merge on a master changes.

Deferring this diff until consensus in a dev team chat thread for the new shared items directory structure and then move this file though.

Rebase on master changes.

In D4805#142222, @ashoat wrote:

I think we should have a separate top-level directory that contains Rust/C++/protos shared between native, services, and web

Removing target_sources in favor of add_library only.

In D4893#142119, @ashoat wrote:

Did we make sure the library's license doesn't have a copyleft?

In D4805#142010, @jon wrote:

In D4805#141989, @max wrote:

In D4805#141985, @varun wrote:

Is it possible to move this to the project root directory so I can use it in native as well?

I think we can do this, but we don't have a folder for such things yet. We are sharing the protobuf files from native/cpp/CommonCpp.
We can use something like native/cpp/CommonCpp/CMake/corrosion-cxx.cmake. What do you think @varun? Also cc to @jon, @karol, @tomek.

Sounds good to me

Moving cmake file to the native/cpp/CMake folder.

Minor refactoring to the AmqpManager::connect() to connect after waiting and run first connect outside of the reconnection loop.

Rebase on parent changes.

Rebase/Merge on master changes.
Change constant name from AMQP_RECONNECT_ATTEMPT_INTERVAL to AMQP_RECONNECT_ATTEMPT_INTERVAL_MS.

In D4805#141985, @varun wrote:

Is it possible to move this to the project root directory so I can use it in native as well?

Rebase/merge on master changes.

Merge on parent changes.

Merge on a parent changes.

Rebase on master changes.
Simplify while loop.

In D4752#139305, @tomek wrote:

It seems like the agreement is to keep this but with a different name, because these are not unit tests. That would mean that we should no longer run tests that require the cloud as a part of run_unit_tests.sh.

In D4752#138943, @ashoat wrote:

Cool, so let's maybe make a task to rename it, or alternately we could rename in this diff

In D4752#138779, @karol wrote:

Yes, I checked and these tests are not really unit.

We could just name them gtests or something. They, in fact, use s3 and dynamoDB.

@jon suggested good ideas and I've put them into a possible follow-up for 0.5 as an ENG-1680 task because these changes a breaking and not really relevant to this diff.
The main purpose of this diff is to fix the running of the Tunnelbroker Nix dev environment without breaking changes in a 0.4 version (which is "frozen" for such breaking updates).
But we should consider making the following updates.

Rebase on the parent changes.

In D4540#139298, @tomek wrote:

The first line of crate documentation https://crates.io/crates/a2 says that

HTTP/2 Apple Push Notification Service for Rust using Tokio and async sending.

So this is only for APN, which is reflected in function name sendNotifToAPNS.
The Linear task is about handling all the notifs

Rust notification library integration into the C++ codebase including helper functions.

E.g. in a comment there

We should be able to communicate directly with APNs and FCM.

So what is the plan for FCM? I can't see any info about it.

Changes to use &str instead of String.
Remove sound and badge hardcoding.

As I've created a ENG-1679 for the discussion about the detach() vs join() problem and this current solution in this diff passed tests and solved the main problem we can go with it now and defer our discussion and decision to the ENG-1679 .
What do you think @tomek and @karol ?

In D4740#141765, @ashoat wrote:

There is way too much churn happening on this diff. Have the people involved scheduled a meeting yet?