No love for Dart protobuf?

[hagerf]

Hi there Dart team at Google!

I've been using protobufs with Dart for the last 2 years and while it's working well in the current implementation, I've noticed that this library hasn't been worked on for a while. The last release was 16 months ago and the last commit (excluding automatic) was over 10 months ago.

There are also a growing number of issues being raised that have no response at all, and some of those are quite serious questions e.g. regarding support for editions (a Protobuf feature!). Last response from any maintainer to any issue was (as far as I see) on March 30th. Responding to questions regarding user errors or similar I can understand is a chore and maybe not necessary, but many of these requests are quite reasonable, i.e. using proper Dart enums, using sealed classes and Uint8Lists instead of lists of integers.

Then there are multiple pull requests that either seem abandoned by maintainers or haven't been reviewed for several months like this one.

I guess I would just like to hear from the Dart team at Google if there is any reason for this (maybe you're waiting for a Dart feature like macros to be finished or similar?) and if there are any plans on giving this library some love so that it would feel more like an up-to-date and current library rather than its current slightly outdated feel.

Thanks, and thanks for your work.

[daviddomkar]

Wondering the same things myself. Waiting for #790 PR for almost 2 years with no response from devs. :/

[mraleph]

Dart team does not have engineering resources to actively maintain this library. To the best of our knowledge the package works for most use cases - so it is kinda frozen in this state. If there is some severe bug - we can probably patch things up and release a new version. But at the moment we can't support any feature work, which unfortunately includes any sort of PR reviews as well.

We have been searching for ways to get out of this state for a while now - but sadly no news to share about that.

[hagerf]

@mraleph Ok, thank you for your answer. It was something I think we all suspected but hoped would not be true.

It is a bit sad, but to be honest it's true that the package works for most use cases currently. But as the both Dart and Protobuf evolves this will probably not be the case forever (most critically supporting editions).
Anyways I hope this situation can be resolved, either under the google umbrella or maybe somewhere else.

[bivens-dev]

Dart team does not have engineering resources to actively maintain this library. To the best of our knowledge the package works for most use cases - so it is kinda frozen in this state. If there is some severe bug - we can probably patch things up and release a new version. But at the moment we can't support any feature work, which unfortunately includes any sort of PR reviews as well.

We have been searching for ways to get out of this state for a while now - but sadly no news to share about that.

Granted, I’m sure this has already been considered and found to be not suitable for some reason but I was wondering if @mraleph or @osa1 could share some insights on the path of using FFI to wrap the C++ core library for the server-side implementation and JS interop to wrap gRPC-web as a way of reducing both the maintenance burden while also keeping up with all the latest and greatest features perhaps most notably that protobuf is moving to a versioned format and will be taking much of the ecosystem with it leaving Dart in an unfortunate position cut off from all of that.

[xentopos]

@mraleph @bivens-dev

Wouldn't another option be to find a maintainer from the community? This is kind of a critical interop library, and while it might be working for most use cases, there are a number of ways where it's currently suboptimal (e.g. having to check presence for optional fields in a language with explicit null-safety checks, with open PRs which have been languishing for over a year).