spantaleev / matrix-docker-ansible-deploy

šŸ³ Matrix (An open network for secure, decentralized communication) server setup using Ansible and Docker
GNU Affero General Public License v3.0
4.9k stars 1.04k forks source link

Removing Sliding Sync proxy role? #3790

Open luixxiul opened 17 hours ago

luixxiul commented 17 hours ago

Playbook Configuration: irrelevant

Matrix Server: irrelevant

Ansible: irrelevant

Problem description:

https://matrix.org/blog/2024/11/15/this-week-in-matrix-2024-11-15/ says:

Work on Sliding Sync ā€“ which provides a significantly faster and more scalable sync experience in Matrix clients ā€“ has moved to focus on native support, and away from the proxy.

The Sliding Sync Proxy on the Matrix.org homeserver will be decommissioned on November 21st, and client support in Element X will be removed on January 17th.

More details for users as well as server and client developers in Matrix.org's latest blog post

And the blog post (https://matrix.org/blog/2024/11/14/moving-to-native-sliding-sync/) says:

The Matrix.org Foundation does not have the resources to keep up maintenance of the proxy service or its codebase, and plans to decommission the proxy from Mid-November and archive the sliding-sync repo.

Recognising that the community needs time to adopt sliding sync natively, Element will keep client support for the old API (MSC3575) until the 17th of January, 2025.

I am wondering what would be the future for Sliding Sync proxy role of this playbook. Will it be removed or kept in the same way as other unmaintained / deprecated components?

Additional context:

spantaleev commented 14 hours ago

Some people's clients are still using it, so keeping it around a while longer is preferable.

Keeping the role in the playbook is not too heavy of a burden, but still.. removing the role after a few months will probably be better, as it would prevent confusion.

The other unmaintained / deprecated components are somewhat different - they are unmaintained, but they still have their use and some (most) of them still work. We have no telemetry, so we don't know how many users each has, but I believe almost all deprecated components are still useful to at least a few people. It would be better if we can figure out which are broken and/or unused (how?), so we can remove dead-weight, but it's a difficult challenge.

As for sliding-sync, after Element X mobile apps (and various forks of them) stop using the old sliding-sync proxy and some time passes (so that people would upgrade/migrate away from these old versions of the apps), we can safely remove the role. I don't see what purpose and to who it might serve.