You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Feature - Bump the version of our stacks-core dependency
1. Description
The signer binary pulls in crates from the a fork of the stacks-core repo but from a slightly modified version of the develop branch. This code is likely out of date with the latest Nakamoto changes so we will likely need to update it.
1.1 Context & Purpose
We need the latest version of stacks-core crates so that there is no discrepancy of the types when we communicate with stacks core.
We currently cannot rely on the official stacks-core crates because of a dependency conflict with sqlx. Our sqlx dependency pulls in sqlite as a dependency because the macros and/or migrate features pull it in, and our libstacks dependency pulls in sqlite as well. There is a conflict because the versions don't match and they need to match exactly (because sqlite is basically a C dependency). There is no way around this conflict without doing one of the following:
Maintain our own version of stacks-core with a version of rusqlite that is compatible with the version used by sqlx. This is what we currently do.
Update stacks-core's to a version of rusqlite that is compatible with the version used by sqlx.
Refactor stacks-core so that the sqlite dependency is always optional. It's supposed to be optional for many of the crates there, just not for the libstacks crate.
Change sqlx so that we do not pull in the sqlite dependency ever (either the official crate our by fixing it and using our own version).
Remove our sqlite dependency from sqlx by not using the features that pull in sqlite.
I don't want to do (1) anymore but it's the lowest amount of work. I'd love to do (3) but that is probably out of the question. (5) is the second lowest lift, but requires us to give up on all sqlx macros.
2. Technical Details:
2.1 Acceptance Criteria:
Use the latest version of the stacks-core crates.
3. Related Issues and Pull Requests (optional):
The text was updated successfully, but these errors were encountered:
stacks-core recently updated their dependencies (on the develop branch) to use the latest rusqlite, which pulls in the latest libsqlite-sys. The latest version of sqlx, version 0.8.0, uses the same libsqlite-sys version.
We need to remove the migrate, json, and time sqlx features in order to not make it seem like we don't pull in libsqlite-sys. After refactor: automatically set created_at columns #436 was merged, we don't actually use the json and time features.
Feature - Bump the version of our stacks-core dependency
1. Description
The signer binary pulls in crates from the a fork of the stacks-core repo but from a slightly modified version of the develop branch. This code is likely out of date with the latest Nakamoto changes so we will likely need to update it.
1.1 Context & Purpose
We need the latest version of stacks-core crates so that there is no discrepancy of the types when we communicate with stacks core.
We currently cannot rely on the official stacks-core crates because of a dependency conflict with
sqlx
. Oursqlx
dependency pulls in sqlite as a dependency because themacros
and/ormigrate
features pull it in, and our libstacks dependency pulls in sqlite as well. There is a conflict because the versions don't match and they need to match exactly (because sqlite is basically a C dependency). There is no way around this conflict without doing one of the following:sqlx
. This is what we currently do.sqlx
.sqlx
so that we do not pull in the sqlite dependency ever (either the official crate our by fixing it and using our own version).sqlx
by not using the features that pull in sqlite.I don't want to do (1) anymore but it's the lowest amount of work. I'd love to do (3) but that is probably out of the question. (5) is the second lowest lift, but requires us to give up on all
sqlx
macros.2. Technical Details:
2.1 Acceptance Criteria:
3. Related Issues and Pull Requests (optional):
The text was updated successfully, but these errors were encountered: