CircleCI: fix op-node-docker-build

[syntrust]

Error log:

ERROR: process "/bin/sh -c cd da-server && go build -o da-server main.go" did not complete successfully: exit code: 1

The error happens because there is no code to build, so we update the submodules before build.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI.
The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs.
But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

LGTM

[syntrust]

Error log:

ERROR: process "/bin/sh -c cd da-server && go build -o da-server main.go" did not complete successfully: exit code: 1

The error happens because there is no code to build, so we update the submodules before build.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI.
The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs.
But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

LGTM

[syntrust]

Error log:

ERROR: process "/bin/sh -c cd da-server && go build -o da-server main.go" did not complete successfully: exit code: 1

The error happens because there is no code to build, so we update the submodules before build.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI.
The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs.
But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

LGTM

[syntrust]

Error log:

ERROR: process "/bin/sh -c cd da-server && go build -o da-server main.go" did not complete successfully: exit code: 1

The error happens because there is no code to build, so we update the submodules before build.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI.
The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs.
But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

LGTM

[syntrust]

Error log:

ERROR: process "/bin/sh -c cd da-server && go build -o da-server main.go" did not complete successfully: exit code: 1

The error happens because there is no code to build, so we update the submodules before build.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI.
The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs.
But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

LGTM

[syntrust]

Error log:

ERROR: process "/bin/sh -c cd da-server && go build -o da-server main.go" did not complete successfully: exit code: 1

The error happens because there is no code to build, so we update the submodules before build.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI.
The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs.
But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

LGTM

[syntrust]

Error log:

ERROR: process "/bin/sh -c cd da-server && go build -o da-server main.go" did not complete successfully: exit code: 1

The error happens because there is no code to build, so we update the submodules before build.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI.
The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs.
But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

LGTM

[syntrust]

Error log:

ERROR: process "/bin/sh -c cd da-server && go build -o da-server main.go" did not complete successfully: exit code: 1

The error happens because there is no code to build, so we update the submodules before build.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI.
The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs.
But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.

However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

[syntrust]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

[qzhodl]

I’m curious why the tests passed in the PR without cloning the code. Could you clarify?

From what I can figure out, the PR is tested using the Makefile but not with CircleCI. The Makefile defines the dependency path as devnet-up: pre-devnet: submodules. Heresubmodules pulls code from the da-server using the command git submodule update --init.
However, in .circleci/config.yml, this command is only present in the install-contracts-dependencies step used by contract-related jobs. But *-docker-build jobs like op-node-docker-build in this issue build go code in Docker containers and have no submodule introduced before da-server.

Got it. I would suggest re-running the tests involved in the PR to ensure that this change has no unintended side effects.

In order to avoid the possible side effects and keep the impact in the CircleCI scope, I would suggest this change instead. It is also future-compatible for possible new submodules. Idea?

LGTM