Flyte enables production-grade orchestration for machine learning workflows and data processing created to accelerate local workflows to production.

Flyte

I have a workflow that has a conditional task A that executes if some parameter is true. Then there is a task B that needs to execute after A.
Is there a way to represent this dependency without defining a sub-workflow for A?
Looks like I can't represent this as condition &gt;&gt; node (syntax error)

is it like this
```if true:
   A()
   B()
else:
   # do nothin```
is it like this
```if true:
   A()
B()```

The 2nd option
But I need to make sure A() runs before B()

Another behavior that I just noticed
```variable c_out = C()
if true:
    A(c_out=c_cout)```
A started running before C was done. Is that expected?

<@U03HQE6THNV> both of those should just work

<https://docs.flyte.org/projects/cookbook/en/latest/auto/core/control_flow/conditions.html#example-3|Example# 3> - shows using outputs from condition to chain

<https://docs.flyte.org/projects/cookbook/en/latest/auto/core/control_flow/conditions.html#example-4|Example #4 shows using outputs from previous to a condition>

if for example 3  you do not want to pass outputs and use some chaining then

<@U03HQE6THNV> I just tried this, it worked

```from flytekit import task, workflow, conditional

@task
def foo():
    print("Called")


@task
def foo2():
    print("Called")

@workflow
def wf(b: bool):
    c = conditional("blah").if_(b.is_true()).then(foo()).else_().fail("haha")
    c &gt;&gt; foo()```

I had foo_node = create_node(foo)
c &gt;&gt; foo_node gave me the error

Here the type of c is a promise, but the type of foo_node is a node. Therefore, it doesn’t work.
It’s okay to use `c &gt;&gt; foo()` without creating node
```foo_node = create_node(foo)
c &gt;&gt; foo_node```