Prepare 1.2.0 release #12933
Prepare 1.2.0 release #12933
Conversation
|
One or more of the following people are relevant to this code:
|
There was a problem hiding this comment.
We might have had some issues with the migration of release notes. I didn't see: https://github.com/Qiskit/qiskit/blob/main/releasenotes/notes/default-level-2-generate-preset-passmanager-ec758ddc896ae2d6.yaml in 1.2.0, but it definitely should be there.
|
It seems that PR didn't make it into the |
releasenotes/notes/1.2/circuit-gates-rust-5c6ab6c58f7fd2c9.yaml
Outdated
Show resolved
Hide resolved
releasenotes/notes/1.2/add-random-clifford-util-5358041208729988.yaml
Outdated
Show resolved
Hide resolved
| - To expose backend information with no access to execution (just a hardware description), | ||
| consider constructing a :class:`.Target` directly. | ||
| - To provide access to execution capabilities, consider relying on the primitives interfaces | ||
| instead. | ||
| - Alternatively, to continue providing simultaneous :class:`.Target` (hardware information) | ||
| and ``run`` (execution) capabilities, consider moving to :class:`.BackendV2` | ||
| (see `the migration guide<https://qisk.it/backendV1-to-V2>`_). |
There was a problem hiding this comment.
This is a weird suggestion bulleted list, the normal workflow would be to just have people migrate to BackendV2. Like I get there is a narrative around using primitives for execution, but migrating to BackendV2 isn't about execution and a target. You can think of it as providing an interface for building targets. Just saying build a target isn't useful if you're maintaining a provider.
releasenotes/notes/1.2/deprecate_assemble-67486b4d0a8d4f96.yaml
Outdated
Show resolved
Hide resolved
releasenotes/notes/1.2/linear-binary-matrix-utils-rust-c48b5577749c34ab.yaml
Outdated
Show resolved
Hide resolved
releasenotes/notes/1.2/port_star_prerouting-13fae3ff78feb5e3.yaml
Outdated
Show resolved
Hide resolved
releasenotes/notes/1.2/port_star_prerouting-13fae3ff78feb5e3.yaml
Outdated
Show resolved
Hide resolved
| --- | ||
| fixes: | ||
| - | | ||
| Fixed a missing decorator in :class:`.C3SXGate` that made it fail if :meth:`.C3SXGate.to_matrix` | ||
| was called. The gate matrix is now return as expected. |
There was a problem hiding this comment.
Shouldn't we delete these and the other 1.1.x release notes that are erroneously being detected as part of the 1.2.0 release since we've already documented the fixes?
There was a problem hiding this comment.
Yep, they're already deleted locally, I didn't push it to not interrupt CI on that last PR 🙂
| @@ -13,25 +13,19 @@ features_misc: | |||
| decreases the memory overhead of a :class:`.CircuitInstruction` and | |||
| :class:`.DAGOpNode` object at the cost of decreased runtime on multiple | |||
There was a problem hiding this comment.
This probably should be "increased runtime" since it gets slower when you build without a cache.
| @@ -1,7 +1,13 @@ | |||
| --- | |||
| deprecations_providers: | |||
| - | | |||
| The `Qobj` structure and related classes are now deprecated, they were introduced as part of the `BackendV1` workflow and are no longer necessary for interacting with :class:`.BackendV2` backends. Remote backend interaction should be done via `QPY` or `OpenQASM` instead. | |||
| The Qobj structure and related classes (see :mod:`qiskit.qobj`) are now deprecated. They were | |||
| introduced as part of the :class:`.BackendV1` workflow and are no longer necessary for | |||
There was a problem hiding this comment.
It was actually part of BaseBackend (i.e. BackendV0) the concrete difference between BaseBackend and BackendV1 was BaseBackend only took a Qobj as the input for job submission while BackendV1 used a QuantumCircuit. But this is probably just a historical default and not worth updating here.
| interacting with :class:`.BackendV2` backends. Remote backend interaction should be done via | ||
| QPY (see :mod:`qiskit.qpy`) or OpenQASM (see :mod:`qiskit.qasm2` and :mod:`qiskit.qasm3`) instead. |
There was a problem hiding this comment.
I would say:
| interacting with :class:`.BackendV2` backends. Remote backend interaction should be done via | |
| QPY (see :mod:`qiskit.qpy`) or OpenQASM (see :mod:`qiskit.qasm2` and :mod:`qiskit.qasm3`) instead. | |
| interacting with :class:`.BackendV2` backends. :class:`.QuantumCircuit` serialization should be done via | |
| QPY (see :mod:`qiskit.qpy`) or OpenQASM (see :mod:`qiskit.qasm2` and :mod:`qiskit.qasm3`) instead. |
| then ``max_operands`` will default to 4 and a random circuit with a random gate distribution will | ||
| be generated. If both ``num_operand_distribution`` and ``max_operands`` are specified at the same | ||
| time then ``num_operand_distribution`` will be used to generate the random circuit. | ||
|
|
||
| Example usage:: |
There was a problem hiding this comment.
We could use the plot directive here if we want a circuit visualization showing the random circuit instead of just the code.
| prelude: > | ||
| The Qiskit 1.2.0 release focuses mainly on improving performance | ||
| and quality of the compiler. Also, it is the last release supporting | ||
| Python 3.8. By Qiskit 1.3.0 the minimal required Python version |
There was a problem hiding this comment.
| Python 3.8. By Qiskit 1.3.0 the minimal required Python version | |
| Python 3.8. In Qiskit 1.3.0 the minimal required Python version |
Summary
Prepare the 1.2.0 release.
Details and comments
Opening this already so people can have a look at whether I missed any typos 🙂 Prelude is currently a draft, maybe we should put some more concrete numbers on the performance improvements and mention some new features.