Sounds good. Thanks for the feedback.
Do you have any suggestions for the name? Currently the class name is
called AsyncWrapper and the file name is named async_dofn.py. Would
async_wrapper_dofn be better?

On Tue, May 19, 2026 at 9:57 AM Reuven Lax <[email protected]> wrote:

> I'm pretty sure the Scio s implementation is working different. Hence the
> reason why I suggest we call this something other than AsuncDoFn.
>
> On Tue, May 19, 2026, 4:09 AM Kellen Dye <[email protected]> wrote:
>
>> There is an implementation of this already in scio that could be
>> donated/modified/etc. cc @Claire McGinty <[email protected]>
>>
>> Cheers,
>> Kellen
>>
>>
>> On Mon, May 18, 2026 at 11:12 PM Reuven Lax via dev <[email protected]>
>> wrote:
>>
>>> Can we name it something else? This doesn't seem like AsyncDoFn to me,
>>> and we should probably have named the Python version something else
>>> initially.
>>>
>>> On Mon, May 18, 2026 at 1:59 PM Tejas Iyer via dev <[email protected]>
>>> wrote:
>>>
>>>> Hello,
>>>>
>>>> I'm Tejas, and I am planning to implement an asynchronous DoFn wrapper
>>>> for the Apache Beam Java SDK. Currently, this capability exists in the
>>>> Python SDK, and I would like to mirror this functionality in Java to allow
>>>> developers write and manage asynchronous DoFns in their pipelines.
>>>>
>>>> I have created a feature request to track this work here:
>>>> https://github.com/apache/beam/issues/38529
>>>>
>>>> Because this is a straightforward implementation based on the existing
>>>> Python SDK, it shouldn't warrant a full design doc, but I wanted to give
>>>> the community a heads-up.
>>>>
>>>> Please let me know if you have any early feedback, concerns, or if
>>>> anyone would be interested in reviewing the code once the PR is ready.
>>>>
>>>> Thanks,
>>>> Tejas
>>>>
>>>

Reply via email to