Adding Series Dependency to Patchwork
Ferruh Yigit
ferruh.yigit at amd.com
Fri Jul 19 19:40:51 CEST 2024
On 7/12/2024 9:15 PM, Adam Hassick wrote:
> Hi All,
>
> We've gotten a review of our initial submission to add the dependency
> parsing to the Patchwork dashboard. Stephen recommends that we change
> our format to use the message ID of patches or cover letters rather
> than the ID of the patch or series in the database. So, instead of
> adding a dependency by adding "Depends-on: series-5678" one would add
> "Depends-on: <20240712120000-1-user at example.com>". We can keep the
> option of using the patchwork web URLs that was discussed in the
> original issue on GitHub.
>
I wasn't aware that patchwork has target to abstract/hide the integer
IDs, in this case since we want to enable this feature in the patchwork,
it makes sense to follow their path.
No need to push for integer IDs, that was something we thought as an
option, it is not crucial for us, we can change it.
But there is another thing Stephen mentioned, perhaps using URL instead
of message ID.
Even message ID gives more data, it may still require some effort for
someone investigating the patch to access to that dependent
patch/series. But using URL gives instant access.
If using URL doesn't make like harder for the patchwork implementation,
I am for using URL. From patch developer perspective, I think providing
message ID or patchwork URL are similar level of hassle.
If we go with the URL option, does is still required to differentiate as
"patch-xxx" or "series-yyy", previously they were different IDs, but
with URL can patchwork deduce if it is series or patch? If so this can
bring a simplification.
> The main reasoning for this is that our format doesn't make it clear
> exactly what the dependency is or where it's found outside of the
> context of Patchwork. This discussion can be viewed here:
> http://patchwork.ozlabs.org/project/patchwork/patch/20240617221900.156155-3-ahassick@iol.unh.edu/
>
> Does this change sound reasonable?
>
> Regards,
> Adam
More information about the ci
mailing list