I also think there is great value for this type of matrix. Not only will it help the broader community understand the current implementation status, but more importantly I believe it will help encourage them
> But I do have concern that it will be quickly out of date if it covers many projects. I agree this is a valid concern. I think the best way to keep it up to date is crowdsource it / bring attention to it -- and I do think many of the implementations will want to keep their status up to date I have commented on the previous effort[1] and will see if I can help it along Andrew (Apache Arrow + Apache DataFusion PMC) [1] https://github.com/apache/parquet-site/pull/34#issuecomment-2098014168 On Mon, May 6, 2024 at 9:21 PM Gang Wu <[email protected]> wrote: > Hi, > > There was an effort on this: > https://github.com/apache/parquet-site/pull/34 > > It would be good if we can have something like what Apache Arrow does: > - https://arrow.apache.org/docs/status.html > - > https://arrow.apache.org/docs/cpp/parquet.html#supported-parquet-features > > But I do have concern that it will be quickly out of date if it covers many > projects. > > Best, > Gang > > On Tue, May 7, 2024 at 4:05 AM Prem Sahoo <[email protected]> wrote: > > > +1 on this > > Sent from my iPhone > > > > > On May 6, 2024, at 2:27 PM, Vinoo Ganesh <[email protected]> > wrote: > > > > > > I’d love to have this on the website. Thanks, Ed! > > > > > >> On May 6, 2024, at 19:27, Ed Seidl <[email protected]> wrote: > > >> > > >> Hi all, > > >> Given the recent confusion on this list concerning Parquet V1 vs V2, I > > was wondering if there was any interest in the community to create a > > feature matrix that users could consult to see which implementations > would > > work with features they consider important. My group is compiling our own > > version of this for internal purposes, but we would consider cleaning it > up > > and posting it somewhere (perhaps on the Parquet site?) if there was > > interest in this. Since the format evolves slowly, once a first pass is > > complete, maintaining the list hopefully won't be burdensome...as > > implementers add features, they could submit PRs for their column in the > > matrix. > > >> > > >> Thoughts? Or does such a thing exist already somewhere? > > >> > > >> Thanks, > > >> Ed > > >
