Let’s not allow the monthly release cadence to get in the way of high-impact conversations, like re-evaluating the status of generic tables. Remember how we all knew the community needed 1.0, but we kept piling on blockers? A release only matters if it actually delivers value to users.
Absolutely keep tagging anything that’s truly a 1.1.0 blocker, for example, things that would break the build or create serious regressions. But let’s treat the milestone label as a best-effort goal, not a gate. Yufei On Tue, Jul 29, 2025 at 8:31 AM Dmitri Bourlatchkov <di...@apache.org> wrote: > From my POV the "1.1.0" milestone works just as well and causes less > agitation than a "blocker" label :) > > We can use it for collecting work items for a release. Whether something is > a blocker probably needs to be discussed on the dev ML anyway. > > Cheers, > Dmitri. > > On Tue, Jul 29, 2025 at 9:13 AM Eric Maynard <eric.w.mayn...@gmail.com> > wrote: > > > As far as this thread is concerned, what does everyone think about just > > using a 1.1.0-blocker label similar to the one we had for 1.0 and > treating > > that as the source of truth? > > > > If we take that approach, we can hash out the status of some change to > > generic tables as a release blocker on the relevant issue or mailing list > > thread. > > > > —EM > > > > On Tue, Jul 29, 2025 at 21:55 Jean-Baptiste Onofré <j...@nanthrax.net> > > wrote: > > > > > By the way, specifically about Generic Tables I would still keep it as > > > experimental for the following reasons: > > > 1. We would need more feedback from the community to both evaluate > > > it's an interesting feature (from community standpoint) and it works > > > as the community expects > > > 2. Generic table has been included in 1.0.0, I think would consider at > > > least a couple of features (due to the first point) before considering > > > as non experimental feature. So probably 1.2.0 (e.g. September > > > release) would be a good time to evaluate Generic Table (also with the > > > help of the community). > > > 3. Very selfish :), I have a new proposal that I plan to submit soon > > > that could be a good "complement" of Generic Table. > > > > > > Just my $0.01 :) > > > > > > Regards > > > JB > > > > > > On Tue, Jul 29, 2025 at 1:00 PM Jean-Baptiste Onofré <j...@nanthrax.net> > > > wrote: > > > > > > > > Yeah you are right. It’s probably feature per feature. A discussion > on > > > dev can happen to remove the experiment flag. My point is that it’s not > > > related to a release: as soon as the discussion happens and > experimental > > > flag is removed it will be included in the next release (monthly). > > > > > > > > Wdyt ? > > > > > > > > Regards > > > > JB > > > > > > > > Le mar. 29 juil. 2025 à 11:40, Eric Maynard < > eric.w.mayn...@gmail.com> > > > a écrit : > > > >> > > > >> I agree with everything you said JB, but I’m not that it addresses > the > > > >> question around generic tables and the experimental label. In the > case > > > of > > > >> generic tables the feature is already in 1.0. It will be in 1.1.0. > > > >> > > > >> However we are labeling it “experimental” and the question here is > > about > > > >> under what conditions that label will be removed. Is it some number > of > > > >> releases? A vote? > > > >> > > > >> We should clarify this process, as right now it seems rather > > arbitrary. > > > It > > > >> appears that there is at least some tenuous connection to releases > in > > > the > > > >> minds of some community members, as this labeling of the feature as > > > >> experimental was reported to be a 1.0 blocker. > > > >> > > > >> —EM > > > >> > > > >> On Tue, Jul 29, 2025 at 13:37 Jean-Baptiste Onofré <j...@nanthrax.net > > > > > wrote: > > > >> > > > >> > Hi > > > >> > > > > >> > As we plan a monthly release pace now, I think we should use “best > > > effort” > > > >> > about what’s included in a release. If it’s not in this month > > release > > > it > > > >> > can included in next month one. > > > >> > It means that I plan 1.2.0 in September, 1.3.0 in October, etc. > IMHO > > > we > > > >> > should remove the target release number from roadmap gh discussion > > and > > > >> > instead list what we want by priority: as soon as it’s ready we > ship > > > it in > > > >> > the month release. > > > >> > > > > >> > With this approach, and as we will have a release every month, I’m > > > not sure > > > >> > a dedicated meeting will help much. Instead I propose we update > the > > > >> > issues/PRs with target milestone, and if not done at release date, > > we > > > >> > postpone to next month release. > > > >> > > > > >> > The purpose is to not being too ambitious in terms of what’s > > included > > > but > > > >> > have more frequent and predictable release dates for our users. > > > >> > > > > >> > Thoughts ? > > > >> > > > > >> > Regards > > > >> > JB > > > >> > > > > >> > Le mar. 29 juil. 2025 à 00:14, Dmitri Bourlatchkov < > > di...@apache.org> > > > a > > > >> > écrit : > > > >> > > > > >> > > Hi Yufei, > > > >> > > > > > >> > > > Can you elaborate on it? > > > >> > > > > > >> > > I proposed discussing (and elaborating on arguments about) the > > > Generic > > > >> > > Tables API in a separate thread. > > > >> > > > > > >> > > Cheers, > > > >> > > Dmitri. > > > >> > > > > > >> > > On Mon, Jul 28, 2025 at 5:06 PM Yufei Gu <flyrain...@gmail.com> > > > wrote: > > > >> > > > > > >> > > > Hi Dmitri, > > > >> > > > > > > >> > > > Thanks for chiming in. Are there any volunteers to work on > Helm > > > chart > > > >> > > > separation and CLI client publishing? > > > >> > > > > > > >> > > > > The related use cases and the future of them are still not > > clear > > > >> > > > > > > >> > > > Can you elaborate on it? The Delta table use case is pretty > > clear > > > to me > > > >> > > > that Polaris can host Delta tables and they are accessible > from > > > Spark. > > > >> > > > > > > >> > > > Yufei > > > >> > > > > > > >> > > > > > > >> > > > On Mon, Jul 28, 2025 at 1:56 PM Dmitri Bourlatchkov < > > > di...@apache.org> > > > >> > > > wrote: > > > >> > > > > > > >> > > > > Hi Yufei, > > > >> > > > > > > > >> > > > > Re: Generic Tables API - I do not think it is ready to be a > > > standard > > > >> > > > > feature in 1.1. The related use cases and the future of them > > are > > > >> > still > > > >> > > > not > > > >> > > > > clear, as far as I can tell. It may be worth having a > separate > > > >> > > discussion > > > >> > > > > thread for this. > > > >> > > > > > > > >> > > > > +1 to a separate helm repo (I believe this was discussed in > > > another > > > >> > > > thread > > > >> > > > > too). > > > >> > > > > > > > >> > > > > +1 to include CLI into the binary bundle. > > > >> > > > > > > > >> > > > > Cheers, > > > >> > > > > Dmitri. > > > >> > > > > > > > >> > > > > On Mon, Jul 28, 2025 at 3:12 PM Yufei Gu < > > flyrain...@gmail.com> > > > >> > wrote: > > > >> > > > > > > > >> > > > > > The timeline LGTM! > > > >> > > > > > > > > >> > > > > > We’ll need to align on the scope. In addition to several > > > ongoing > > > >> > > > features > > > >> > > > > > that should be ready by then, like MinIO support and KMS > > > support > > > >> > (PR > > > >> > > > > > #1424). There are a few key items(not a complete list) > that > > > need > > > >> > > > > > discussion: > > > >> > > > > > > > > >> > > > > > - Generic Table & Catalog Federation Status: We need to > > > decide > > > >> > > > whether > > > >> > > > > > the generic table feature and catalog federation should > > be > > > moved > > > >> > > out > > > >> > > > > of > > > >> > > > > > preview. Personally, I’d like to see the generic table > > > feature > > > >> > > > > graduate > > > >> > > > > > from preview. > > > >> > > > > > - Helm Chart Repository & Release Strategy: I propose > > > moving the > > > >> > > > Helm > > > >> > > > > > Chart to a separate repository and releasing it > > > independently. > > > >> > > > > Ideally, > > > >> > > > > > this should be in place by the 1.1.0 release. > > > >> > > > > > - CLI Client Release: I’d like to include the CLI > client > > > within > > > >> > > the > > > >> > > > > > binary bundle and get it published along with 1.1 > > release. > > > >> > > > > > > > > >> > > > > > I'd also propose a dedicated community sync for the 1.1 > > > release. > > > >> > > > > Thoughts? > > > >> > > > > > > > > >> > > > > > Yufei > > > >> > > > > > > > > >> > > > > > > > > >> > > > > > On Fri, Jul 18, 2025 at 8:22 AM Dmitri Bourlatchkov < > > > >> > > di...@apache.org> > > > >> > > > > > wrote: > > > >> > > > > > > > > >> > > > > > > A mid-August release sounds good to me. > > > >> > > > > > > > > > >> > > > > > > I've added MinIO-related issues to the milestone: > > > >> > > > > > > > > > >> > > > > > > * https://github.com/apache/polaris/issues/1901 > > > >> > > > > > > * https://github.com/apache/polaris/issues/1530 > > > >> > > > > > > > > > >> > > > > > > Cheers, > > > >> > > > > > > Dmitri. > > > >> > > > > > > > > > >> > > > > > > On Fri, Jul 18, 2025 at 1:07 AM Jean-Baptiste Onofré < > > > >> > > > j...@nanthrax.net> > > > >> > > > > > > wrote: > > > >> > > > > > > > > > >> > > > > > > > Hi folks, > > > >> > > > > > > > > > > >> > > > > > > > I propose to have 1.1.0-incubating release around > August > > > 20. > > > >> > > > > > > > > > > >> > > > > > > > On github: > > > >> > > > > > > > 1. I updated 1.1.0 milestone due date > > > >> > > > > > > > 2. I will move the open issues still on 1.0.0 > milestone > > to > > > >> > 1.1.0 > > > >> > > > > > > > 3. I will close the 1.0.0 milestone (as it has been > > > released) > > > >> > > > > > > > > > > >> > > > > > > > My main focus is to review/update/improve the release > > > guide and > > > >> > > > move > > > >> > > > > > > > forward on semi-automatic release. > > > >> > > > > > > > > > > >> > > > > > > > Thoughts ? > > > >> > > > > > > > > > > >> > > > > > > > If you agree, feel free to create/assign issues for > the > > > 1.1.0 > > > >> > > > > > milestone. > > > >> > > > > > > > > > > >> > > > > > > > Thanks ! > > > >> > > > > > > > Regards > > > >> > > > > > > > JB > > > >> > > > > > > > > > > >> > > > > > > > > > >> > > > > > > > > >> > > > > > > > >> > > > > > > >> > > > > > >> > > > > > > >
