There's a fair bit of work happening on 3.5-dev right now. It's probably best to not slow down anyone's momentum on that to hold 3.5-dev frozen. If we have to do a small 3.5.3 release alongside 3.6.0 to ensure functional parity I guess that's fine. I think we can reopen 3.5-dev again. Please let me know if there are any concerns.
On Wed, Jan 5, 2022 at 2:39 PM Stephen Mallette <[email protected]> wrote: > Hi folks, I've been away for a bit during the holiday but back now > thinking about this release. The issue for 3.4-dev that made me want to > hold release was resolved. The fix had nothing to do with the code itself > and more to do with the fact that memory requirements for the tests we'd > added over the years seemed to have quietly reached a point where they were > sometimes exceeding the memory available of the GH Actions instances > running them. After a bit of profiling Maven executions (something I've > never had to do ever) I was able to pretty drastically reduce the memory > consumption and now everything works smoothly. > > There is one PR out there that should be merged for 3.4-dev (it's been > open for a while pending the CI problem): > > https://github.com/apache/tinkerpop/pull/1534 > > and will try to get that done immediately. As I'm not aware of any other > work, for 3.5-dev and 3.4-dev I think we can proceed as if this were code > freeze week and move to release. I think we should hold on release 3.6.0 > and continue to land PRs on master during the release process and then keep > the freeze on 3.5-dev until 3.6.0 is out the door (recall that 3.4-dev is > basically retired now). Please let me know if there are concerns. > > On Wed, Dec 29, 2021 at 2:05 PM Stephen Mallette <[email protected]> > wrote: > >> I think we will need to take a bit of a delay on the code freeze for this >> Friday. 3.4-dev is unstable in CI - i've been trying to get to the root of >> the problem for a week now and it's still not completely resolved. That's >> prevented me from looking at other issues I wanted to get done and I think >> that there is definitely some outstanding work from folks who are off for >> the holidays. So, let's revisit code freeze a week from now and see where >> things are. Thanks! >> >> On Mon, Dec 13, 2021 at 3:37 PM Stephen Mallette <[email protected]> >> wrote: >> >>> In the worst case, if there were things we wanted added to 3.6.0 that >>> weren't quite ready for code freeze we could release 3.5.2 and hold the >>> freeze on its branch while 3.6.0 finished. I just wouldn't want a 3.6.0 to >>> force a few commits into a fast turnaround release of 3.5.3. That might >>> give some buffer if we wanted, but it would be nice if it could all just >>> land at once. >>> >>> On Mon, Dec 13, 2021 at 3:07 PM Kelvin Lawrence <[email protected]> >>> wrote: >>> >>>> The work for adding a regex TextP is sitting on the TINKERPOP-2652 >>>> branch. I still need to write the docs and add one last Java test but >>>> otherwise it's close to PR ready. I would like to try and get that into 3.6 >>>> which I think I should be able to do before the cutoff. >>>> In general capturing what we have now into a set of releases makes >>>> sense to start the year with a pretty clean mainline branch that we can >>>> iterate on. >>>> Kelvin >>>> >>>> On Monday, December 13, 2021, 06:44:46 AM CST, Stephen Mallette < >>>> [email protected]> wrote: >>>> >>>> I'd like to propose we do a release for the new year, with code freeze >>>> starting conveniently on Friday December 31. >>>> >>>> I think we should also consider making this the last release of 3.4.x. >>>> >>>> It occurred to me that 3.6.x doesn't have a name yet. Working in reverse >>>> (never done it this way before), I have a Victorian/Steampunk sorta >>>> Gremlin >>>> image that could be used for a logo, and searching around for related >>>> music >>>> came across a whole genre of Steampunk Music. First artist group I >>>> clicked >>>> on, Deus Ex Vapore Machina, has a song called "Tinkerheart" - sorta just >>>> fits I'd say. >>>> >>> >>>
