Hi, > On Jul 6, 2016, at 8:25 AM, Ted Wilmes <[email protected]> wrote: > > Sounds good and I will include benchmark numbers. I probably won't have > large scale Spark cluster #'s at the time of PR but maybe can get those out > during the freeze. I did have one question, should this work go into the > 3.1 line or just 3.2.x?
Okay. There are some integration tests on MatchStep that consume a lot of memory due to path-data. I think we can get a good idea of speed on Spark simply by how much faster the integration tests run! :) But, of course, large scale cluster testing would be awesome. Regarding 3.1.x or 3.2.x — it depends. I suspect the changes are pretty complex and thus, it would be best to put into 3.2.x only, but if not, then yes, target 3.1.x and up merge to 3.2.x. Up to you. Thanks, Marko. > > Thanks, > Ted > > On Wed, Jul 6, 2016 at 9:15 AM, Marko Rodriguez <[email protected]> > wrote: > >> @Ted: When you get that PR out, I will immediately focus on reviewing it. >> Please make sure your branch has master/ merged into it so we know we know >> that if the code/design is good, then there won’t be any hiccups on merge. >> >> *** Also, benchmarks that dropping paths is “good” will be very important >> so hopefully you have that coming too. >> >> Thanks, >> Marko. >> >> http://markorodriguez.com >> >> >> >>> On Jul 6, 2016, at 8:07 AM, Stephen Mallette <[email protected]> >> wrote: >>> >>> We're starting to go through reviews now with some earnest. Unless there >>> are concerns, I think that we should target code freeze for Friday at >> this >>> point. >>> >>> On Wed, Jul 6, 2016 at 10:01 AM, Ted Wilmes <[email protected]> wrote: >>> >>>> Quick update, I plan on getting a PR in late tonight or tomorrow morning >>>> for 1254. >>>> >>>> Thanks, >>>> Ted >>>> >>>> On Fri, Jul 1, 2016 at 6:04 PM, Stephen Mallette <[email protected]> >>>> wrote: >>>> >>>>> Thanks for the update Ted. We may need an extra day or two to clear up >>>> some >>>>> other discussion on some of the other PRs - i don't think we can start >> a >>>>> code freeze on Monday as I'd proposed. >>>>> >>>>> On Fri, Jul 1, 2016 at 6:49 PM, Ted Wilmes <[email protected]> wrote: >>>>> >>>>>> Hi everyone, >>>>>> I had some time this week and made further progress on 1254. Still >>>>> working >>>>>> out some kinks but I'll see how far I can get over the weekend. Have >> a >>>>>> good 4th. >>>>>> >>>>>> --Ted >>>>>> >>>>>> On Tue, Jun 28, 2016 at 11:50 AM, Marko Rodriguez < >>>> [email protected]> >>>>>> wrote: >>>>>> >>>>>>> Hi, >>>>>>> >>>>>>> Ted — I think we should get that work into the next release. Thus, if >>>>> you >>>>>>> need more time (reasonable amount), then I say we delay accordingly. >>>>>>> >>>>>>> NOTE: The Gremlin-Python stuff will not get into the next release. It >>>>> has >>>>>>> gotten really complex/powerful and is currently 100+ commits ahead of >>>>>>> master/ ! :) .. As such, given the gargantuan undertaking, we will >>>> save >>>>>>> this work for a future release. >>>>>>> >>>>>>> Thanks, >>>>>>> Marko. >>>>>>> >>>>>>> http://markorodriguez.com >>>>>>> >>>>>>> >>>>>>> >>>>>>>> On Jun 28, 2016, at 10:36 AM, Ted Wilmes <[email protected]> >>>> wrote: >>>>>>>> >>>>>>>> I would like to get TinkerPop-1254 in before code freeze. Shooting >>>>> for >>>>>>>> finishing it up by end of this week but I'll drop a note if it >>>> looks >>>>>>> like I >>>>>>>> won't make it. >>>>>>>> >>>>>>>> --Ted >>>>>>>> >>>>>>>> On Tue, Jun 28, 2016 at 5:35 AM, Stephen Mallette < >>>>>> [email protected]> >>>>>>>> wrote: >>>>>>>> >>>>>>>>> It's end of June - time to start firming up for release. We have a >>>>>>> number >>>>>>>>> of outstanding PRs that need votes/merge. Perhaps we try to get >>>>> those >>>>>>> all >>>>>>>>> in this week and begin code freeze next Monday (7/4)? >>>>>>>>> >>>>>>>>> I think we'll have to postpone the GLV work with gremlin-python >>>> for >>>>>> this >>>>>>>>> release. There's just too much left to do to get that in "right". >>>>> Are >>>>>>> there >>>>>>>>> any other open issues of importance? >>>>>>>>> >>>>>>>>> On Mon, Jun 6, 2016 at 9:18 AM, Marko Rodriguez < >>>>> [email protected] >>>>>>> >>>>>>>>> wrote: >>>>>>>>> >>>>>>>>>> Yes, end of June-ish is best for me as I have few things on my >>>>> plate >>>>>>> the >>>>>>>>>> first half of this month. >>>>>>>>>> >>>>>>>>>> Thanks, >>>>>>>>>> Marko. >>>>>>>>>> >>>>>>>>>> http://markorodriguez.com >>>>>>>>>> >>>>>>>>>> >>>>>>>>>>> On Jun 6, 2016, at 4:46 AM, Stephen Mallette < >>>>> [email protected]> >>>>>>>>>> wrote: >>>>>>>>>>> >>>>>>>>>>> We didn't really discuss a date for release on this thread. I >>>> was >>>>>>>>>> thinking >>>>>>>>>>> that we could start looking at the week of July 4th as the >>>> target >>>>>> week >>>>>>>>>> for >>>>>>>>>>> VOTE and nail down a date as we get closer. >>>>>>>>>>> >>>>>>>>>>> On Wed, May 25, 2016 at 2:42 PM, Jason Plurad < >>>> [email protected]> >>>>>>>>> wrote: >>>>>>>>>>> >>>>>>>>>>>> I'd think from a TinkerPop branding perspective, it probably >>>>> helps >>>>>> to >>>>>>>>>> have >>>>>>>>>>>> the name in there. It's Apache TinkerPop, not Apache Gremlin. >>>>>>>>>>>> >>>>>>>>>>>> I just took a quick look on a mirror, and some other Apache >>>>>> projects >>>>>>>>>>>> (Spark, Kafka, HBase, NiFi, Pig, Zookeeper) don't even include >>>>>> apache >>>>>>>>> in >>>>>>>>>>>> their distributables, so maybe we can just do: >>>>>>>>>>>> >>>>>>>>>>>> tinkerpop-gremlin-console-x.y.z.zip >>>>>>>>>>>> tinkerpop-gremlin-server-x.y.z.zip >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> On Wed, May 25, 2016 at 11:03 AM Stephen Mallette < >>>>>>>>> [email protected] >>>>>>>>>>> >>>>>>>>>>>> wrote: >>>>>>>>>>>> >>>>>>>>>>>>> jason i think that was a suggestion to conform more to >>>> standard >>>>>>>>> apache >>>>>>>>>>>>> releases from someone in incubator. if it was mandatory we >>>> would >>>>>>> have >>>>>>>>>>>>> burned for that too many times to count at this point. i'm >>>> good >>>>> to >>>>>>>>>> change >>>>>>>>>>>>> it if everyone else is. what do we want them to be? >>>>>>>>>>>>> >>>>>>>>>>>>> apache-tinkerpop-console-x.y.z.zip >>>>>>>>>>>>> apache-tinkerpop-server-x.y.z.zip >>>>>>>>>>>>> >>>>>>>>>>>>> or the full business: >>>>>>>>>>>>> >>>>>>>>>>>>> apache-tinkerpop-gremlin-console-x.y.z.zip >>>>>>>>>>>>> apache-tinkerpop-gremlin-server-x.y.z.zip >>>>>>>>>>>>> >>>>>>>>>>>>> i guess we lost "-incubating" now so the latter doesn't look >>>> so >>>>>> bad >>>>>>>>> to >>>>>>>>>> me >>>>>>>>>>>>> anymore. >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> On Wed, May 25, 2016 at 10:56 AM, Marko Rodriguez < >>>>>>>>>> [email protected]> >>>>>>>>>>>>> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>> Hi, >>>>>>>>>>>>>> >>>>>>>>>>>>>> Yes, an imminent release is good. There are 2 severe bug >>>> fixes >>>>> in >>>>>>>>>>>> master/ >>>>>>>>>>>>>> (3.2.1) that I would like to get out there. 3.2.0 had lots of >>>>>>>>> internal >>>>>>>>>>>>>> changes to OLAP and I paid the price by incurring bugs. :| >>>>>>>>>>>>>> >>>>>>>>>>>>>>> Somebody had mentioned that our distributables are supposed >>>> to >>>>>> be >>>>>>>>>>>> named >>>>>>>>>>>>>>> apache-tinkerpop*.zip instead of apache-gremlin*.zip. Maybe >>>>>> that's >>>>>>>>>>>>>>> something that should be done along with this release. >>>>>>>>>>>>>> >>>>>>>>>>>>>> There is really no such thing as "tinkerpop" besides the >>>> source >>>>>>> code >>>>>>>>>>>>> which >>>>>>>>>>>>>> is distributed as apache-tinkerpop-*.zip. The two other >>>>>>>>> distributions >>>>>>>>>>>> are >>>>>>>>>>>>>> gremlin-console and gremlin-server and I think we should keep >>>>>> those >>>>>>>>>>>>> naming >>>>>>>>>>>>>> conventions as they are so they reflect what is being >>>>>> distributed. >>>>>>>>>>>> Thus, >>>>>>>>>>>>> I >>>>>>>>>>>>>> think the naming of our artifacts is correct. >>>>>>>>>>>>>> >>>>>>>>>>>>>> Thanks, >>>>>>>>>>>>>> Marko. >>>>>>>>>>>>>> >>>>>>>>>>>>>> http://markorodriguez.com >>>>>>>>>>>>>> >>>>>>>>>>>>>> On May 25, 2016, at 8:38 AM, Jason Plurad <[email protected] >>>>> >>>>>>>>> wrote: >>>>>>>>>>>>>> >>>>>>>>>>>>>>> Somebody had mentioned that our distributables are supposed >>>> to >>>>>> be >>>>>>>>>>>> named >>>>>>>>>>>>>>> apache-tinkerpop*.zip instead of apache-gremlin*.zip. Maybe >>>>>> that's >>>>>>>>>>>>>>> something that should be done along with this release. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> -- Jason >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> On Wed, May 25, 2016 at 9:55 AM, Stephen Mallette < >>>>>>>>>>>>> [email protected]> >>>>>>>>>>>>>>> wrote: >>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> cool, Ted. it would be good to have another hand there. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> On Wed, May 25, 2016 at 9:13 AM, Ted Wilmes < >>>>> [email protected] >>>>>>> >>>>>>>>>>>>> wrote: >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> I think a release sounds good. I'd be interested in >>>>>> witnessing >>>>>>>>> the >>>>>>>>>>>>> the >>>>>>>>>>>>>>>>> post-PMC vote release steps so that I might be able to >>>> help >>>>>> out >>>>>>>>> on >>>>>>>>>>>> an >>>>>>>>>>>>>>>>> upcoming release. >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> --Ted >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> On Wed, May 25, 2016 at 5:35 AM, Marvin Froeder < >>>>>>>>> [email protected] >>>>>>>>>>>>> >>>>>>>>>>>>>>>> wrote: >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>> Your are right, for some reason I though it was on the >>>>>>>>> artifactId >>>>>>>>>>>> as >>>>>>>>>>>>>>>> well >>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>> On Wed, May 25, 2016 at 10:22 PM, Stephen Mallette < >>>>>>>>>>>>>>>> [email protected] >>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>> wrote: >>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>> I don't think we need to relocate anything. The >>>>>> "-incubating" >>>>>>>>> is >>>>>>>>>>>>> just >>>>>>>>>>>>>>>>> in >>>>>>>>>>>>>>>>>>> the version name, so we will just remove it for future >>>>>>>>> releases. >>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>> On Wed, May 25, 2016 at 4:55 AM, Jean-Baptiste Musso < >>>>>>>>>>>>>>>>> [email protected]> >>>>>>>>>>>>>>>>>>> wrote: >>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>> I think this is a good idea. This could make these >>>>> releases >>>>>>>>> look >>>>>>>>>>>>>>>> more >>>>>>>>>>>>>>>>>>>> "stable": I've often felt that the -incubating suffix >>>>>> somehow >>>>>>>>>>>> made >>>>>>>>>>>>>>>>>>>> releases look "alpha-ish" / "beta-ish", even though >>>> they >>>>>> were >>>>>>>>>>>> not. >>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>> Naming aside, bug fixes never hurt. >>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>> Jean-Baptiste >>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>> On Wed, May 25, 2016 at 12:49 AM, Stephen Mallette < >>>>>>>>>>>>>>>>>> [email protected] >>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>> wrote: >>>>>>>>>>>>>>>>>>>>> We've seen a lot of good fixes/optimizations to 3.1.3 >>>>> and >>>>>>>>> 3.2.1 >>>>>>>>>>>>>>>>> and I >>>>>>>>>>>>>>>>>>>>> wonder if we shouldn't exercise our new found TLP >>>> powers >>>>>> to >>>>>>>>> do >>>>>>>>>>>> a >>>>>>>>>>>>>>>>>>> release >>>>>>>>>>>>>>>>>>>>> and get rid of the "-incubating" at the end of our >>>>>> "current" >>>>>>>>>>>>>>>>>>>> distributions >>>>>>>>>>>>>>>>>>>>> and artifacts. thoughts? >>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>> >>>>>>> >>>>>>> >>>>>> >>>>> >>>> >> >>
