Regarding the two possible bugs in Gremlin.Net: For TINKERPOP-2043, my current assumption is that the fix for TINKERPOP-2026 also solves this issue as TINKERPOP-2043 is caused by a failing close of WebSocket connections which is now silently catched with the fix from TINKERPOP-2026. So, I don't think that this problem can still occur. I just want to wait for confirmation from the user before we close the issue.
TINKERPOP-2019 is a lot more complicated and I still have basically no idea what causes this bug. However, I don't think that this issue should stop the release which includes other important fixes and improvements as we don't know when and if we are able to get down to this one. Since we also plan to completely rework how the ConnectionPool works in Gremlin.Net with TINKERPOP-1774 and TINKERPOP-1775, I don't think that it's worth the effort to spend a lot of time on trying to reproducing the bug. So, yes, I'm in favour of moving forward with the release without these two issues. -----Ursprüngliche Nachricht----- Von: Stephen Mallette <[email protected]> Gesendet: Freitag, 5. Oktober 2018 13:14 An: [email protected] Betreff: Re: [DISCUSS] 3.2.10, 3.3.4, 3.4.0 Releases It seems like we're closing in on the final items: + TINKERPOP-2019/TINKERPOP-2043 - Possible bugs in .NET - Florian, it + looks like you are digging around in these issues trying to get to the bottom of things - do you feel that we can release without these? + TINKERPOP-1972 - Two failing tests in .NET around inject() - Florian + is issuing a PR soon + https://github.com/apache/tinkerpop/pull/922 - gremlin-js script submission - i'm looking into the final issue with this now. I'm going to give another pass through the JIRA issue list, but I think this looks pretty good. Maybe we could code freeze today if PR 922 can get settled. Any concerns? On Wed, Oct 3, 2018 at 11:04 AM Stephen Mallette <[email protected]> wrote: > Here's the latest update on 3.2.10/3.3.4 release issues: > > + TINKERPOP-2019/TINKERPOP-2043 - Possible bugs in .NET > + TINKERPOP-1972 - Two failing tests in .NET around inject() (I can't > + seem > to get to the bottom of this one) > + https://github.com/apache/tinkerpop/pull/922 - gremlin-js script > submission (critical imo - i would hold release over this) > + TINKERPOP-2055 which should have PR soon and fixes special case > + numbers > in GraphSON > > We'll see what other odds/ends pop up.... > > > > On Thu, Sep 27, 2018 at 4:07 PM Florian Hockmann > <[email protected]> > wrote: > >> Cool, I just tried it out and it works as expected. In case anyone on >> here also wants to test the template: >> >> It must first be installed once with: >> >> dotnet new -i Gremlin.Net.Template::3.4.0-rc2 (The '::3.4.0-rc2' >> part is only necessary here because dotnet won't install pre-release >> versions otherwise.) >> >> and then a new project can be created with: >> >> dotnet new gremlin -o MyFirstGremlinProject >> >> where -o with its argument is optional and specifies the name of the >> project to create. >> >> >> Am 27.09.2018 um 16:59 schrieb Robert Dale: >> > Oh, it's its own package. I'm all setup for it now. Thanks. >> > >> > Robert Dale >> > >> > >> > On Thu, Sep 27, 2018 at 10:53 AM Stephen Mallette >> > <[email protected] >> > >> > wrote: >> > >> >> yeah - that part works...it's the "Template" package that failed >> >> to >> upload. >> >> i figured it out though: my API key was limited to publishing just >> >> to Gremlin.Net and not the Gremlin.Net.Template. It's there now: >> >> >> >> https://www.nuget.org/packages/Gremlin.Net.Template/ >> >> >> >> On Thu, Sep 27, 2018 at 10:32 AM Robert Dale <[email protected]> >> wrote: >> >> >> >>> I got an email that states: >> >>> >> >>> The package Gremlin.Net 3.4.0-rc2 >> >>> <https://www.nuget.org/packages/Gremlin.Net/3.4.0-rc2> was >> >>> recently published on NuGet Gallery by tinkerpop. If this was not >> >>> intended, please contact support >> >>> <https://www.nuget.org/packages/Gremlin.Net/3.4.0-rc2/ReportMyPac >> >>> kage >> >. >> >>> The link works. Looks like it worked. >> >>> >> >>> Robert Dale >> >>> >> >>> >> >>> On Thu, Sep 27, 2018 at 10:21 AM Stephen Mallette < >> [email protected]> >> >>> wrote: >> >>> >> >>>> So, 3.4.0-rc2 for Gremlin.Net is published. The template >> >>>> attempted to publish but failed with: >> >>>> >> >>>> [exec] Pushing Gremlin.Net.Template.3.4.0-rc2.nupkg to ' >> >>>> https://www.nuget.org/api/v2/package'... >> >>>> [exec] PUT https://www.nuget.org/api/v2/package/ >> >>>> [exec] ServiceUnavailable >> https://www.nuget.org/api/v2/package/ >> >>>> 458ms >> >>>> [exec] PUT https://www.nuget.org/api/v2/package/ >> >>>> [exec] 403 (The specified API key is invalid, has expired, >> >>>> or >> does >> >>> not >> >>>> have permission to access the specified package.) >> >>>> [exec] Forbidden https://www.nuget.org/api/v2/package/ 702ms >> >>>> >> >>>> I'm not sure what's going on there....does that make any sense >> >>>> to >> you, >> >>>> Florian? >> >>>> >> >>>> >> >>>> >> >>>> On Thu, Sep 27, 2018 at 9:26 AM Florian Hockmann < >> >> [email protected] >> >>>> wrote: >> >>>> >> >>>>> It should at least. We have never tested it, but the docs say >> >>>>> that >> >> 'mvn >> >>>>> clean install -Dnuget' can be used to create the package. So, >> >>>>> the >> >> other >> >>>>> Maven commands should work the same way. If it doesn't directly >> >>>>> work >> >> or >> >>>>> if you run into any other problems, then I can also give it a try. >> >>>>> >> >>>>> >> >>>>> Am 27.09.2018 um 15:20 schrieb Stephen Mallette: >> >>>>>> uh....i don't remember if there is anything special i need to >> >>>>>> do to >> >>>> make >> >>>>>> that happen. does it just deploy with the standard deploy >> >>> instructions? >> >>>>>> On Thu, Sep 27, 2018 at 9:13 AM Florian Hockmann < >> >>>> [email protected] >> >>>>>> wrote: >> >>>>>> >> >>>>>>> Great, do you also plan to include the Gremlin.Net.Template >> >>>>>>> in >> >> this >> >>>>>>> prerelease? >> >>>>>>> >> >>>>>>> >> >>>>>>> Am 26.09.2018 um 13:00 schrieb Stephen Mallette: >> >>>>>>>> Just a quick note that I plan to do the .NET 3.4.0-rc2 >> >>>>>>>> release >> >>>>> tomorrow. >> >>>>>>>> Here's the updated todo list for the 3.2.10/3.3.4: >> >>>>>>>> >> >>>>>>>> + TINKERPOP-2025 - >> >>>>>>>> + https://github.com/apache/tinkerpop/pull/935 >> >>>> ready >> >>>>> to >> >>>>>>>> merge >> >>>>>>>> + TINKERPOP-2019/TINKERPOP-2043 - Possible bugs in .NET >> >>>>>>>> + TINKERPOP-1906 - Make status messages/attributes from the >> >> server >> >>>> more >> >>>>>>>> available in .NET (maybe already done on TINKERPOP-1913 to >> >>>>>>>> some >> >>>> degree) >> >>>>>>>> + TINKERPOP-1972 - Two failing tests in .NET (I can't seem >> >>>>>>>> + to get >> >>> to >> >>>>> the >> >>>>>>>> bottom of this one) >> >>>>>>>> + https://github.com/apache/tinkerpop/pull/920 - minor >> >>> refactoring, >> >>>>> just >> >>>>>>>> had some activity on it, so it looks like this one will get in: >> >>>>>>>> + https://github.com/apache/tinkerpop/pull/922 - gremlin-js >> >> script >> >>>>>>>> submission (critical imo - i would hold release over this) >> >>>>>>>> + https://github.com/apache/tinkerpop/pull/928 - this one >> >>>>>>>> + just >> >>> needs >> >>>>> to >> >>>>>>> be >> >>>>>>>> merged i think - it's past cooling down period (florian has >> >>>>>>>> been >> >> on >> >>>>>>> holiday) >> >>>>>>>> + https://github.com/apache/tinkerpop/pull/929 - this one is >> >>>>>>>> + all >> >>> set >> >>>>>>> afaik >> >>>>>>>> and just needs to be merged to tp32 (jorge is handling that >> >>>>>>>> one) >> >>>>>>>> + https://github.com/apache/tinkerpop/pull/939 - groovy >> >> dependency >> >>>>>>> cleanup >> >>>>>>>> - easy one >> >>>>>>>> >> >>>>>>>> Note that I removed PR 903 as it's slated for 3.4.0 at this >> >> point. >> >>>>>>>> >> >>>>>>>> >> >>>>>>>> On Fri, Sep 21, 2018 at 9:46 PM Stephen Mallette < >> >>>> [email protected] >> >>>>>>>> wrote: >> >>>>>>>> >> >>>>>>>>> we need to add TINKERPOP-2025 to this too: >> >>>>>>>>> >> >>>>>>>>> +TINKERPOP-2025 - which is related to the build - Kuppitz, >> >>>>>>>>> +i >> >> think >> >>>> you >> >>>>>>> had >> >>>>>>>>> said that you were looking at that one....don't think we >> >>>>>>>>> can >> >>> release >> >>>>>>>>> without that >> >>>>>>>>> + Open PRs <= 903 (maybe with the exception of 920) - of >> >> critical >> >>>> note >> >>>>>>> to >> >>>>>>>>> me is 922 - that's on the critical path to me >> >>>>>>>>> + TINKERPOP-2019/TINKERPOP-2043 - Possible bugs in .NET >> >>>>>>>>> + TINKERPOP-1906 - Make status messages/attributes from the >> >> server >> >>>>> more >> >>>>>>>>> available in .NET (maybe already done on TINKERPOP-1913 to >> >>>>>>>>> some >> >>>>> degree) >> >>>>>>>>> + TINKERPOP-1972 - Two failing tests in .NET (I can't seem >> >>>>>>>>> + to >> >> get >> >>> to >> >>>>> the >> >>>>>>>>> bottom of this one) >> >>>>>>>>> >> >>>>>>>>> I just closed out this: >> >>>>>>>>> >> >>>>>>>>> + TINKERPOP-2030 - Bug in Java driver around keep-alive >> >>>>>>>>> >> >>>>>>>>> >> >>>>>>>>> On Fri, Sep 21, 2018 at 7:47 AM Stephen Mallette < >> >>>>> [email protected]> >> >>>>>>>>> wrote: >> >>>>>>>>> >> >>>>>>>>>> So, it seems like we're good to push off 3.4.0 for a bit. >> >>>>>>>>>> Let's >> >>>> look >> >>>>> to >> >>>>>>>>>> do the 3.4.0-rc2 for .NET next week - any volunteers to >> >>>>>>>>>> handle >> >>>> that? >> >>>>>>>>>> As for 3.2.10 and 3.3.4 code freeze, perhaps we look to do >> >>>>>>>>>> that >> >>> in >> >>>>> two >> >>>>>>>>>> weeks October 5 which would give us a release around the >> >>>>>>>>>> week >> >> of >> >>>>>>> October >> >>>>>>>>>> 15. If that's agreeable then please raise any issue that >> >>>>>>>>>> are >> >>>> thought >> >>>>>>> to be >> >>>>>>>>>> important for release so that we can track them here. >> >>>>>>>>>> Here's a >> >>> few >> >>>> of >> >>>>>>>>>> concern: >> >>>>>>>>>> >> >>>>>>>>>> + Open PRs <= 903 (maybe with the exception of 920) - of >> >> critical >> >>>>> note >> >>>>>>> to >> >>>>>>>>>> me is 922 - that's on the critical path to me >> >>>>>>>>>> + TINKERPOP-2030 - Bug in Java driver around keep-alive >> >>>>>>>>>> + TINKERPOP-2019/TINKERPOP-2043 - Possible bugs in .NET >> >>>>>>>>>> + TINKERPOP-1906 - Make status messages/attributes from >> >>>>>>>>>> + the >> >>> server >> >>>>> more >> >>>>>>>>>> available in .NET (maybe already done on TINKERPOP-1913 to >> >>>>>>>>>> some >> >>>>> degree) >> >>>>>>>>>> + TINKERPOP-1972 - Two failing tests in .NET (I can't seem >> >>>>>>>>>> + to >> >> get >> >>>> to >> >>>>>>> the >> >>>>>>>>>> bottom of this one) >> >>>>>>>>>> >> >>>>>>>>>> >> >>>>>>>>>> >> >>>>>>>>>> >> >>>>>>>>>> On Tue, Sep 18, 2018 at 12:33 PM Stephen Mallette < >> >>>>>>> [email protected]> >> >>>>>>>>>> wrote: >> >>>>>>>>>> >> >>>>>>>>>>> Given the importance of TINKERPOP-1913 for the CosmosDB >> >>>> community, I >> >>>>>>>>>>> think we should consider pushing out a 3.4.0-rc2 for .NET. >> >>>>>>>>>>> >> >>>>>>>>>>> >> >>>>>>>>>>> >> >>>>>>>>>>> On Tue, Sep 18, 2018 at 11:10 AM Robert Dale < >> >> [email protected] >> >>>>>>> wrote: >> >>>>>>>>>>>> +1 >> >>>>>>>>>>>> >> >>>>>>>>>>>> Robert Dale >> >>>>>>>>>>>> >> >>>>>>>>>>>> >> >>>>>>>>>>>> On Tue, Sep 18, 2018 at 4:13 AM Jorge Bay Gondra < >> >>>>>>>>>>>> [email protected]> >> >>>>>>>>>>>> wrote: >> >>>>>>>>>>>> >> >>>>>>>>>>>>> I agree we should start looking at timelines for 3.2.10 >> >>>>>>>>>>>>> and >> >>>> 3.3.4 >> >>>>>>> and >> >>>>>>>>>>>> don't >> >>>>>>>>>>>>> bother yet with a 3.4 release. >> >>>>>>>>>>>>> >> >>>>>>>>>>>>> El mar., 18 sept. 2018 a las 1:51, Stephen Mallette (< >> >>>>>>>>>>>> [email protected] >> >>>>>>>>>>>>>> ) >> >>>>>>>>>>>>> escribió: >> >>>>>>>>>>>>> >> >>>>>>>>>>>>>> I've been floating around "end of summer" for a >> >>>>>>>>>>>>>> release >> >> time >> >>>>> frame >> >>>>>>>>>>>> for >> >>>>>>>>>>>>> some >> >>>>>>>>>>>>>> time now. Well, end of summer is basically here and I >> >>>>>>>>>>>>>> feel >> >>> like >> >>>>>>>>>>>> 3.4.0 >> >>>>>>>>>>>>> still >> >>>>>>>>>>>>>> needs work. I feel like we should give it a bit more >> >>>>>>>>>>>>>> time >> >> to >> >>>>>>>>>>>> develop and >> >>>>>>>>>>>>>> then give it some fresh consideration in the next >> >>>>>>>>>>>>>> couple of >> >>>>> months. >> >>>>>>>>>>>>>> That said, 3.2.10 and 3.3.4 have a lot of good bug >> >>>>>>>>>>>>>> fixes >> >> and >> >>>>> minor >> >>>>>>>>>>>>>> features. Perhaps those shouldn't be delayed any further. >> >>> Maybe >> >>>>> we >> >>>>>>>>>>>> could >> >>>>>>>>>>>>>> look to code freeze in next few weeks on those >> >>>>>>>>>>>>>> branches and >> >>>>>>>>>>>> release. Then >> >>>>>>>>>>>>>> we could do a smaller 3.2.11 and 3.3.5 when 3.4.0 >> >>>>>>>>>>>>>> feels >> >> more >> >>>>> ready. >> >>>>>>>>>>>>>> Anyway any concerns about heading down that direction? >> >>>>>>>>>>>>>> >> >>>>> >> >> >>
