You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tinkerpop.apache.org by Stephen Mallette <sp...@gmail.com> on 2021/02/19 12:16:55 UTC
[DISCUSS] Revert to Groovy 2.x for 3.5.0
Given the performance problems with Groovy 3.0 that I've detailed in this
thread:
https://lists.apache.org/thread.html/rbd7b09e53f7a0e421057192ed564907755252b8bd43085ecb16f98e4%40%3Cdev.tinkerpop.apache.org%3E
and further documented in this issue:
https://issues.apache.org/jira/browse/TINKERPOP-2526
and compounded by what I can only describe as unprecedented disinterest in
the problem by the Groovy Community, I think it wise that we revert 3.0 and
go back to 2.x. Thankfully, I think it was just this commit:
https://github.com/apache/tinkerpop/commit/cc3c5cb83e253b9949076628a7cfaade7f86f40e
I will then reopen the Groovy 3.0 issue:
https://issues.apache.org/jira/browse/TINKERPOP-2373
and link it to the performance blocker with TINKERPOP-2526.
If there are no objections in the next 72 hours, I'll assume lazy consensus
and push on in this direction.
I'm a bit shocked by the lack of feedback I've gotten on this issue from
Groovy to be honest. Just silence. It's understandable in that it is
perhaps a difficult/complex problem to get into but it's been months now
and without at least some direction and communication with Groovy project
members I could spend weeks developing a fix that may not even be
acceptable to be integrated into their code base.
I don't want to read too much into this situation but it highlights our
continued dependence on Groovy despite our attempts to get better
separation as we went to TinkerPop 3. I'm starting to feel concerned that
this dependence is shifting further to "risk". I don't think we need to
make any immediate decisions, but with the expected move to processing
Gremlin with antlr (i.e. gremlin-script)[1] it might be time to try to cut
the cord further. Perhaps we switch to JShell for the Gremlin Console or
develop our own repl around gremlin-script?? Maybe we also deprecate Groovy
as a ScriptEngine in Gremlin Server and only host gremlin-script?? I
suppose that all of this is ideas for a different thread and for a
different day.
[1]
https://lists.apache.org/thread.html/rc9288877898d583b3307eebf09949d5887081da38186dd82e59077ae%40%3Cdev.tinkerpop.apache.org%3E
Re: [DISCUSS] Revert to Groovy 2.x for 3.5.0
Posted by Stephen Mallette <sp...@gmail.com>.
Docs are publishing now for 3.5.0 after the revert:
https://tinkerpop.apache.org/docs/3.5.0-SNAPSHOT/reference/
On Fri, Feb 26, 2021 at 8:17 PM Stephen Mallette <sp...@gmail.com>
wrote:
> I've reverted to Groovy 2.5.x and the 3.0 issue is reopened:
>
> https://issues.apache.org/jira/browse/TINKERPOP-2373
>
> We're back to being fast on master. Going to generate docs for the first
> time in months. Hopefully it won't be too big a mess. I will post back here
> when I've got a new 3.5.0-SNAPSHOT published.
>
> On Fri, Feb 19, 2021 at 7:06 PM Josh Perryman <jo...@gmail.com>
> wrote:
>
>> As an interested observer, I think your concerns regarding Groovy are
>> valid. All of your considerations look entirely reasonable. Perhaps move
>> it
>> to "risk" as suggested, and note it in the next board report. Maybe the
>> Apache board has some pull in this situation?
>>
>> Reducing dependency on Groovy doesn't add a whole lot of value, but not
>> being able to build the documentation is a definite problem.
>>
>> It's a shame. I rather like Groovy, though my actual professional use of
>> it
>> has been limited.
>>
>> -Josh
>>
>> On Fri, Feb 19, 2021 at 6:17 AM Stephen Mallette <sp...@gmail.com>
>> wrote:
>>
>> > Given the performance problems with Groovy 3.0 that I've detailed in
>> this
>> > thread:
>> >
>> >
>> >
>> https://lists.apache.org/thread.html/rbd7b09e53f7a0e421057192ed564907755252b8bd43085ecb16f98e4%40%3Cdev.tinkerpop.apache.org%3E
>> >
>> > and further documented in this issue:
>> >
>> > https://issues.apache.org/jira/browse/TINKERPOP-2526
>> >
>> > and compounded by what I can only describe as unprecedented disinterest
>> in
>> > the problem by the Groovy Community, I think it wise that we revert 3.0
>> and
>> > go back to 2.x. Thankfully, I think it was just this commit:
>> >
>> >
>> >
>> https://github.com/apache/tinkerpop/commit/cc3c5cb83e253b9949076628a7cfaade7f86f40e
>> >
>> > I will then reopen the Groovy 3.0 issue:
>> >
>> > https://issues.apache.org/jira/browse/TINKERPOP-2373
>> >
>> > and link it to the performance blocker with TINKERPOP-2526.
>> >
>> > If there are no objections in the next 72 hours, I'll assume lazy
>> consensus
>> > and push on in this direction.
>> >
>> > I'm a bit shocked by the lack of feedback I've gotten on this issue from
>> > Groovy to be honest. Just silence. It's understandable in that it is
>> > perhaps a difficult/complex problem to get into but it's been months now
>> > and without at least some direction and communication with Groovy
>> project
>> > members I could spend weeks developing a fix that may not even be
>> > acceptable to be integrated into their code base.
>> >
>> > I don't want to read too much into this situation but it highlights our
>> > continued dependence on Groovy despite our attempts to get better
>> > separation as we went to TinkerPop 3. I'm starting to feel concerned
>> that
>> > this dependence is shifting further to "risk". I don't think we need to
>> > make any immediate decisions, but with the expected move to processing
>> > Gremlin with antlr (i.e. gremlin-script)[1] it might be time to try to
>> cut
>> > the cord further. Perhaps we switch to JShell for the Gremlin Console or
>> > develop our own repl around gremlin-script?? Maybe we also deprecate
>> Groovy
>> > as a ScriptEngine in Gremlin Server and only host gremlin-script?? I
>> > suppose that all of this is ideas for a different thread and for a
>> > different day.
>> >
>> > [1]
>> >
>> >
>> https://lists.apache.org/thread.html/rc9288877898d583b3307eebf09949d5887081da38186dd82e59077ae%40%3Cdev.tinkerpop.apache.org%3E
>> >
>>
>
Re: [DISCUSS] Revert to Groovy 2.x for 3.5.0
Posted by Stephen Mallette <sp...@gmail.com>.
I've reverted to Groovy 2.5.x and the 3.0 issue is reopened:
https://issues.apache.org/jira/browse/TINKERPOP-2373
We're back to being fast on master. Going to generate docs for the first
time in months. Hopefully it won't be too big a mess. I will post back here
when I've got a new 3.5.0-SNAPSHOT published.
On Fri, Feb 19, 2021 at 7:06 PM Josh Perryman <jo...@gmail.com>
wrote:
> As an interested observer, I think your concerns regarding Groovy are
> valid. All of your considerations look entirely reasonable. Perhaps move it
> to "risk" as suggested, and note it in the next board report. Maybe the
> Apache board has some pull in this situation?
>
> Reducing dependency on Groovy doesn't add a whole lot of value, but not
> being able to build the documentation is a definite problem.
>
> It's a shame. I rather like Groovy, though my actual professional use of it
> has been limited.
>
> -Josh
>
> On Fri, Feb 19, 2021 at 6:17 AM Stephen Mallette <sp...@gmail.com>
> wrote:
>
> > Given the performance problems with Groovy 3.0 that I've detailed in this
> > thread:
> >
> >
> >
> https://lists.apache.org/thread.html/rbd7b09e53f7a0e421057192ed564907755252b8bd43085ecb16f98e4%40%3Cdev.tinkerpop.apache.org%3E
> >
> > and further documented in this issue:
> >
> > https://issues.apache.org/jira/browse/TINKERPOP-2526
> >
> > and compounded by what I can only describe as unprecedented disinterest
> in
> > the problem by the Groovy Community, I think it wise that we revert 3.0
> and
> > go back to 2.x. Thankfully, I think it was just this commit:
> >
> >
> >
> https://github.com/apache/tinkerpop/commit/cc3c5cb83e253b9949076628a7cfaade7f86f40e
> >
> > I will then reopen the Groovy 3.0 issue:
> >
> > https://issues.apache.org/jira/browse/TINKERPOP-2373
> >
> > and link it to the performance blocker with TINKERPOP-2526.
> >
> > If there are no objections in the next 72 hours, I'll assume lazy
> consensus
> > and push on in this direction.
> >
> > I'm a bit shocked by the lack of feedback I've gotten on this issue from
> > Groovy to be honest. Just silence. It's understandable in that it is
> > perhaps a difficult/complex problem to get into but it's been months now
> > and without at least some direction and communication with Groovy project
> > members I could spend weeks developing a fix that may not even be
> > acceptable to be integrated into their code base.
> >
> > I don't want to read too much into this situation but it highlights our
> > continued dependence on Groovy despite our attempts to get better
> > separation as we went to TinkerPop 3. I'm starting to feel concerned that
> > this dependence is shifting further to "risk". I don't think we need to
> > make any immediate decisions, but with the expected move to processing
> > Gremlin with antlr (i.e. gremlin-script)[1] it might be time to try to
> cut
> > the cord further. Perhaps we switch to JShell for the Gremlin Console or
> > develop our own repl around gremlin-script?? Maybe we also deprecate
> Groovy
> > as a ScriptEngine in Gremlin Server and only host gremlin-script?? I
> > suppose that all of this is ideas for a different thread and for a
> > different day.
> >
> > [1]
> >
> >
> https://lists.apache.org/thread.html/rc9288877898d583b3307eebf09949d5887081da38186dd82e59077ae%40%3Cdev.tinkerpop.apache.org%3E
> >
>
Re: [DISCUSS] Revert to Groovy 2.x for 3.5.0
Posted by Josh Perryman <jo...@gmail.com>.
As an interested observer, I think your concerns regarding Groovy are
valid. All of your considerations look entirely reasonable. Perhaps move it
to "risk" as suggested, and note it in the next board report. Maybe the
Apache board has some pull in this situation?
Reducing dependency on Groovy doesn't add a whole lot of value, but not
being able to build the documentation is a definite problem.
It's a shame. I rather like Groovy, though my actual professional use of it
has been limited.
-Josh
On Fri, Feb 19, 2021 at 6:17 AM Stephen Mallette <sp...@gmail.com>
wrote:
> Given the performance problems with Groovy 3.0 that I've detailed in this
> thread:
>
>
> https://lists.apache.org/thread.html/rbd7b09e53f7a0e421057192ed564907755252b8bd43085ecb16f98e4%40%3Cdev.tinkerpop.apache.org%3E
>
> and further documented in this issue:
>
> https://issues.apache.org/jira/browse/TINKERPOP-2526
>
> and compounded by what I can only describe as unprecedented disinterest in
> the problem by the Groovy Community, I think it wise that we revert 3.0 and
> go back to 2.x. Thankfully, I think it was just this commit:
>
>
> https://github.com/apache/tinkerpop/commit/cc3c5cb83e253b9949076628a7cfaade7f86f40e
>
> I will then reopen the Groovy 3.0 issue:
>
> https://issues.apache.org/jira/browse/TINKERPOP-2373
>
> and link it to the performance blocker with TINKERPOP-2526.
>
> If there are no objections in the next 72 hours, I'll assume lazy consensus
> and push on in this direction.
>
> I'm a bit shocked by the lack of feedback I've gotten on this issue from
> Groovy to be honest. Just silence. It's understandable in that it is
> perhaps a difficult/complex problem to get into but it's been months now
> and without at least some direction and communication with Groovy project
> members I could spend weeks developing a fix that may not even be
> acceptable to be integrated into their code base.
>
> I don't want to read too much into this situation but it highlights our
> continued dependence on Groovy despite our attempts to get better
> separation as we went to TinkerPop 3. I'm starting to feel concerned that
> this dependence is shifting further to "risk". I don't think we need to
> make any immediate decisions, but with the expected move to processing
> Gremlin with antlr (i.e. gremlin-script)[1] it might be time to try to cut
> the cord further. Perhaps we switch to JShell for the Gremlin Console or
> develop our own repl around gremlin-script?? Maybe we also deprecate Groovy
> as a ScriptEngine in Gremlin Server and only host gremlin-script?? I
> suppose that all of this is ideas for a different thread and for a
> different day.
>
> [1]
>
> https://lists.apache.org/thread.html/rc9288877898d583b3307eebf09949d5887081da38186dd82e59077ae%40%3Cdev.tinkerpop.apache.org%3E
>