I like this hierarchy of naming:
Provider
Graph System Provider
Graph Database Provider
Graph Processor Provider
Graph Driver Provider
Graph Language Provider
and "provider" does seem like the right term. That hierarchy nicely
describes the sub-communities within TinkerPop.
On Thu, Oct 8, 2015 at 12:07 PM, Marko Rodriguez <[email protected]>
wrote:
> Hello,
>
> Please note this ticket:
> https://issues.apache.org/jira/browse/TINKERPOP3-876
>
> Finally, note this branch where I provide an update to all the docs
> replacing "vendor" with "provider."
>
> https://github.com/apache/incubator-tinkerpop/commit/34bcf7eaa9deb42ae13ef1f8cb21de79cc5aa180
>
> Doing this update made me see the following hierarchy:
>
> Provider
> Graph System Provider
> Graph Database Provider
> Graph Processor Provider
> Graph Driver Provider
> Graph Language Provider
>
> Note that GraphSystem[Database/Process] is the common parlance of the
> graph computing space.
>
> Here are some examples to show how "provider" feels better than
> "implementor." The primary reason is that "implementation" is a noun used
> throughout to describe the "provider's implementation" (not the
> "implementor's implementation").
>
> "A graph system provider…"
> "The provider should implement…"
> "If the provider's implementation requires.."
>
> Try and do that with just "implementor." Watch:
>
> "A graph system implementor…"
> "The implementor should implement…"
> "If the implementor's implementation requires.."
>
> Thanks,
> Marko.
>
> http://markorodriguez.com
>
> On Oct 5, 2015, at 2:28 PM, Marko Rodriguez <[email protected]> wrote:
>
> > Seems that the DISCUSS is leaning towards removing "vendor" terminology.
> >
> > I will now make this a VOTE. If the VOTE is not heavily sided towards
> one side, we can bring up with gremlin-users@.
> >
> > Thanks everyone,
> > Marko.
> >
> > http://markorodriguez.com
> >
> > On Oct 5, 2015, at 2:26 PM, Ran Magen <[email protected]> wrote:
> >
> >> +1 Imolementers
> >>
> >> On Mon, 5 Oct 2015 at 20:26 Bob Briody <[email protected]> wrote:
> >>
> >>> I've always liked the terms implementer and implementation. Plus, as
> >>> Stephen stated, this was how it used to roll anyway.
> >>>
> >>> Everyone else is just a "user", right? Point being, "users" is a good
> term
> >>> to group application developers, driver developers, and whoever else
> into a
> >>> set of personas that operate via the APIs and are basically just not
> >>> implementers.
> >>>
> >>> On Mon, Oct 5, 2015 at 12:10 PM, Jason Plurad <[email protected]>
> wrote:
> >>>
> >>>> +1
> >>>>
> >>>> Down with vendors, Up with implementers.
> >>>>
> >>>> On Mon, Oct 5, 2015 at 12:03 PM, Stephen Mallette <
> [email protected]>
> >>>> wrote:
> >>>>
> >>>>> If memory serves, we've only used "vendor" for as long as TP3 has
> been
> >>>>> around. I swear those who built on top of Blueprints were
> "Blueprints
> >>>>> Implementers". If they built Gremlin (e.g. gremlin-scala) then we
> had
> >>>>> "Gremlin Implementers". I'm not sure we went so far to say "Rexster
> >>>>> Implementers", but the idea is sound. Anywho, "Implementers" does
> the
> >>>>> trick for me.
> >>>>>
> >>>>> On Mon, Oct 5, 2015 at 11:52 AM, Andrew Fitzgerald <
> >>>>> [email protected]> wrote:
> >>>>>
> >>>>>> +1 for dropping "vendor"
> >>>>>> On Oct 5, 2015 11:39 AM, "David Robinson" <[email protected]>
> >>>> wrote:
> >>>>>>
> >>>>>>> We may agree, Marko, on the discussion around "vendor", but some
> >>>> things
> >>>>>>> just aren't worth it.
> >>>>>>> By strict definition, not opinion, it does define someone selling
> >>>>>>> something.
> >>>>>>>
> >>>>>>> If the term is offensive, let's pick a new term we try to en-grain
> >>> in
> >>>>> our
> >>>>>>> behavior and move on.
> >>>>>>>
> >>>>>>> We can focus on more important things - like helping our
> >>> vendors...I
> >>>>> mean
> >>>>>>> "TP Implementors" use this cool think called TinkerPop/Gremlin.
> >>>>>>>
> >>>>>>> Here are two suggestions to place on the vote list for terminology:
> >>>>>>>
> >>>>>>> a) Application Developers - those leveraging the Gremlin Language /
> >>>>> APIs
> >>>>>> on
> >>>>>>> top of an implementation - be that Titan, Orient, Neo4J, Flink,
> >>>> Spark,
> >>>>>>> whatever.
> >>>>>>>
> >>>>>>> a) TinkerPop Implementers (or Implementers for short) - those that
> >>>>>>> implement an underlying system, whether for sale or not, that
> >>> expose
> >>>>> the
> >>>>>>> Gremlin Language / API.
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>> On Mon, Oct 5, 2015 at 11:13 AM, Marko Rodriguez <
> >>>> [email protected]
> >>>>>>
> >>>>>>> wrote:
> >>>>>>>
> >>>>>>>> Hello Pp-otik-ner
> >>>>>>>>
> >>>>>>>> It seems that the mentors are adverse to the term "vendor" even
> >>>>> though
> >>>>>> as
> >>>>>>>> the mentees have explained Hadoop, Spark, Gremlin-Scala,
> >>>> gremlin-php,
> >>>>>>> etc.
> >>>>>>>> are all considered "vendors." That is, anything that implements
> >>> the
> >>>>>>>> TinkerPop3 API regardless of them being commercial or otherwise
> >>> is
> >>>> a
> >>>>>>> vendor.
> >>>>>>>>
> >>>>>>>> With that said (and known), we can continually go back and forth
> >>>> with
> >>>>>>> "No.
> >>>>>>>> 'vendor' means this." "Uh uh, it doesn't -- it means this." "That
> >>>>> makes
> >>>>>>> no
> >>>>>>>> sense cause to me it means this."
> >>>>>>>>
> >>>>>>>> If we are going to get TinkerPop out of the malaise of personal
> >>>>>> opinions
> >>>>>>>> and arguments about meaning in the English language (in
> >>>> zeitgeist), I
> >>>>>> say
> >>>>>>>> we bring this to a collective VOTE which includes the whole
> >>>> community
> >>>>>>> (i.e.
> >>>>>>>> gremlin-users@ as well). I would frame the vote as:
> >>>>>>>>
> >>>>>>>> "Should TinkerPop abstain from its use of the word
> >>> 'vendor'
> >>>>> (to
> >>>>>>>> categorize graph system and graph language implementers) because,
> >>>> to
> >>>>>> you,
> >>>>>>>> it strongly implies commercial interest?"
> >>>>>>>>
> >>>>>>>> With that vote tally, we can then do accordingly and from then
> >>> on,
> >>>> no
> >>>>>>>> individual's personal opinion about the meaning of "vendor" will
> >>> be
> >>>>>>>> considered a valid argument given that language is a socially
> >>>>>> constructed
> >>>>>>>> phenomena.
> >>>>>>>>
> >>>>>>>> Thoughts?,
> >>>>>>>> Marko.
> >>>>>>>>
> >>>>>>>> http://markorodriguez.com
> >>>>>>>>
> >>>>>>>>
> >>>>>>>
> >>>>>>
> >>>>>
> >>>>
> >>>>
> >>>>
> >>>> --
> >>>> Have a good one,
> >>>> Jason
> >>>>
> >>>
> >
>
>
