GitHub user FlorianHockmann opened a pull request:
https://github.com/apache/tinkerpop/pull/953
TINKERPOP-1972 Fix inject step in Gremlin.Net
https://issues.apache.org/jira/browse/TINKERPOP-1972
We had two issues with the `inject` step in Gremlin.Net. The first one
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/947#discussion_r222751451
--- Diff:
gremlin-javascript/src/main/javascript/gremlin-javascript/test/unit/graphson-test.js
---
@@ -46,6 +46,30 @@ describe('GraphSONReader
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/928
@deejvince: Did you see [the integration
test](https://github.com/apache/tinkerpop/blob/master/gremlin-dotnet/test/Gremlin.Net.IntegrationTest/Driver/GremlinClientTests.cs#L233)
added
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/928
I also wanted to implement it like that as first, but unfortunately the
[`ClientWebSocket.Options`
property](https://docs.microsoft.com/dotnet/api
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/928
Yes, that should be possible. With this PR, you can provide a delegate to
configure the ClientWebSocketOptions which should also include custom headers.
---
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/925
No, there is no reason for not merging this. I'm just currently on
vacation. So, I can't do it myself right now.
But if you or anyone else wants to merge this, then feel free to do that.
---
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/915
I'm currently out of office but I just gave this a quick look and my review
comments seem to be mostly addressed.
So: VOTE +1
---
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/903
> looks like you are focused on getting this one merged
No, I don't really mind waiting for TINKERPOP-1775 and implementing both
together. I just had two motivations for merg
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/903
Thanks for the insight @spmallette. That means for this PR that we don't
have to hurry at the moment.
> Max in-flight should be a threshold that, when reached, more connecti
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/903
> I think implementing TINKERPOP-1774 with the current implementation of
the pool (dequeuing) comes at a cost of introducing complexity (in particular
AsyncAutoResetEvent).
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/903
Ok, I think I finally got your proposition. Took me way longer than it
should have to be honest ð
The changes would basically look like this if I got everything right
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/903
>>What I meant here is that the connection cannot be used again until
SendAsync completed. So, my suggestion for request pipelining is simply that
the connection is taken out of th
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/903
>>Why do you want to keep the connections in the pool when a read/write
operation is executed on them? They can't be used at that time anyway and I
think that the current design is
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/915
> Is there a good reason for not exposing the StatusCode on the
ResponseException as well? I'm not familiar enough with the protocol to say
whether that's an internal detail or someth
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/903
>I think the ultimate solution would be for the pool to balance request
between connections, and without taking them out of the pool each time a
request/response is issued.
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/903
Thanks for the clarification. I think I understand now how that would look
like. It is of course a nice improvement. However, I wonder what would be the
concrete advantages for the user
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/926
LGTM, VOTE +1
---
GitHub user FlorianHockmann opened a pull request:
https://github.com/apache/tinkerpop/pull/925
Make closing of connections robust TINKERPOP-2026
https://issues.apache.org/jira/browse/TINKERPOP-2026
This should solve the problem described in the ticket where apparently
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/903#discussion_r214426405
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/AsyncAutoResetEvent.cs ---
@@ -0,0 +1,103 @@
+#region License
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/903
> We should allow sending multiple requests without waiting for a response
and support unordered responses (based on the request id)
Wouldn't that mean that the server se
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/903#discussion_r214384104
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/AsyncAutoResetEvent.cs ---
@@ -0,0 +1,103 @@
+#region License
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/903#discussion_r214373377
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ConnectionPool.cs ---
@@ -33,91 +34,135 @@ internal class ConnectionPool : IDisposable
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/903#discussion_r214371239
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/AsyncAutoResetEvent.cs ---
@@ -0,0 +1,103 @@
+#region License
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/903#discussion_r214055785
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ConnectionPool.cs ---
@@ -33,91 +34,134 @@ internal class ConnectionPool : IDisposable
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/912
> They settings are @Deprecated not removed - for example:
Ah, thanks for the clarification. I indeed only looked at the changes to
the docs. Deprecating those setti
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/912
To be honest, I haven't taken a detailed look at the changes of this PR
yet, but it seems to also remove / rename a bunch of config properties like
`ssl.keyFile` for example. Doesn't
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/915
In that case, you can change it with:
```
git config --global user.name "John Doe"
```
and then reset your two commits so you can write a new commit message and
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/912
Since this is a breaking change, wouldn't it be better to target `master`
so that this can go into 3.4.0?
---
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/903#discussion_r210924567
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ConnectionPool.cs ---
@@ -33,91 +34,134 @@ internal class ConnectionPool : IDisposable
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/915#discussion_r210893971
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/IResultSet.cs ---
@@ -0,0 +1,43 @@
+#region License
+/*
+ * Licensed to the Apache
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/915#discussion_r210903830
--- Diff:
gremlin-dotnet/test/Gremlin.Net.IntegrationTest/Driver/GremlinClientTests.cs ---
@@ -51,7 +52,18 @@ public async Task
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/915#discussion_r210902761
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ResultSet.cs ---
@@ -0,0 +1,85 @@
+#region License
+/*
+ * Licensed to the Apache
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/915#discussion_r210902499
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ResultSet.cs ---
@@ -0,0 +1,85 @@
+#region License
+/*
+ * Licensed to the Apache
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/915#discussion_r210901000
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/Connection.cs ---
@@ -116,9 +118,9 @@ private async Task>
ReceiveAs
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/915#discussion_r210901961
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ResultSet.cs ---
@@ -0,0 +1,85 @@
+#region License
+/*
+ * Licensed to the Apache
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/915#discussion_r210901733
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ResultSet.cs ---
@@ -0,0 +1,85 @@
+#region License
+/*
+ * Licensed to the Apache
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/915#discussion_r210895827
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/IResultSet.cs ---
@@ -0,0 +1,43 @@
+#region License
+/*
+ * Licensed to the Apache
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/911
Thanks for implementing this so quickly, @spmallette
LGTM, VOTE +1
---
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/842
I think I found the cause of the problem: We're using `dynamic` in
Gremlin.Net which requires the `Microsoft.CSharp` package. We always got that
package because Newtonsoft.Json also had
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/842
I'm looking into this. At a first glance it seems to be based on the
combination of the upgraded Newtonsoft.Json version and the addition of the
.NET Standard 2.0 target.
---
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/842#discussion_r209610435
--- Diff: gremlin-dotnet/src/Gremlin.Net/Gremlin.Net.csproj ---
@@ -48,7 +48,7 @@ Please see the reference documentation at Apache
TinkerPop
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/842#discussion_r209602159
--- Diff: gremlin-dotnet/src/Gremlin.Net/Gremlin.Net.csproj ---
@@ -48,7 +48,7 @@ Please see the reference documentation at Apache
TinkerPop
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/842
Great, thanks for making those changes.
My VOTE +1 from nearly 3 months ago still holds with these recent changes.
---
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/842#discussion_r209342928
--- Diff:
gremlin-dotnet/test/Gremlin.Net.UnitTest/Structure/IO/GraphSON/GraphSONWriterTests.cs
---
@@ -346,6 +347,70 @@ public void
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/842#discussion_r209346542
--- Diff:
gremlin-dotnet/src/Gremlin.Net/Structure/IO/GraphSON/TimeSpanDeserializer.cs ---
@@ -0,0 +1,37 @@
+#region License
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/909
Great to now also have the computer tests for the GLVs :tada:
VOTE +1
---
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/842
@danielcweber the build is now failing because some of the added files miss
the required license header. Could you please add that?
---
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/904
At least they are working on it for a future release:
Ticket: https://issues.apache.org/jira/browse/SPARK-24601
PR: https://github.com/apache/spark/pull/21596
Although
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/904
Ah ok, I didn't see that. I guess I'm still not really used to the way
Maven works ð
But in that case: VOTE +1
---
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/904
I just searched whether the library would also be used somewhere else and
found it [in
`spark-gremlin`](https://github.com/apache/tinkerpop/blob/tp32/spark-gremlin/pom.xml#L211
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/903
I rebased and added a commit that should address your comments. The
reference docs now contain a _Configuration_ section for Gremlin.Net.
I was also able to run all tests
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/903#discussion_r207262352
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ConnectionPoolSettings.cs
---
@@ -0,0 +1,55 @@
+#region License
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/903#discussion_r207222032
--- Diff: docs/src/upgrade/release-3.4.x.asciidoc ---
@@ -177,6 +177,17 @@ when dealing with that event. To make this easier, the
event now
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/903#discussion_r207220819
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ConnectionPoolSettings.cs
---
@@ -0,0 +1,55 @@
+#region License
GitHub user FlorianHockmann opened a pull request:
https://github.com/apache/tinkerpop/pull/903
Gremlin.Net: Add ConnectionPool min and max sizes TINKERPOP-1774
https://issues.apache.org/jira/browse/TINKERPOP-1774
The Gremlin.Net `ConnectionPool` now has min and max sizes
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/882
Just wondered: Does the _Shortest Path_ recipe still make sense with the
new `shortestPath()` step? And should the section in the reference docs about
shortest paths and the recipe link
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/897
Shouldn't we also add the `ConnectedComponent` class to the GLVs so users
can easily do something like this in other languages:
```java
g.V().hasLabel('person
GitHub user FlorianHockmann opened a pull request:
https://github.com/apache/tinkerpop/pull/895
TINKERPOP-2012 Add .NET Standard 2.0 target to Gremlin.Net
https://issues.apache.org/jira/browse/TINKERPOP-2012
This simply adds .NET Standard 2.0 as an additional target
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/894
Thanks for handling the `.gitignore`. I also reviewed the code and it looks
good to me.
So: VOTE +1
---
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/894
Looks like I accidentally added that `Dockerfile`. I should have checked
more carefully what my commit added but shouldn't we prevent that from
happening again with an entry
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/885#discussion_r197855449
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ConnectionPool.cs ---
@@ -45,19 +44,28 @@ public ConnectionPool(ConnectionFactory
GitHub user FlorianHockmann opened a pull request:
https://github.com/apache/tinkerpop/pull/885
Add check for connection state before returning it TINKERPOP-1978
https://issues.apache.org/jira/browse/TINKERPOP-1978
This PR simply adds the requested check of whether
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/867
Rebased again to fix a merge conflict.
---
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/867#discussion_r195899631
--- Diff: .travis.yml ---
@@ -20,9 +20,13 @@ install:
before_install:
- wget -q
https://packages.microsoft.com/config/ubuntu/14.04
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/867
I just rebased and made the changes so that now the template is only packed
with:
`mvn clean install -Dnuget`.
---
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/867
> we require nuget (to run nuget pack) only as part of the deploy phase.
I understand that part and I don't have anything against it, especially
when we add something to the &
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/802
> Please make a ticket in JIRA as a reminder for that and assign it a fix
version of 3.4.0 so we don't forget.
Done: [TINKERPOP-1980](https://issues.apache.org/jira/bro
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/802
Just pushed two CTR commits to hopefully get the tags right:
[aac534e11](https://github.com/apache/tinkerpop/commit/aac534e11fa406bedc61dbc060d06740b898e170)
adds a check to only
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/802
> I believe that aspect of this works fine. Nothing deploys when the
version is a SNAPSHOT.
But do we want to add the `3.2` tag for a `3.2.10-rc1` image? That means
that us
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/802
> Separately, do I need to edit the repository short/long description
manually? can that not be automatically done on deploy using content from our
github repo?
I'm afr
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/802
> I only see the 3.2 tag - where the 3.2.10-rc1? shouldn't that be up there
too?
Yes and the `3.2` tag should only exist for non-SNAPSHOT versions, but I
guess that we did
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/802
> I also just pushed a minor update to make the command more like our
others with -DdockerImages rather than explicitly using the -P directly which
we typically dont' do. Also disab
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/802
@spmallette you now voted +1 on this but didn't you want to push release
candidates of the Docker images during the review?
---
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/802#discussion_r191811843
--- Diff: docs/src/dev/developer/release.asciidoc ---
@@ -111,7 +111,8 @@ under release is protected. Tweaks to documentation and
other odds
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/842
In case someone is following here but is not on the dev mailing list:
https://lists.apache.org/thread.html/32c47e8682c2848a9cc90eaff0a0da1f9bf4bd5a94b746a3e143f341
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/842
> @FlorianHockmann Why would we do that? If we can construct (i.e.
deserialize) e.g. a DateTime of whatever a server gives us, fine. Doesn't mean
we have to send (i.e. serialize) obsc
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/842
After thinking a bit more about this, my conclusion was that deprecating
those types doesn't really make sense unless we plan on GraphSON 4.0 as we
couldn't remove the types before
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/868
@ashwini-ms did you see my comment regarding whether `IGremlinClient`
should return a `ResultSet` now instead of an `IReadOnlyCollection`? It's now a
bit hidden as GitHub collapsed
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/802
No, you're right. That is something that is very unlikely to occur and I
can't think of other (more likely) reasons where the build of the docker images
could fail as they really only
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/802
> what is the advantage to building these images if a .docker file is
present?
A reason to keep this option is that it makes it easier to detect when
changes break the Doc
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/802#discussion_r190320705
--- Diff: docs/src/reference/gremlin-applications.asciidoc ---
@@ -353,6 +353,38 @@ variable initialization code into the console.
Like
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/802#discussion_r190319039
--- Diff:
gremlin-driver/src/test/java/org/apache/tinkerpop/gremlin/driver/ResultQueueTest.java
---
@@ -281,7 +281,7 @@ public void
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/842
I'm ok with either direction - either to only support types symmetrically
or try to support at the reading site as much types as possible - as long as we
make that decision clear
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/867
> so if the version changed from 4..4.1
Good point, I haven't considered that. I amended my last commit to include
the solution you proposed. The downloaded exe now sim
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/867
You're right. It shouldn't be necessary to have an internet connection when
all dependencies are already available. I thought that nuget.exe would only be
downloaded once as we have
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/802
Rebased and removed the hard-coded version. The major and minor version are
now parsed with the `build-helper-maven-plugin` and added as tags to the image
when it's not a `SNAPSHOT
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/867
> Or, would it be crazy to only package templates if mono is present?
I think that would be the best solution here. It wouldn't introduce any new
dependencies to the bu
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/802#discussion_r189340476
--- Diff: gremlin-console/pom.xml ---
@@ -336,5 +336,59 @@ limitations under the License
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/868#discussion_r189275849
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ResultSet.cs ---
@@ -0,0 +1,88 @@
+#region License
+
+/*
+ * Licensed
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/867
Rebased and I extended the `gremlin-archetypes` section in the reference
docs by splitting it into two sections: _Maven Archetypes_ and _Gremlin.Net
Template_. In the future, we can add
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/842
Great and thanks for your contribution @danielcweber!
VOTE +1
---
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/842#discussion_r188710881
--- Diff:
gremlin-dotnet/test/Gremlin.Net.UnitTest/Structure/IO/GraphSON/GraphSONReaderTests.cs
---
@@ -39,7 +39,13 @@ public class
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/842#discussion_r188686243
--- Diff:
gremlin-dotnet/test/Gremlin.Net.UnitTest/Structure/IO/GraphSON/GraphSONReaderTests.cs
---
@@ -39,7 +39,13 @@ public class
Github user FlorianHockmann commented on the issue:
https://github.com/apache/tinkerpop/pull/842
Yeah, but they all extend `NumberConverter`, don't they? I remember that
there was a decision for gremlin-python a while back to separate serializers
and deserializers in different
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/868#discussion_r188655675
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ResultSet.cs ---
@@ -0,0 +1,88 @@
+#region License
+
+/*
+ * Licensed
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/868#discussion_r188657021
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ResultSet.cs ---
@@ -0,0 +1,88 @@
+#region License
+
+/*
+ * Licensed
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/868#discussion_r188657628
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ResultSet.cs ---
@@ -0,0 +1,88 @@
+#region License
+
+/*
+ * Licensed
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/868#discussion_r188658381
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ResultSet.cs ---
@@ -0,0 +1,88 @@
+#region License
+
+/*
+ * Licensed
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/868#discussion_r188654244
--- Diff: gremlin-dotnet/src/Gremlin.Net/Driver/ResultSet.cs ---
@@ -0,0 +1,88 @@
+#region License
+
+/*
+ * Licensed
Github user FlorianHockmann commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/868#discussion_r188661041
--- Diff:
gremlin-dotnet/test/Gremlin.Net.IntegrationTest/Driver/GremlinClientTests.cs ---
@@ -167,6 +168,25 @@ public class GremlinClientTests
1 - 100 of 207 matches
Mail list logo