Got it. On Mon, Jun 29, 2015 at 4:15 PM, Stephen Mallette <spmalle...@gmail.com> wrote:
> yeah - you can do something like that with websockets. i think i need more > data than a simple string will hold though in that subprotocol header. > it's more than just the subprotocol name. i also have to send back > serializer configuration information which is more hierarchical if i want > to dynamically try to instantiate the right serializers on the driver side. > > On Mon, Jun 29, 2015 at 4:04 PM, David Brown <davebs...@gmail.com> wrote: > > > Ah, I must have misunderstood the protocol. I thought that you could do > > something like this: > > > > GET /chat HTTP/1.1 > > Host: server.example.com > > Upgrade: websocket > > Connection: Upgrade > > Sec-WebSocket-Key: dGhlIHNhbXBsZSBub25jZQ== > > Origin: http://example.com > > Sec-WebSocket-Protocol: gyro, json > > Sec-WebSocket-Version: 13 > > > > > > HTTP/1.1 101 Switching Protocols > > Upgrade: websocket > > Connection: Upgrade > > Sec-WebSocket-Accept: s3pPLMBiTxaQ9kYGzzhZRbK+xOo= > > Sec-WebSocket-Protocol: gyro > > > > > > But I guess this is used by more universal style subprotocols, > > registered with IANA etc. > > > > > > > > > > On Mon, Jun 29, 2015 at 2:54 PM, Stephen Mallette <spmalle...@gmail.com> > > wrote: > > > > > I hadn't thought that far ahead. I"m guessing it wouldn't however as I > > > don't think I'd want to hijack that aspect of the websocket protocol - > > > unless of course there was something built into that i'm unaware of > that > > > would allow for some sort of custom payload. Perhaps it's a new for of > > > "control" operation that the driver could just request when it > > > connects..??? Note that this feature would mostly be useful to those > > > drivers that are JVM based and use gryo. GraphSON serializers pipe > > > everything to generic Map/List constructs which make it easy for > non-JVM > > > languages to deserialize and thus don't require any special > > configuration. > > > > > > On Mon, Jun 29, 2015 at 2:32 PM, David Brown <davebs...@gmail.com> > > wrote: > > > > > > > Would this information be sent as part of the Sec-WebSocket-Protocol > in > > > the > > > > handshake header? > > > > > > > > On Mon, Jun 29, 2015 at 8:53 AM, stephen mallette (JIRA) < > > > j...@apache.org> > > > > wrote: > > > > > > > > > > > > > > [ > > > > > > > > > > > > > > > https://issues.apache.org/jira/browse/TINKERPOP3-740?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel > > > > > ] > > > > > > > > > > stephen mallette updated TINKERPOP3-740: > > > > > ---------------------------------------- > > > > > Priority: Major (was: Minor) > > > > > > > > > > It might help in the short term to allow the serializer to accept > the > > > > > {{IoRegistry}} in the yaml file on the driver side. > > > > > > > > > > > Serializer Handshake > > > > > > -------------------- > > > > > > > > > > > > Key: TINKERPOP3-740 > > > > > > URL: > > > > > https://issues.apache.org/jira/browse/TINKERPOP3-740 > > > > > > Project: TinkerPop 3 > > > > > > Issue Type: Improvement > > > > > > Components: driver, io, server > > > > > > Reporter: stephen mallette > > > > > > Assignee: stephen mallette > > > > > > > > > > > > Might be nice if Gremlin Server offered some way to tell the > driver > > > > what > > > > > serializers it had (sorta thinking specifically of gryo) so that it > > > could > > > > > try to use the same ones on the client. That way the user wouldn't > > > need > > > > to > > > > > hand configure them to be the same and if some class wasn't present > > the > > > > > driver could fail early with a clear error. > > > > > > > > > > > > > > > > > > > > -- > > > > > This message was sent by Atlassian JIRA > > > > > (v6.3.4#6332) > > > > > > > > > > > > > > > > > > > > > -- > > > > David M. Brown > > > > R.A. CulturePlex Lab, Western University > > > > > > > > > > > > > > > -- > > David M. Brown > > R.A. CulturePlex Lab, Western University > > > -- David M. Brown R.A. CulturePlex Lab, Western University
