On Fri, Sep 27, 2013 at 2:35 AM, Arun C Murthy wrote:
Looks like vitess impl currently supports only CRAM-MD5, I'm interested in
a few more - happy to contribute.
Also, I have my own RPC system, would it be of interest if I could
re-factor vitess/rpcwrap/auth to support other RPC systems?
I see decent amount of rpc specific code there. I wonder how much will be
left if you separated that part out. But I (we) will be open to hearing you
out. You can either start a discussion on firstname.lastname@example.org
, or send
a pull request.
Along same lines, would there be interest in taking that back in as a go
module itself rather than stay in vitess?
Vitess has many generic packages. Practically, everything herehttps://github.com/youtube/vitess/tree/master/go
except vt is generic (some
are deprecated). But they were built for vitess. So, they may not be
well-rounded enough to be used by the masses. What goes into go needs to be
crafted very carefully because you can't freely change the API once it's
Otherwise, yes. Anything that's useful, generic and robust can be submitted
On Sep 25, 2013, at 4:43 PM, Sugu Sougoumarane wrote:
Vitess uses SASL. You can see our implementation here:https://github.com/youtube/vitess/tree/master/go/rpcwrap/auth.
rpcwrap builds a handshake protocol using the auth package for rpc
Note that we use rpcplus, which is backward compatible with go's rpc, but
also supports streaming and passing client context info.
On Wednesday, September 25, 2013 1:15:42 PM UTC-7, Arun C Murthy wrote:
Any pointers on a golang SASL client? Anyone has seen/done this before?
PS: Appreciate a cc, I'm not sub'ed to the list. Tx.
You received this message because you are subscribed to the Google Groups "golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email to email@example.com.
For more options, visit https://groups.google.com/groups/opt_out.