Qtwebsocket will be provided by the next official Qt 5.3, I would prefer
wait under jqt uses that version before reviewing how to implement low
level functions.
On Feb 12, 2014 3:07 AM, "Joe Bogner" <joebog...@gmail.com> wrote:

> If QtWebSocket is being used then it should be accessible through
> http://qt-project.org/doc/qt-4.8/qabstractsocket.html#peerAddress of the
> socket that is returned from QtWebsocket::QWsSocket* clientSocket =
> server->nextPendingConnection();. I'm assuming it's implemented along the
> lines of
>
> https://gitorious.org/qtwebsocket/qtwebsocket/source/75f9105977502b5ec7ae405820b112c7c576b10a:Example/Server/Server.cpp#L44
>
>
> Perhaps pushing the peer address string to a global variable like wss0_jrx_
> would be enough
>
>
>
>
>
> On Tue, Feb 11, 2014 at 12:33 PM, Pascal Jasmin <godspiral2...@yahoo.ca
> >wrote:
>
> > I'd like to make an application that allows public connections, and in
> > fact is peer to peer so the world can run a node/the code.  Websockest
> > solves an issue of allowing more than 64 connections on windows, and even
> > 1024 connections on linux/mac/android.  Websockets also transparently
> > handles large block transfers.
> >
> > I've posted J code before on an authentication layer in J, and managing
> > multiple connections, and walling off code access to clients. (
> >
> http://www.jsoftware.com/jwiki/PascalJasmin/OOP%20scheduler%20and%20jsockets
> )
> >
> > The one requirement though is that there should be some limit on login
> > attempts per IP, and some way of detecting obvious abuse such as millions
> > of login attempts or sending terrabytes in response to a login prompt.
> >
> > "There is REMOTE_ADDR"
> >
> > no way to access from J yet?
> >
> > " put a websocket server behind a webserver or other proxy. What we
> > candone is unlikely to be better than those specialised products."
> >
> > A user code solution is straightforward if REMOTE_ADDR is available.  Its
> > also preferable if the application provides default code for dealing with
> > access limits and doesn't require any user decisions/settings, much less
> > installing firewalls and forwarding ports.  Its especially preferable if
> > the application is the only server a user is expected to be running.
> >
> > LOCAL_ADDR is marginally helpful too.
> >
> >
> >
> > ----- Original Message -----
> > From: bill lam <bbill....@gmail.com>
> > To: programm...@jsoftware.com
> > Cc:
> > Sent: Tuesday, February 11, 2014 11:29:50 AM
> > Subject: Re: [Jprogramming] QT websockets
> >
> > There is REMOTE_ADDR and also some othe cgi environment variables
> > but I'm not sure if webosket protocol is required to provide
> > cgi environment variables.
> >
> > If security is a real concern, I guess it is best to put a
> > websocket server behind a webserver or other proxy. What we can
> > done is unlikely to be better than those specialised products.
> >
> > Вт, 11 фев 2014, Joe Bogner писал(а):
> > > The websocket demo works well and I can see a lot of potential with it.
> > > Thank you! I also agree with Pascal that it would be useful to get
> > > information about the connecting socket (e.g. originating IP address).
> > The
> > > most basic security workaround is to add a firewall rule to prevent
> > > connections on the port from outside interfaces. It might be helpful to
> > be
> > > able to bind to a specific IP/interface like a socket too.
> > >
> > >
> > >
> > >
> > >
> > >
> > > On Tue, Feb 11, 2014 at 9:21 AM, Pascal Jasmin <godspiral2...@yahoo.ca
> > >wrote:
> > >
> > > > nodeJS calls the function remoteAddress, and pants calls it
> > getpeername.
> > > >
> > > > In terms of "security already handled by websocket protocol", I can
> > see a
> > > > necessity for a user code authentication layer, but passwords might
> be
> > > > compromised, and an IP could attempt millions of logins.  The
> millions
> > of
> > > > login attempts could be an intentional DOS event.  There is a
> > simplicity to
> > > > requiring 192.168.x.x IPs for "root" (privileged) user code access to
> > > > server or simply any access.
> > > >
> > > > Running the demo for instance allows the world to execute ". on your
> > > > machine. ?
> > > >
> > > >
> > > >
> > > > ----- Original Message -----
> > > > From: bill lam <bbill....@gmail.com>
> > > > To: programm...@jsoftware.com
> > > > Cc:
> > > > Sent: Tuesday, February 11, 2014 3:48:36 AM
> > > > Subject: Re: [Jprogramming] QT websockets
> > > >
> > > > I think those low level functions are not part of websocket
> > > > spec.  Security is already handled by websocket protocol.
> > > >
> > > > Since J is single thread, those globals will not cause any
> > > > trouble. A handler can erase them if they are no longer used.
> > > >
> > > > Anyways, IMO, to compare socket with websocket is analogous
> > > > to compre an apple with an pineapple.
> > > >
> > > > Пн, 10 фев 2014, Pascal Jasmin писал(а):
> > > > > demos work fine. (and avast makes no complaints about the zip
> > version of
> > > > J8)
> > > > >
> > > > > I notice that sdgetpeername and sdgetsockname do not work on
> > websockets
> > > > (understandable if they are not sockets) and there is no QT
> equivalent
> > for
> > > > the functions.
> > > > >
> > > > > Is that something not implemented by QT, or not implemented by the
> J
> > > > version?
> > > > >
> > > > > I can provide use cases, but the most obvious ones are centered
> > around
> > > > security and logging, authentication, and tunneling or calling back
> > > > disconnected nodes based on IP addresses.
> > > > >
> > > > > I notice that wss0_jrx_ , wss1_jrx_ are "global" variables that
> > persist
> > > > through connect/disconnect (calls through handler).  Perhaps these
> > > > parameters could be set to peer/sock name/IP for the onopen event,
> if a
> > > > qtgetpeername/sockname is somehow difficult?
> > > > >
> > ----------------------------------------------------------------------
> > > > > For information about J forums see
> > http://www.jsoftware.com/forums.htm
> > > >
> > > > --
> > > > regards,
> > > > ====================================================
> > > > GPG key 1024D/4434BAB3 2008-08-24
> > > > gpg --keyserver subkeys.pgp.net --recv-keys 4434BAB3
> > > > gpg --keyserver subkeys.pgp.net --armor --export 4434BAB3
> > > >
> > > >
> ----------------------------------------------------------------------
> > > > For information about J forums see
> http://www.jsoftware.com/forums.htm
> > > >
> ----------------------------------------------------------------------
> > > > For information about J forums see
> http://www.jsoftware.com/forums.htm
> >
> > > >
> > > ----------------------------------------------------------------------
> > > For information about J forums see http://www.jsoftware.com/forums.htm
> >
> > --
> > regards,
> > ====================================================
> > GPG key 1024D/4434BAB3 2008-08-24
> > gpg --keyserver subkeys.pgp.net --recv-keys 4434BAB3
> > gpg --keyserver subkeys.pgp.net --armor --export 4434BAB3
> > ----------------------------------------------------------------------
> > For information about J forums see http://www.jsoftware.com/forums.htm
> > ----------------------------------------------------------------------
> > For information about J forums see http://www.jsoftware.com/forums.htm
> >
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to