datagrams were fragmented and reassembled at the network layer,
but apparently that's not something one can count on.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@5770 542714f4-19e9-0310-aa3c-eee0fc999fb1
sending it on the stream. However, also increase the maximum
datagram size.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@5670 542714f4-19e9-0310-aa3c-eee0fc999fb1
initiated by the client, but we always want to stop our ticking.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@5519 542714f4-19e9-0310-aa3c-eee0fc999fb1
polled socket system. This much more cleanly and efficiently integrates peer
(and other server to server) connections into the normal server I/O framework
and should also eliminate for good the annoying server hangs that result when
our old blocking client I/O threads got their pants wedged.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@5510 542714f4-19e9-0310-aa3c-eee0fc999fb1
second rate (10 messages per second), and to do the same 10x buffer on the
server that we were doing before so that we avoid disconnects due to network
congestion (which might "save up" some throttled messages and then end up
delivering them in a bunch with some later messages).
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@5424 542714f4-19e9-0310-aa3c-eee0fc999fb1
communicates that limit to the client which queues outgoing messages so that it
does not exceed the limit. When the throttle is changed, the server does not
apply the new throttle until the client has ACKed it to avoid edge cases where
the client is up against the limit when the throttle changes.
ActionScript side of all of this forthcoming as well as more testing.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@5422 542714f4-19e9-0310-aa3c-eee0fc999fb1
Note: this involves generated code from a samskivert patch I only just sent to mdb,
but he's in a meeting, so that won't show up for a little.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@5369 542714f4-19e9-0310-aa3c-eee0fc999fb1
importing receivers from their senders to satisfy their links.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@5333 542714f4-19e9-0310-aa3c-eee0fc999fb1
launcher which manages a lot of clients, but seems generally useful.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@5322 542714f4-19e9-0310-aa3c-eee0fc999fb1
fired in such rapid succession that we have not completed processing the first
call before the next one comes in. We could do the cumbersome thing and keep
track of an _isLoggingOff state but that is so inelegant that I can't bear to
do it. I'd rather just cope in the Presents code and put the onus on the caller
not to do wacky things like this.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@5296 542714f4-19e9-0310-aa3c-eee0fc999fb1
outstanding know that their subscription request failed.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@4873 542714f4-19e9-0310-aa3c-eee0fc999fb1
(which runs on the server) that we do in a normal client (which runs on a
user's computer). Eventually we'll make a custom Communicator for use on the
server that integrates with the ConnectionManager's polled socket I/O which
will be easier now that this bit of factoring has been done.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@4835 542714f4-19e9-0310-aa3c-eee0fc999fb1
separate thread and racing against other client actions. Also nixed a pointless
type check and cast.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@4828 542714f4-19e9-0310-aa3c-eee0fc999fb1
everything in two languages. Especially when one doesn't support inner classes
and has non-C syntax just to make life easier for parser writers.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@4762 542714f4-19e9-0310-aa3c-eee0fc999fb1
and the ProxySubscriber mechanism by which the server proxies for the client.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@4730 542714f4-19e9-0310-aa3c-eee0fc999fb1
another server. The PeerManager needs both to rewrite the oid of the events
before dispatching them and to assign an eventId to them so that they will not
be filtered by PresentsClient when deciding whether or not to send them along
to its client. Also fixed compound event dispatch in the process. Now proxy
subscribers will be notified once of a compound event rather than of each
individual internal event.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@4729 542714f4-19e9-0310-aa3c-eee0fc999fb1
to fail if a director tries to register services groups but was not created
until after the client was logged on.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@4557 542714f4-19e9-0310-aa3c-eee0fc999fb1
notion of a global group (though we implicitly define one in InvocationCodes)
added a mechanism for directors (which generally handle the client side of
invocation services) to register their interest in bootstrap service groups so
that the whole goddamned complex business can happen magically behind the
scenes.
If you instantiate a director, it will automatically register interest in the
service group it needs and everything will work. If you don't use the director
code, you don't get the services and you can safely exclude all of that code
from your client even though the services are still in use on the server (and
presumably used by some other types of clients).
This is going to break all the builds, which I'll soon fix. Then I'll go write
all this in ActionScript. Yay!
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@4552 542714f4-19e9-0310-aa3c-eee0fc999fb1
MetaSOY clients (Swiftly, World, and soon Admin Dashboard) don't have to know
about a bunch of unrelated crap. Fricking complexity++, grumble.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@4551 542714f4-19e9-0310-aa3c-eee0fc999fb1
inner bits are all clear and ready to try logging on to a new server. We
specifically wait to call clientLogonFailed() until after everything is
cleared, but I don't want to change the behavior of clientDidLogoff() for fear
that any one of the zillion directors depends on something being around during
that call.
Instead I added another callback clientDidClear() that is called when a
successful session is fulled cleared out (basically when we're not calling
clientLogonFailed()) so that entities that are waiting for a logoff to finish
so that they can logon somewhere else have a clean way of doing so.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@4255 542714f4-19e9-0310-aa3c-eee0fc999fb1
make connections to other servers in the cluster and can exchange events (in a
limited fashion).
This is different than Liz's project wherein servers share an oid space and one
can interchangably work with distributed objects from any server. This package
provides a means by which certain services (by default, presence and chat) can
be communicated between servers to allow communication between players
scattered around a bunch of otherwise independent server instances.
This is less general purpose but also less likely to encourage people to write
code that tightly couples multiple servers and then falls over because it
generates gobs of network traffic as events are flung willy nilly behind the
scenes.
git-svn-id: svn+ssh://src.earth.threerings.net/narya/trunk@4238 542714f4-19e9-0310-aa3c-eee0fc999fb1
Dave Hoover