[sldev] Point of no return in teleporting?

[John Hurliman]

Soft Noel wrote:
> On Wed, April 11, 2007 6:52 pm, John Hurliman wrote:
>   
>> Teleporting in Second Life is something of a gamble. If the grid is
>> healthy and concurrency is normal, your teleport will generally always
>> succeed and things are fine. However at times like right now, you may or
>> may not ever reach your destination, and you may or may not get
>> disconnected in the process.
>>     
>
> Just in case the bot code is introducing a new problem --
>
> Are you certain you're getting disconnected when using the regular viewer?
> I've not had that happen. What I do have happen now and then is that I get
> pegged in place, unable to move, which looks a lot like a disconnect.
>
> There's a blob of code that does an early out in send_agent_update() if
> the teleport state isn't TELEPORT_NONE. This effectively swallows outbound
> movement events once teleport goes active, and the state doesn't seem to
> get reset when a teleport fails. I haven't been able to reproduce it with
> the debugger attached to verify yet.
>
>   

Yes the official viewer was disconnecting me, and one way to trigger it 
for sure is if the CAPS server is down but the simulator is up. You will 
receive the UDP portion of the transaction (TeleportStart, 
TeleportProgress, TeleportProgress) but the TeleportFinish CAPS event 
will never trigger since you don't have a CAPS connection. Another way 
to see this is if you teleport around fast enough that the CAPS 
connection isn't established in time. I'll be changing the libsecondlife 
code soon to disallow any actions that are guaranteed to fail without 
CAPS if the connection isn't established yet, would it make sense to do 
something like that in the official viewer as well?

John