dataserver... more than one ?!?

hi...

i have a script , that uses llGetNotecardLine to read the config-notecard, so it uses a dataserver.

now i want to use llRequestAgentData although.... what now ?

as the dataserver-konzept is somewhat strange to me, here is my limit of understanding.

can i have two dataserver in one LSL-script ?
or can one dataserverserver handle multiple types of requests ?

Wiki doesn´t gave me an idea here....

how to ?


cheers
ralf

using one dataserver to serve two purposes could be done by grabbing the key for the request and storing it in a global.

psudo-code:

key notecard;
key userdata;

default {
state_entry()
{
  notecard = llGetNotecardLine...
  userdata = llRequestAgentData...
}

dataserver(key returnkey, string data)
{
  if (returnkey == notecard)
  {
    //handle the notecard data
  }
  else if (returnkey == userdata)
  {
    //handle the userdata
  }
}
}

You'll need to save the request keys, and check them in the dataserver event. Something like:

global variable key notecardRequestId;
global variable key agentRequestId;

...

some_event_handler

  notecardRequestId = llGetNotecardLine(whatever);

...

  agentRequestId = llRequestAgentData (whatever);

...

dataserver

  if (id == notecardRequestId)
    // This is in response to the notecard request

  else if (id == agentRequestId)
    // This is in response to the agent request

  else
    // Bogus dataserver event, ignore

well..

that make scene and even gives the word "server" more content....

in times i learned, we would have said subroutine, but times change


will try that

thx

cheers
ralf

I'd say that subroutine would be closer to a function within the script, the dataserver however is accessing the asset server (or an associated database) in order to pluck data from it.

You say tomato and well, so do I.

EDIT: What they said below

You can think of any function as a subroutine, however event handlers are a special case.
Event handlers will block processing of any other script code while they run, think NMI.
They are called in responce to an EXTERNAL trigger. Both the names of the event handlers, their parameters and their triggers are predefined, all we do is flesh out what we want to do in this particular instance.

Anything written within the bounds of a state is an event handler.

MySubroutine()
{
}

integer MuFunction()
{
    return 0;
}

default
{
    state_entry()
    {
    }
}

In LSL, any code, function or subroutine will block processing of any other script code, because everything we do is within the scope of an event handler. Otherwise we'd have to manually deal with concurrency issues.

-peekay

True, I see thast not obvious from what I've written my apologies for the ambiguity.

>> Event handlers will block processing of any other script code while they run

Yep, my favorite trick for interuppting an event handler (say, a user clicks on a prim) is to have a second script which can change an attribute (change the alpha or some other param). The event handler in the first script does a check for the change in its loop.

example: I have a photo frame that scrolls llSetText() above the object. The text might be a paragraph, and it could take 10 seconds or more to get through. Having a second script that can do something with a touch event gives me a way to break out of the loop.