reading from notecards

You're kidding, right?

I have to call the llGetNotecardLine() function and then *wait* for a dataserver event to fire in order to recieve that line? Then, if the returned data isn't EOF, I have to repeat the cycle?

Instead of being able to call LLNotecard2List(string notecardname) and slurp the ENTIRE notecard into a sorted list with one function?

Or llNoteCard2List(string notecardname, integer start_line, integer end_line) to specify which lines need slurpin?

You got it. I think the contents of the notecard are stored on a different server somewhere, so your sim has to send a request out and wait for a response from that other server. And since the notecard can be arbitrarily large, you're only allowed to read one line at a time. This isn't like reading a file off a hard drive on a computer, the environment is totally different.

I find reading 1 line at a time silly, and therefore never use it ...
and from a traffic standpoint if your gonna have a large notecard, you have already done it, so now instead of 1 lump of info being called and transfered, you have 20 llget cycles and a new session for each one.

dunno seems kinda dumb to me

It's, technically, a backend database request. So no, it's not silly. If you want data that is immediately accessible, store it in the script or in the prim the script is in

Store notecard data in the script? Umm, no. It seems like scripts hardly have enough memory for code / heap / stack, much less storing gobs of notecard data. Never mind that scripts often need to be locked down while the notecard remains editable.

How do we store notecard data in the prim, exactly? Convert it to a texture? Convert it to geometry? Thatd be a strange lookin prim.

The fact that we STILL have no persistent storage for scripts is just....grrrr.

For example, you could store quite a lot of raw data in seperate scripts, then retreive it using link messages. Just put it in strings stored in a list.

thats what i do, make a dumb script with linked messages for the user to edit if need be

oh and its alot faster too

Sorry about my bitchy reply, Jes. The hour was late. Storin text in seperate scripts sounds like a great idea. Thanks!

Stoopid noob question -- but how might one go about doing this? Is there a small example? This may solve a problem I have with something I'm doing now.

Thanks a bunch!

I know it's a whole year since Osgeld posted this, but I'm looking for the answer so I hope that reviving this thread will shed more light on the question ....

I have doors in a house (wow!!). They each have their own script - don't want all the access points to open every time someone goes in/out.

The owner of the building should be able to create a single 'allowed' list - ideally the list will reside in one prim; the owner can edit it; the doors will each consult this list when touched, to check whether they should open.

I thought this was a job for a notecard - something I haven't yet done from scratch - and, while looking for advice about how to do it properly, came across this thread.

It seems I should put the list in a 'control' script, then make the door scripts fetch the allowed list from there? Can anybody tell me how to do it and/or point me towards instructions, please?

Thank you ....
Cherry

Hi Cherry

Some doors already have lists for admissable ppl (should that be certified?). You could write a control script that you interface with though a hud and which in turn relays your instruction to every door. Or you might update one notecard in the controller and then have it reiterate the notecard to the doors.

I also have the controller script lock all the doors at once after issueing a "close door" command. This might need shouting for some of the bigger houses that are not linkable as a single prim.

hth

Ed

using the control script isn't difficult. Just have a script in a single prim that takes a name or other info, checks to see if it's on the list, and returns a "yes" or "no" appropriately. (or any other data you want to send back.)

The fun part with what you're describing, however, will be how to send and recieve the input. If the prims are linked (and i doubt they will be for a house full of doors) you can use MessageLinked. However, chances are, this house is not linked, so you're going to have to use llSay on some non-zero channel to pass the info back and forth. If the doors are far enough from the controlling prim, then use llShout.

As for using scripts versus notecard, though I do realize that this post is old, scripts are faster, though if you sell the item you have to make sure the new owners know not to edit the functioning part of the script and only edit the list of information. Perhaps you can provide an interface for them to add, or remove information from the list without getting into the script itself, though this adds the problem of the changes being lost on script resets.

A notecard is a better way to go for setting variables, or other settings that a script uses, as it only needs to read these on script reset.

I am looking to use a notecard for a script I am working on, but lukily all I need to do is pick a random line and read the one line, about once every 5 minutes or so when the script is active.

since that time sims have changed, back when sim fps's were in the 15,000 range, and a average notecard line took 1 or 2 seconds the choice was obious

nowdays notecards are almost instant if stored in the objects inventory, and in most cases just a shade slower than linked messages

course the advantages are

not having to make a config script (and adding to overall sim stress)
and the user doesnt have to fool with a script

There are other ways of storing small amounts of data, such as using the script name and description fields. Its limited but does work. And ist suprising how much data you can pack into 255 characters if you are careful.

More recently Squirrel Wood extended the idea by using multiple blank notecards where the name of the notecards was used as the data store. seehere for details.

Be carefull! I believe a name can only hold 63 and a description can only hold 127 or so characters.

Hello

I just wanted to say thanks for all the advice .... I've decided to code this dang thang from scratch (ie, without lifting chunks from other people's scripts) and it's going to take a while
It'll be good for me!!

I'm not sure I'll make it as far as mastering Squirrel Wood's mystery blank-notecard technique but who knows? At my current rate of Wiki reading, anything could happen (including, but not limited to, a new glasses prescription).

Btw, the house I'm making isn't large but I am trying to work to best practice. When I've finally got the door scripts right, I hope you'll point out possible improvements

Cheers,
Cherry

Some random suggestions:

Use "," instead of "return)" in your notecard and after getting the first line run it through llCSV2List().

These days I try to keep the number of in-world scripts in my sim down to a minimum and have adopted a set of "do and die" utility scripts that clean up after themselves like:

default { state_entry() {

	 llSetText( llGetObjectDesc(), <1,1,1>, 1.0 );

 	 // or llSetTextureAnim( blah, blah );
 	 // or llSetTargetOmega( blah, blah );
 	 // or llParticleSystem( blah, blah, BLAH, blah, blah );
 	 // or whatever other persistent effects I want to configure...
  	
	 llRemoveInventory( llGetScriptName() );
} }

Don't forget you can also use the NAMES of items in inventory to store data as well!

Unfortunately, there doesn't seem to be a way to query things like "for sale" status, orig/copy and sale price of prims or inventory, it'd be a useful ability for vendomat gizmos.

I don't mind reading notecard lines one at a time... what I DO mind is animation overriders having to actively poll for agent state instead of having a change in agent state trigger an event. :-/

Those are all really good ideas! Thanks!

With regards to removing the script after it's done its job - I've only just discovered that scripts use up sim memory & time, even when inactive

Are they still wasteful if set to 'Not Running'?
This has implications for items that are to be transferred: it might make sense to include the delete line in animation/loop/particle scripts, then provide a control prim to start the script - which will then delete itself.

What do you think?

Cherry

In the case of offsite data, i tend to set a var to true or false. so unless the script is reset it should keep the data, including logging off. but if it looses it it grabs it from the http server for that person. and they can also issue a save state which pushes back to the outside server but i try to limit my calls.

names can hold upto 255
discriptions are 127

Well I originally picked up Hank Ramos's free dataserver example script and have pretty much used it ever since. No point reinventing the wheel , unless you like triangular wheels...




Squirrel's Idea is very nice and simple. It removes the need for any understanding of the more arcane Dataserver event. Its drawback is that it requires a number of empty/unused notecards to be added to the object.