Persistent LC server?

[Richard Gaskin]

Monte Goulding wrote:

 > Hi Richard
 >
 > I had a look at the engine code for you and I don't think you will
 > get this to work without some modifications to the source. If you
 > look at the server implementation of X_main_loop you will see that
 > unlike all the other engines that repeatedly do stuff while the
 > MCQuit flag is false the server engine just does it's stuff so it
 > will naturally quit at the end.

Very interesting.  Thanks for that background.

In your code base exploration have you discovered why the Linux Server 
engine is nearly 50% larger than that non-server Linux standalone engine?

I can't imagine the extra merge functionality alone accounts for all 
that space.


 > Is there a reason you couldn't use the -u flag on a regular
 > standalone?

None at all.  In fact, I've been using standalones for CGIs for some 
time (I started back when LC Server didn't support stack files as 
libraries).  I've grown to prefer them for many tasks, esp. given my 
habit of including only function calls (rather than complete handlers) 
in my HTML anyway.


 > My biggest question though is what event is triggered on the
 > persistent server when you request some part of your array?
 > LC Server scripts are basically just one big startup handler
 > without the handler declaration.

Right.  If I were to pursue this I'd probably write a quasi-event look, 
watching a particular socket.

At the moment, though, I'm on a different path, since I discovered a way 
to make an array-like file I can access elements within from a cold 
start at about 0.02ms per access, fast enough for CGI use.

--
  Richard Gaskin
  Fourth World
  LiveCode training and consulting: http://www.fourthworld.com
  Webzine for LiveCode developers: http://www.LiveCodeJournal.com
  Follow me on Twitter:  http://twitter.com/FourthWorldSys