gratuitous inclusion of oracle in 7.0.2-RC2

[Richard Gaskin]

J. Landman Gay wrote:

 > The project I've been on lately has hundreds of stacks and thousands
 > of support files with gigabytes of data, they're just not all in
 > the main app. It's easier that way. The app itself takes under 15
 > seconds to compile.

That's a good strategy for many reasons, but I still believe compiling 
any standalone, even a complex one, should never take more than a minute.

At the heart of the build process for the desktop all that's happening 
is the mainstack file is bound to a copy of the runtime engine, which by 
itself takes about a quarter-second.  Depending on your settings there 
may be other steps, like moving stacks in or out of the standalone file, 
clearing a few custom props, etc., but 15 minutes is a looooooooooong 
time for a tool famous for offering zero-compile-time.

I've run code base analytics on fairly complex stacks, and the worst of 
them took less than a minute.  Perhaps what LC is doing is even more 
complex, but given the speed with which LC can traverse objects and 
parse text I'm fairly confident what we're seeing here is a bug, likely 
from some redundant script analysis.

Let's look at the worst cases first: If any of you have stacks that take 
longer than 5 minutes to build a desktop standalone, and you're in a 
position to share the stack with RunRev's support team so they can 
figure out why it takes so long, please open a bug report against this 
so we can find out what's going on.

Even better:  if anyone here has some time on their hands and is up for 
a challenge, it may be worthwhile poking around in the Standalone 
Builder scripts to see if you can find where the bottleneck lies.

-- 
  Richard Gaskin
  Fourth World Systems
  Software Design and Development for the Desktop, Mobile, and the Web
  ____________________________________________________________________
  Ambassador at FourthWorld.com                http://www.FourthWorld.com