Hello:

We've created a new YouTube video on efficient Basic+ coding techniques. Please visit the Tutorials Page (as seen on the navigation menu on the left) for all the latest.

Thanks-

Revelation Software

Excellent.

The more I see, the more I like.

The more I like the more I look.

In order to not keep checking this site or You Tube for new tutorials, I have added Revelation Software to my You Tube subscriptions and receive emails when Revelation adds new vids.

Thank you, Revelation!

Clark Westfield

Nice .. you might want to 'package' things like this so the new user could download and install the code to follow along.

But nice .. pretty soon mike wont have to be on the road so much

Great video Mike however I have one comment. On your testing of the difference between Select and Btree.extract the results from test1 (Select) appear to be faster than test2 (btree.extract) but you state that Btree.extract should be faster than Select…

John

and to squeeze that extra bit of performance again

it's more efficient just to concatenate the field mark regardless then remove the trailing one

pre.code {

  background-color: #E5E5E5;

  border: 1px solid #000000;

  width: 450px;

  padding: 5px;

  font-family: courier, verdana, arial, serif;

  margin: 0px 10px auto;

}

   Array="
   For i=1 to 15000
      Array := i : @Fm
   Next
   Array-1, 1='

shaves of another few percent. Oh yes, and use TimeGetTime instead of DosTime. It's more granular.

pre.code {

  background-color: #E5E5E5;

  border: 1px solid #000000;

  width: 450px;

  padding: 5px;

  font-family: courier, verdana, arial, serif;

  margin: 0px 10px auto;

}

  WINMM
  LONG STDCALL timeGetTime(VOID) 

The Sprezzatura Group

World leaders in all things RevSoft

you mean instead of the

if len(…..

yep .. just stuff it on. if you need to remove it at the end great but no reason to IF and LEN

I think the cause of BTREE usage being slow is due to my belief that using BTREE is sometimes like putting out a lit match with a fire hose. Too much for too little.

If all you are looking for is to get a list of invoices dated 01/01/07, then using an index that just simply reads one row, with field mark delimited list of rows that are dated 01/01/07 is very quick. Just a hint of suggestion.

That sounds like a relational index…

Another update-

We've added two new videos, one dealing with Importing Pick-based applications into OpenInsight, and one for the installation and configuration of the OEngine Server.

Thanks again

Guys-

I just tested both ways- and it looks like they're both very similar, but with the demonstrated example about .01 seconds faster for the 15,000 iterations.

Statistically, I'd call that a tie.

Mike

I agree… I think I should have mentioned that. My bad.

A 'statistical' tie for sure but since we do strings

Keep trying to tell those SQL guys they need to think strings

How about a rss feed.

View this thread on the Works forum...