FM7 loops slowly...why use a loop ?

[Ugo DI LUCA]

Hi,

Someone was complaining that FM7 is 4 times slower than FM6.

The script was generating 12,000 records, each filled with numbers from gMin to gMax, and while 1 minute 12 seconds where needed with 6, FM7 needed 4 minutes 30 seconds to complete the same operation with exactly the same loop.

This alternate technique will create 72,000 records in less than 1:12:00 so in this case FM7 is 6 times quicker than FM6

Check it out.

RandomCreation.fp7.zip

[Brian C]

Looks good and makes sense. Thanks for providing a faster way of accomplishing this.

Could you point me towards the slower example or explain what was done in it? That way I can compare and understand what to avoid doing in the future!

Thanks!

[Brian C]

I must say, your example is blindingly fast.

Would it be safe to assume that the original concept used a single table approach and used the random function in some fashion to derive the random result? This seems to be the obvious and most direct approach I suppose.

[Ugo DI LUCA]

Hi Brian,

Here's the loop as it was formerly designed involving FM5.5. This is the file I posted which involve 2 methods so to demonstrate that the second method there wasn't reliable. But even though it creates 6,000 records with one method and 6,000 with the other, it is still fast.

Try to convert it to 7, or recreate it from scratch with 7 and you'll see how slow it can now be.

If only FileMaker had fixed the repeating bug on import, you could create related records with this method. When it was working with FM7v2, it was really smart.

Alea.zip