I have an application that needed 20 table of occurances of the same table. I just wonder whether there is a limit to the number of TO ? If I have that many TO, will it slow down the system or more prone to data corruption?

Number of tables per file: 1 million.

Click here and this should answer all questions on limitations

There's a subtle difference here - TABLES and TABLE OCCURRENCES.

But 1 mill seems sufficient in either instance.

1 mil is more than sufficient for me. I guess my other question is: Is it normal to have 20 over TO? I tried to 'trim' them down but still ended with about 20 TO. I like to hear from other developers whether this is normal ..

I have 400?

It doesn't really matter, as long as you can control their layout, know where they all are so you can find them quickly.

You have to remember that all they are pointers to tables -- they store no actual data, hence take up no real room.

I tried to 'trim' them down but still ended with about 20 TO. I like to hear from other developers whether this is normal ..

This is highly denpendant on the approach you lean against, 4 tables, does easily translate into 20+ TO's if you're anchor bouy'ing with say 5-6 TOG's.

The notion of trimming down doesn't nessersarily lead anywhere healthier, you should use the number of TO's required to solve a problem, but if all your TO's is tied to one single web, isn't cutting the number the same as getting organized.

Each TO is a metaphor that hardly takes up any place, I once saw somebody measuring what it did to the filesize, and we're talking less that 30 kbyte as I recall it.

But another way to put it, is how many globals does your entire solution hold, this figure reflect inversely how healthy your relational measures are working! One of the big issues to be aware of is that the RG we use in filemaker NOT is an ERD!!!

--sd