---
If you described anything other than UDF related issues, I missed it.
---

You certainly did. I don't think you quite understand the Preview implications.

---
so I don't see them as a good excuse for having _pagetotal implemented as it is.
---

Nobody is making "excuses" for anything, Doru. And nobody is asking you to agree with me, or with MS, or anything of the kind <s>.

However there are many differences of opinions about how to use reports <s>. For example, it is possible to run a report and have UDFs trigger a table of contents that is built dynamically during the report run. It is literally impossible to do this without doing it *during* the report run, if you want page numbers in your table of contents <s>. Can we do it with a "silent" prepass that we force on our own? Yes -- but you can get a pagetotal the same way .

Many such practices exist, whether you approve of them or not, whether you have the need or not. They all require some sort of evaluation ("am I in the report pass that should trigger this behavior?") and PageTotal is not particularly special in this respect.

When the idea of an "automagic" prepass became necessary for _pagetotal in particular, IMHO there is **no way** that this prepass should have become standard behavior.

Having the ability to *assess* the nature of the current report pass *is* a good idea -- for _pagetotal and for any other work that one might want to do, for purposes of work that should not occur during a Preview but should occur during a Print, etc.

Giving us the ability to assess the pass, force a prepass, *or* prevent a prepass, is what they decided to do -- with the default behavior remaining backward compatible *unless* the new _pagetotal was explicitly used. Pretty much the best of the choices available.

This type of reasoning is fairly standard when a change is made in VFP, and it was very thoughtfully done in this case -- always assuming that _pagetotal was such a dire need in the first place that it deserved all this care and thought -- as you know I don't think it was <s>.

And, yes, there are times that the VFP team takes a different approach when they make a change. But the change in GROUP BY reflected "bad" or "good" design based on conformance to SQL standards, not anybody's whim. To me that's kind of a different story, but IAC, you don't have to agree with me about that either.

>L<