>I think that has been suggested several time to Advisor, but it never worked, because evaluating and creating surveys takes too much time and people won't spend the time to answer them if they don't get anything free in return.

I suppose setting up a survey is too much to ask. How about specifically encouraging readers to drop a line letting you know if any individual Answers or Tips are considered either of substantial value or too little value. How else do you folks get any feedback about your choices?

>Yes, it would do, but this is actually a frequently asked question. Many people are not aware of this limit in VFP. And while you could answer this question in one sentence, our goal in Ask Advisor is to provide more information on each question, cover some related topics and explain why things are the way they are.

OK, I'll accept that. I just thought all that stuff with calculating the sizes from the bytes in the header, etc, made the whole answer way too long, given the topic.

>This raises the question on how much comments a tip needs... What do you think?

Steve add comments pretty often, and they generally help in this area. I would like to see a screening mechanism that looks at the question of whether it would be appropriate for a less-than-expert VFP developer to implement the tip as-is. If not, I would either edit it or not use the tip.

I would further avoid publishing tips that are only appropriate from people working in a "VFP desktop sandbox". By that I mean approaches that, if included in applications, preclude migration to other back ends, extensions to n-tier, etc. An example is Joanne Rogers tip on the Minimum() function in SQL. While this is a perfectly valid solution to the problem she posed, the approach of using UDFs is SQL statements leads developers not to learn solutions that are valid within an ANSI-SQL context, and thus constrain the solution to local VFP data.

I suppose my point here is really about publsihng tips that fit in reasonably with the mainstream types of applications and components that VFP developers are likely to be working on from here forward.

-- Randy