There are lots of times when target printers are narrowly specified and you don't have the target printers available.

In such cases, you can and IMHO should ask your users to cooperate in the test. Install the appropriate driver, design with the target printer in mind (for example, pay attention to the non-printable page margins, whether you use "printable page" or "whole page" to design).

Then print to a file. I don't mean ASCII, I mean create the .PRN file that would represent the spooled output to that printer.

You can give this PRN with a small batchfile to the user for test printing. Usually COPY /B > LPT1 is enough to do the trick; if it's a network printer they'll need a little more help but you can do it.

The point is that users are aware that their environment is necessarily not exactly the same as yours. Or if they're not, it's good for you to make them aware of this important thing and prepare them for testing with it in mind! This is true for other reasons than narrowly-specified target printers of course <s>.

Getting them involved with small, easy tests shows them that you are aware of potential differences and are planning for it. You are proactively looking to validate your process with the appropriate environment in mind, not a hothouse development setup. It is not an admission of inadequacy, inefficiency, or weakness to ask users to share this validation process with you.

It also allows them to sign off on the results at an early stage. Suppose the printers do print the small point size and you verified this alone. The results might still be illegible by their standards. If you wait until later to find this out, what will you do? Better to negotiate a different solution now if necessary. If they *do* sign off, they can't object later. <s>

Sorry, this is probably more of a general answer than you wanted. I used to teach RAD/JAD and am still passionate about it <s>.

>L<