>1. What do you mean by a 'Multi-Use DLL'? Is this something special or just a VFP DLL?

Each class yo build can be marked to execute in a certain way. Multi-use means
(at least for DLL server) that multiple instances of the class can be created
on separate threads. Single use means that only a single instance total can
be created or a COM error will occur.

DLL compoents should always be marked for multi-use. EXE components also have
those settings but they mean something different. EXE components marked as
Single Use start a new process (read EXE viewable in task manager) for each
object instance. If you create three instance you'll end up with three distinct
instances of VFP running your COM server. Multi-use means that a single instance
is shared by all users of the object.


>2. In deciding whether to build a DLL or an EXE, what would you say should be the prime consideration? Stablity? Speed? Other factors?

With ASP, using EXE servers is not a good idea. Build DLL servers. DLL servers
tend to be less stable and can crash the host when they fail, but htey are
slightly faster in call interface (not operation of your component though -
just parameter and return value passing).


>3. When you are talking about 'running the component'. Does this happen on the server or on the user's machine?

On the server. Clients access a Web page which responds and runs the component
on the server.


>4. If I set up an ASP that calls my DLL, how do I setup and distribute this? Is it something that each user has to install or just something I set up on the server?

You don't. You only have a single server (the Web server) which uses your
component. Everybody accessing the app just uses a browser (or HTTP client
appliaction, which is something you may be able to do to get the ID
non-interactively via code - see the Distributed HTTP apps article on the
articles page you visited before).


>1. How do I find out what the SYSTEM path is set to?

I think it's in that article - you need to track back the starting location
of the DLL then SET PATH to that and any related directories. You don't
want to be writing anything into the SYSTEM directory.

>2. If the SYSTEM path is a directory on the server, how do I 'get at my data' if the data is sitting on the client machine?

See above.


+++ Rick ---