Optikal

is this a .Net question????

There is a problem with your UpdateCommand, InsertCommand, or DeleteCommand on your data adapter. Make sure the SQL/sproc you are using is fully tested. This is one of my major gripes with using data adapters, is that its difficult to debug the Update() method, so much crap goes on, and you have no visibility as to how or where it fails. (If anybody knows of a way to debug whats going on easily I'd love to hear it).

You should be able to download that many rows easily. I have code that downloads several hundred thousand records and it works fine. Try creating a simple test project that just connects and downloads the table to see if it works, if it does then it must be something else you are doing. PS - If you just want to update the data, by far the most efficient way would just be to send an UPDATE sql query via ExecuteNonQuery(), rather than having to download and re-upload the data. This may or may not be possible depending on your situation.

You can pass data to a new thread by setting instance members of the object which contains the instance method which you use to launch the thread. A full explanation and example can be found here: http://msdn.microsoft.com/library/en-us/cpguide/html/cpconcreatingthreads.asp

noah: yup, sounds like you understand it

You can write to the GraphicsStream returned by LockVertexBuffer

Sounds like you have a memory leak/bloat. Are you creating a whole bunch of data in memory and not freeing it?

pointsFvF is a stream containing your list of vertices you want a bounding-box for. numVertices is the number of vertices you just passed in. vertexFormat is the format your vertices in the first parameter are in. min and max are filled in by the function and are vertices representing 2 opposite corners of the resulting bounding box.

Your web service application is going to have to expose this information somehow. A log file sounds like a good idea to me.

All good points, but alot of them I don't think I could use. Email - we already have a custom component that allows us to send email, so we only need a few lines of code currently. Event Logging - not somthing we've ever used Colors/Style - good if your developing commercial software, but I doubt we'd want to create flashy apps here Registry - again, we have a custom component that wraps this functionality up and makes it available with minimal coding Menus - can you elaborate a bit on this. I know I like the in-form designer better than VB6, but I don't know much else about differences. Multithreading - thats a good one, I can think of quite a few instances where multithreading would be nice Inheritance - not a business benefit, just a nicety for programmers Good input though, keep it coming :).

What do you mean by RealTime language?

Afterlife, OOP is great and I love it, but its not a business benefit. I have to focus on what types of things will this great new OOP capability allow us to do that we couldn't previously do, and thats worth investing $500,000 to get.

No, No, and No (however we do support offline clients, and its a huge pain in the butt). I will definately look into those on the weekend. Very good ideas. Thanks.

well, I believe our ops guys are in the process of upgrading our clients to Win2000 (though this will probably take a LONG time). But lets assume that all our clients are running Win2k+, and continue this discussion, because I'm still interested in trying to make a business case out of this.

Well longhorn doesn't come out until 2006 at the earliest, and considering the majority of our clients are running Win95 (I just realized I don't even know if the CLR will run on 95), I don't think theres any chance of us going to Longhorn until 2010 or so.

I believe another benefit would be decreased memory consumption of the clients. I would have to do some testing to verify this, but I suspect since the overhead of COM is gone, the memory usage *should* be lower in .Net. We currently have problems with the insanely high memory usage of VB6 clients.

I believe MIS is just a Bachelor in Commerce with an MIS major, correct me I'm wrong. Its good if you want a job as a business/systems analyst, wouldn't recommend it if you want to be a programmer. For a programmer I would recommend a CS degree possibly with a minor in Business (which is what I'm doing).

Thanks for the link, it was helpful. But it mostly still leaves me with the same problem. Most of the stuff on that list is either stuff that I can already do in VB6, or its productivity/IDE enhancements. I'll try and go through the list: 1) Deployment - already have a custom VB6 system in place. This would have been nice to have 5 years ago, but now that we have a system in place the benefit is moot. 2) Robust Code - stict type checking is the only concrete benefit from this point. The todo list falls under IDE/productivity enhancement, and structured exception handling is nice, but doesn't allow us to do anything we couldn't do in VB6 (albeit, the syntax is much nicer). 3) Powerful Windows Apps - everything listed here is available in VB6, with the exception of controls docking which I will include in my business case document as its something we've wanted for a long time 4) ADO.Net - this doesn't allow us to do anything we can't do with ADO 2.x, the syntax is just nicer 5) Simplified Component Creation - Nothing we couldn't do in VB6. 6) Enhanced Control Creation - not something we do alot of here, so doesn't really apply. 7) .Net Framework - can do all this stuff in VB6, the framework just provides nicer syntax (ie. better productivity) 8) Integrated Crystal Reports - We already use CR 8.5 with VB6, and we can already design reports in the VB6 IDE via the CR designer, so nothing new 9) ASP.Net - I agree this is a HUGE advancement over the old ASP. But we do all fat-client development here, so this doesn't apply (however I may make a separate business case looking at changing our architecture to ASP.Net instead of fat client). 10) Existing Investments Carry Forward - Not really a benefit, just risk mitigation. So after these 10 points I'm still left with just the following benefits: - Improved IDE and language features leads to increased developer productivity - Control docking allows automatic form resizing - Stict type checking assists in creating more robust code The productivity benefit is probably the most substantial out of those, the other 2 seem minor. I don't think these are enough benefits yet to make a business case for migrating to .Net. More input from anybody would be highly appreciated.

At my work, all our development is done currently in VB6 in a client/server architecture using fat clients and an Oracle database server. I want to put together a business case document, that looks at switching our development to .Net. But I'm having a hard time brainstorming benefits that actually have an impact. About the only one I've come up with so far is increased productivity due to the VS IDE enhancements. The problem is, alot of the issues we have that .Net solves nicely, we've already invested time to come up with a custom solution in VB6. For example, deployment in .Net is much better than VB6, but we already have a custom deployment framework in place, that we spent a considerable amount of time developing, and hence deployment for us is extremely easy using our custom framework (all we do is copy the exe to a fileshare and it automatically propagates to all clients as they need it). So I don't know if I could spin the deployment features of .Net as a benefit to us. Was hoping some of you that have made the switch to .Net could help me brainstorm some benefits that will actually help me persuade management that we should make the switch.

Take a look at the data layer in MS's .Net Pet Store sample application (it can be found on MSDN). That handles abstracting away the data-access code.

I'm not a big C++ user, but as far as I know you will have to do it without a designer. That means using Win32 API's to create your GUI's "oldskool like".

Yes VS.Net uses a new version of MSDN for its help files. And VS.Net is 5 or 6 CD's (somewhere around there anyways). You can get it on DVD also which makes things much easier.

Can you show us the problem code?

I would create a cRoom class, that contains all the data for a certain room. Inside the class have pointers for North/South/East/West that point to the instance of the cRoom class for the rooms connecting to this one.

Since your array changes size dynamically I would try using a System.Collections.ArrayList to store the data. And you could probably do an ArrayList of arrays, or an ArrayList of ArrayLists