New Netgame Launcher - Version 0.50 Available

SDK

The new version of MSERV.DLL is now ready. I haven't finished documenting it yet, but an SDK can be downloaded here, at the Launcher's page. There's a Visual Basic wrapper included which should be fairly straightforward to use, so the current dearth of documentation shouldn't be much of a problem. I'll post again when the docs are finished, though.

EDIT: Having just read your STH: LOTBB2 page, Hotdog, (and I look forward to seeing it at SAGE), I've remembered you're using VB4. The VB code I've written is designed for VB6, but I don't think I've done anything too outlandish, so the necessary modification should be minimal. I hope.

Aw, that's alright. I'll just have to redraw the forms, but the code should work.
Thanks, Oogland!

EDIT: It doesn't work. There's a problem in your wrapper, there's a run-time error 6: Overflow. I'll try to fix it on my own.

EDIT 2: I got it working. Changing the line
DIM MSsList(512) As serverlist
to
DIM MSsList(0) As serverlist
fixed the problem. Huh.

EDIT #3: I need to know one thing: Does your function return -1 if it can't connect, 0 if there are no games running and 1 if there are? Or, does it return the number of games running, -1 if it can't connect?

It returns a negative number if it can't connect: -2, I think, but that doesn't matter; negative return signals error. Otherwise, it returns the number of servers registered.

That overflow error could either be due to my lack of sufficiently thorough testing (I made a few last-minute changes to the array stuff; seemed okay, but maybe not) or possibly a difference in VB4's handling of LBound and UBound. Or summat. I rejigged it to handle any size of array, instead of the stupid 16-element hardwiring it used to use.

I'll try to take another look at it tomorr... today. Sorry for the lack of coherence: it's late, as I've already indicated. :-)

EDIT: I can't reproduce the problem, but now that I've actually looked up the error code (rather than taking it to mean "Subscript out of bounds", which was silly of me), it seems that it could be to do with the size of integer constants, or possibly string lengths. There was a definite bug in the code with respect to the latter, which has now been fixed, in v1.1.2. A possible solution to integer overflow - if that is indeed the problem - has also been added.

Thanks, but I think that I got it to work.

Okay. The other bug fix is still necessary, though. Changing all instances of the number 75 to 74 in QueryMasterServer should sort things out; otherwise, you'll get strange things happening when more than one server is registered.