Vedexent

I can empathise. From reading the thread though, do I understand correctly the weather "injection" works on multiplayer mission files using the current binary, but that finding an appropriate trigger inside DCS at load time is a problem, and that such an injection alters the main file, which is considered undesireable? If so, I think it might be possible for me to write a quick Python script, scheduled to run hourly (or more often), which would copy a "master copy" of a mission file into the Multiplayer directory and run your weather injection binary on the new copy, whenever the file access time exceeded the file creation time for the copy in the multiplayer directory - that is, if the file had been loaded into the server. It's a "band-aid" solution, but it might work on our server as an interim solution.

Hey Falcon60 - good to see you and Tycoon74 out there tonight.

Awesome idea, just what we were looking for, but ... what's the status on the MP support?

Just to let people know: The Mission server is up as close to 24/7 as I can make it: listed under 314th in the public server list. The teamspeak server is here: ts32.gameservers.com:9105 - you are encouraged to use it. The website is here: http://314thsquadron.enjin.com/ - lots listed here, including the server rules. Server is running Serveman 3 - so feel free to select (or at least vote on) any one of a number of available missions. Most missions have the Su-25 as the central aircraft, but the Caucasus mission in the list is true multi-aircraft mission, with pretty much everything but the Mustang available. Everyone is welcome.

This happens to me occasionally, and it has to do with Windows permissions. The reasons are due to to some re-installation and changing directories, but it turned out I didn't have full write permissions to the directory that all the default training missions were in. It behaved in exactly the manner you're describing. You might check to make sure you have full read/write/delete permissions for the directory, and - if this isn't the case - add those permissions, or change the missions directory to one you create while not in Administrator mode.

Fighting Lemmings! Impressive work! :D

I haven't seen a debriefing screen - but it does save a track file on your system. Perhaps if you play the track back it would show a debrief screen?

My apologies to anyone flying - or trying to fly - on the server last night. Apparently Windows restarted after I left for the evening and took everything down.

I fly all the time. I never see you. Maybe you don't fly ... or maybe you just don't fly on my server, hmm?

That must be why Xplane is used as a professional small plane pilot training simulator :P XPlane lacks the integrated combat simulation, and it's 3D cockpits are not nearly as good as the A-10C or the Mustang. But XPlane also has a far broader selection of aircraft. If all I wanted to do was fly airplanes, by myself, XPlane might be a better platform. However, there really is only one high fidelity multi-user, multi-role combat aircraft platform out there. Which really is a shame. As much as I like DCS and ED, having a little competition keeps you motivated to stay on top of your game.

Unless it was accidental, I don't think it was me ... If so, I apologize. That was not my intent.

We did not; we were just kidding.

There are ways to block this without blocking VMs, but it requires a little infrastructure on the part of the game company. Your installation is required that it "call home" periodically, and check that the software key is valid. If the authentication server starts seeing the same key coming from multiple systems, the key is suspended or cancelled, and the cloned systems are locked down until the user contacts the company and explains what's going on, and the company may or may not re-activate the key. That incurs some ongoing expense on behalf of the company, however.

tha Well ... I'd say that's true of many DCS players, yes. But if all I wanted was airframe fidelity, why wouldn't I just use XPlane, which has professional level simulation, and a far wider selection of aircraft? It's the high fidelity combat simulation that is unique to DCS - and given the limitations of AI opponents, that almost automatically implies multiplayer. I concur 100%, however, that the original simulators were not originally designed to be extended this far. DCS World seems to be an organic evolution of the original software far beyond it's original scope, merging several discrete projects under an umbrella wrapper - and I've never seen a software system that has "evolved" into something else not be laden with issues. I don't hold out a whole lot of hope for EDGE upgrade - unless there's is also substantial re-engineering of the basic software framework along with it. Perhaps I'll be pleasantly surprised.

Multiplayer works - or at least it has been working for me and the people I've been flying with. Multiplayer does seem to have issues if you push past certain scenario and map limits, however. Apparently - as previously noted - if your scenarios stay within certain limits things are fairly stable. Yes, we'd like big, complex, integrated operational scenarios which seem possible in the system as it is designed - but it seems there are stability issues with that. I haven't run up against those issues - yet - but we've not tried huge hyper-complex mission scenarios in our group, yet, either.

Thanks! Sadly, for me - I'm a fan of modern aircraft combat, and so far as I know you can't import an Su-25 into a WWII sim :)

Complex and expensive? Yes. Unavoidable? No. As a systems analyst and tester, I know it's not avoidable. It's called continuous integration and regression testing. However, I also know that it's bloody near impossible to retrofit this approach into an existing project. And I recognize that open ended simulations contain testing challenges not found elsewhere. DCS bears all the software hallmarks of a project that has evolved way beyond its original scope (you think?), and as such is a bearcat to keep consistent, so I don't really expect the project team to be able to easily keep things consistent - short of a complete code re-design and re-write (and I don't have millions of dollars in spare change to help fund that).

But what are they migrating to? Apart from Falcon BMS - which arguably is a pretty narrow genre; you don't have multi-role capabilities like you do with DCS - and more arcade like, lower fidelity game/sim hybrids, what else is there? The combat sim market is pretty lean right now - and that might be part of the problem. A stiff competitor might light a fire under ED to clean up the platform. And my question wasn't rhetorical - if there's an up and coming combat flight sim that's drawing in players, I'd love to know what it is.

I'm curious: what settings, parameters, and procedures do you find keep things stable?

Nope - flying the Su-25 and Su-25T on our server these days.

I've had this happen about 50% of the time in multiplayer. All of a sudden I can't hear ATC, although other people on our server can hear the ATC traffic back to me. Seems to be an open bug.

There seems to be more traffic on the weekends. This Sunday our server was pretty much full. There are groups out there - but they aren't as active as one might hope. You can always come check us out - but we're primarily a Russian aircraft group ... not a lot of A10C roles.

Although it's not a question of what hosts a windows VM "better" - it's about what virtualization techniques trigger the apparent DRM. Various virtualization layers have different virtualiztion "clues": signature default values (registry keys, hard disk name, default network card address), dmesg output values (for *nix based guests), subtle non-fatal but known - and thus detectable - implementation bugs, etc. I would guess the DRM in this case is looking for certain signatures, and ignoring others - and Hyper-V must implement one of the known signatures. Edit: If you're wondering if a VM could ever be made "detection proof", check this out - it suggests not. Timing analysis might always be a weakness: http://www.cs.berkeley.edu/~cthompson/papers/virt-detect.pdf

Post reported! You did mean that red button, right? ;)

Ouch; there's another reason to prefer *nix. Well, I don't mind long initial load times, as the server will - hopefully - be up for extended periods of time. In any case, I'll have to experiment with it, and learn to tune things under VMWare. Thanks for putting in the time for experimenting :)