sundog

My previous message wasn't directed at you, Andy. That said I think the current metar.rwx file that is coming in from X-Plane for this hour contains some invalid data that is causing a crash. I'm looking into it.

Looking into it. I think there is some invalid METAR data in the current weather data that is causing trouble.

Please try updating or re-installing your aircraft; it's throwing a lot of errors in your log. And temporarily disable X-Life. See if that helps. That's really just the tip of the iceberg; I see other errors associated with other plugins as well in your log, as well as a bunch of conflicting custom scenery. I think SMP is crashing because one of these other plugin errors corrupted memory, or your custom scenery caused X-Plane to run out of resources when we were trying to create new clouds.

You're free to go back to SMP 3.1.2 if you wish, but I can assure you that 3.2 uses fewer system resources given the same settings, not more.

I really think you're low on video memory, Andy. I think experimenting with your usage of custom scenery and third party aircraft will make a difference. Reducing the cloud draw area setting in SMP may also help, but as I recall you already have that set fairly low. Ultimately you'll have to make some trade-offs, or upgrade your system I'm afraid.

Whatever changed your weather mode sounds like the culprit. If you post your log.txt file we might be able to spot what's going on. SMP and RWC don't attempt to change X-Plane's weather mode in any way, so I suspect there is some other weather-related plugin in the mix doing this.

I agree the main thing that makes CB look odd is when there is not a bunch of cumulus congestus clouds surrounding them, as there would be in real life. We are just doing what the METAR data tells us, though. It's strangely common to see "FEW" cumulus clouds reported along with a "CB" modifier - and if we take that literally, it does end up looking a little odd. I am considering forcing SCT when CB is seen. It's not true to the METAR data, but would look more natural. Open to thoughts on this. As for why the CB clouds extend to the ground - the idea is to capture the reduced visibility from the rain shaft underneath the cloud which does reach the ground. We've experimented with other approaches for representing virga, but have had trouble getting it to look good up close or when you're directly underneath the CB cloud. We've since cracked the code on forcing our own visibility conditions on X-Plane within localized areas though, so that is something I hope to improve upon in future updates.

METAR data does not include information on cirrus clouds. So with RWC, yes you need to "force" cirrus.

If you turn up the visibility in your weather settings, I think you'll see clouds drawn out farther with SMP in your test above. Keep in mind too that the very distant clouds in the default engine are just a static ring around you. With SMP you can see individual clouds come into view at the edge of the visible area, and fly toward them the whole way. We really try to minimize the use of "smoke and mirrors".

Because a cirrus cloud texture that visibly tiles as you move would look even worse. Usually they are high enough above you that it's not noticeable in normal flight, so it's just a conscious tradeoff for better visual quality. If you find this distracting, try experimenting with the different cirrus texture sets we offer in the Skymaxx Pro configuration screen. Some are wispier than others, and contain fewer clouds near the plane, which will make this much less noticeable.

What's happening above is that SkyMaxx Pro is seeing a request to create cumulus clouds at extremely high altitudes, and assuming that data was sent in error. Cumulus clouds do not appear at 30,000 feet in nature. It automatically converts any requests for cumulus clouds above 7,000 meters (about 22,000 feet) to cirrus or cirrocumulus clouds instead, in order to maintain realism. It's not a bug, it's a feature. (Seriously!)

Please send us your log.txt file after experiencing this crash. There are a lot of people using RWC with NOAA successfully, so I suspect you have some sort of other plugin conflict at play, or an out of date plugin that has bug fixes available for it. The log may tell us more. NOAA and XSB don't actually interact directly with RWC at all.

Even 2GB can be insufficient if you're running a complex aircraft and/or custom scenery. Can you try flying with a standard plane and standard scenery as a test, and see if that helps?

Not quite sure I understand. Are you saying you used the X-Plane feature to draw your own weather? RWC won't sync up with that mode, I'm afraid - it's not exposed to us in any way.

It comes down to the details of how FSGRW deals with networked setups like yours, which I don't really know. What I can say is that if FSGRW appears to be updating the METAR.rwx file on both of your PC's (are they both the same size with the same modification time?) then you should be in good shape; just set RWC to "automatic" and it should pick up the weather that FSGRW is broadcasting. My understanding is that FSGRW communicates with X-Plane to set the settings it needs automatically while it is active. If anyone actually has this sort of setup themselves, please chime in!

Well, when you hit "reset" it basically clears out all of the clouds and recreates them - which will have the effect of un-fragmenting your video memory in the process. I'm afraid it's more evidence that your video card just doesn't have enough memory on it to handle what you're asking it to do.

It may have been coincidence that weather conditions got bad enough in the real world to trigger performance issues after you installed the update. Or, perhaps you adjusted some settings after upgrading and that's what affected performance, not the upgrade itself. I have heard legends of mysterious performance problems resolving themselves upon re-installing the same version of SMP, however. I can't think of a logical explanation of why that would work, but may be worth a shot.

Andy, I looked more deeply into your setup to see what might be going on. First, I loaded up the metar.rwx file you provided and flew around Detroit. The weather was pretty crappy as you described, but I didn't see any sort of issues. Then I took a closer look at your system specs, and realized you're using a graphics card that only has 1GB of video memory. I think that explains a lot, and video memory usage is something I don't think your Apple rep looked at. I think you should try my original advice - temporarily remove custom scenery, and fly with a default plane, and see if that clears it up. I think the weather in this region just pushed your system over the edge because you ran low on memory on your video card. Ultimately you'll probably need to make some trade-offs by uninstalling custom scenery, reducing your render settings further, or flying with a less demanding plane. I don't think the SMP 3.2 update has anything to do with it. You could try reverting to 3.1.2 as a test, but I would expect that to make things worse, not better.

No ETA yet as it's not entirely under my control when it goes out. But I'm hoping for very soon. Just checking the boards this morning to see if there are any other issues I need to address with it before releasing it to distribution.

I don't see anything obviously wrong with your file structure there. It's RAM you're running out of, not hard drive space. Try flying with a default plane in that same area; do you still have problems? If you have custom scenery installed for the area you are flying in, does temporarily removing it help? Are there other applications running in the background that you can quit? Looks like you've got a few things running there.

It is normal. The sun is actually below the horizon there (your DSF's just dont extend far enough to obscure it), and unless there's a moon out, clouds will look very dark in that situation in real life.

According to the SMP config screen you posted, you're almost out of system memory despite the fact that your system has 16GB. Do you have some particularly demanding custom scenery installed or something? If memory is swapping, it would definitely explain what you're describing.

SkyMaxx Pro logging an updated plane location does not mean it was responsible for your crash. There are a bunch of errors in your log associated with your custom scenery, and many other plugins that could be causing trouble. Unfortunately your log is inconclusive as to what actually caused the problem. It could be that you're running out of memory (what does the memory usage in the SMP config screen say?) or it could be a bad plugin. Only way to really tell is to go back to a standard installation, and add stuff back in until you have trouble.

My understanding is that your download counter gets reset when new updates are released.

I just meant that switching to sparse or dense particles is the only way to avoid the problem you're seeing until the 3.2.1 update comes out. Not quite ready to talk about what's coming up in SMP 4 yet!