Re: [Flightgear-devel] Broken Clouds

2007-07-26 Thread Melchior FRANZ
* Hans Fugal -- Thursday 26 July 2007: So we likely are looking at a memory corruption bug in OSG Yes. That looks very much like an OSG bug. And valgrind agrees. The textures are perfectly valid. BTW: according to my checker we don't have a single corrupted SGI image file in CVS. m.

Re: [Flightgear-devel] Broken Clouds

2007-07-26 Thread Melchior FRANZ
* Melchior FRANZ -- Thursday 26 July 2007: Yes. That looks very much like an OSG bug. And valgrind agrees. I have to take this back. The code is rather ugly -- it reads more data than are available, and valgrind complains about that. But after that the line decoder breaks correctly. I don't see

Re: [Flightgear-devel] Broken Clouds

2007-07-26 Thread Curtis Olson
On 7/26/07, Vivian Meazza wrote: Incidentally, I have observed in both linux and osx in fg/osg that cloud layers don't work quite like I'd expect. If the weather is scattered clouds at 3000 ft and overcast at 5000, then when I'm sitting on the runway I see blue sky between the scattered

Re: [Flightgear-devel] Broken Clouds

2007-07-26 Thread Vivian Meazza
Hans Fugal Sent: 26 July 2007 14:57 To: FlightGear developers discussions Subject: Re: [Flightgear-devel] Broken Clouds On 7/26/07, Vivian Meazza [EMAIL PROTECTED] wrote: Hans Fugal I have never seen or heard of any problem with osg and clouds of any sort. Xnview reports

Re: [Flightgear-devel] Broken Clouds

2007-07-26 Thread Hans Fugal
On 7/26/07, Vivian Meazza [EMAIL PROTECTED] wrote: Hans Fugal I have never seen or heard of any problem with osg and clouds of any sort. Xnview reports no errors in cirrus.rgba, or with any of the textures. Perhaps you have corrupted the clouds textures locally? Did you open them in any

Re: [Flightgear-devel] Broken Clouds

2007-07-26 Thread Vivian Meazza
Hans Fugal Sent: 26 July 2007 01:14 To: FlightGear developers discussions Subject: [Flightgear-devel] Broken Clouds Textures/Sky/broken.rgba seems to be literally just that, broken. Or rather, corrupted. As do scattered, few, and cirrus. fg/osg (but not plib) frequently crashes

Re: [Flightgear-devel] Broken Clouds

2007-07-26 Thread Norman Vine
Curtis Olson writes: On 7/26/07, Vivian Meazza wrote: Incidentally, I have observed in both linux and osx in fg/osg that cloud layers don't work quite like I'd expect. If the weather is scattered clouds at 3000 ft and overcast at 5000, then when I'm sitting on

Re: [Flightgear-devel] Broken Clouds

2007-07-25 Thread Hans Fugal
Confirmed that osg seems to understand them, try it: osgviewer --image cirrus.rgba So the format is ok, the question is where's the bug? It would certainly seem that it lies in the difference between these two variants of SGI images. It's probably within OSG itself. The other question is, what

[Flightgear-devel] Broken Clouds

2007-07-25 Thread Hans Fugal
Textures/Sky/broken.rgba seems to be literally just that, broken. Or rather, corrupted. As do scattered, few, and cirrus. fg/osg (but not plib) frequently crashes and I've traced it down to the cloud cover (clouds == likely crash, or crazy sky as in

Re: [Flightgear-devel] Broken Clouds

2007-07-25 Thread Csaba Halász
On 7/26/07, Hans Fugal [EMAIL PROTECTED] wrote: ImageMagick does not think these corrupted (identify(1) should return nonzero if it encounters a corrupted file), but some of them are obviously not being interpreted correctly by ImageMagick. Try display cirrus.rgba to see what I mean.