Re: [Stellarium-pubdevel] Observations after the GL branch merge

2012-09-10 Thread Thomas Morris
For the record I don't see performance drop at high zoom (down to 10E-4 deg fov) on either laptop (Intel 945GM) or desktop (ATI HD 6670). Thomas On 10 September 2012 16:14, Ferdinand Majerech wrote: > --safe-mode sets GL1 Qt paint engine. > The GL2 backend reqiures the GL2 paint engine, so it fa

Re: [Stellarium-pubdevel] Observations after the GL branch merge

2012-09-10 Thread Ferdinand Majerech
--safe-mode sets GL1 Qt paint engine. The GL2 backend reqiures the GL2 paint engine, so it fails to initialize and falls back to GL1. On 9/10/12, Bogdan Marinov wrote: > On Mon, Sep 10, 2012 at 2:39 PM, Ferdinand Majerech > wrote: >> [snip] >> Matthew, Bogdan: >> I can't reproduce this - but I t

Re: [Stellarium-pubdevel] Observations after the GL branch merge

2012-09-10 Thread Bogdan Marinov
On Mon, Sep 10, 2012 at 2:39 PM, Ferdinand Majerech wrote: > [snip] > Matthew, Bogdan: > I can't reproduce this - but I think it might be caused by fragment > shaders being too expensive. Could you try if --safe-mode (GL1) > changes anything? No, no change here. Starting it with --safe-mode resul

Re: [Stellarium-pubdevel] Observations after the GL branch merge

2012-09-10 Thread Ferdinand Majerech
Barry: Sorry, I don't mean to be too demanding. I have only one Windows box for testing here. But if you have a bug on any of these machines, at least saving the log (which now prints info about GPU/driver) would be useful. If there are checkered textures permanently, that's definitely a bug. Matt

Re: [Stellarium-pubdevel] Observations after the GL branch merge

2012-09-10 Thread Bogdan Marinov
On Sun, Sep 9, 2012 at 8:20 PM, Matthew Gates wrote: > [snip] > 3a. When zoomed in right on a texture so it fills a good portion of > the screen (e.g. zooming so the moon fills the FOV), the frame rate > drops off very badly to a few FPS, and even less than 1 fps at very > high zoom. Anyone else

Re: [Stellarium-pubdevel] Observations after the GL branch merge

2012-09-09 Thread Matthew Gates
On 9 September 2012 18:38, Ferdinand Majerech wrote: > 3: > Still working on this. We seem to get better FPS on some GPU/drivers, > worse on others. Particularly bad seem to be Atoms with integrated GPU. > > What are you using (CPU/GPU/OS/driver)? Kubuntu 12.04; kernel 3.2.0-29-generic-pae Accord

Re: [Stellarium-pubdevel] Observations after the GL branch merge

2012-09-09 Thread Barry Gerdes
:29 +0200 > From: kiithsa...@gmail.com > To: stellarium-pubdevel@lists.sourceforge.net > Subject: Re: [Stellarium-pubdevel] Observations after the GL branch merge > > The checkerboard is a placeholder used in case a texture failed to load > or something was unsupported. It is al

Re: [Stellarium-pubdevel] Observations after the GL branch merge

2012-09-09 Thread Ferdinand Majerech
The checkerboard is a placeholder used in case a texture failed to load or something was unsupported. It is also used while the texture is loading. Previously, with planets, if texture didn't load / was still loading the planet simply wouldn't be drawn, I thought this would be better as it is more

Re: [Stellarium-pubdevel] Observations after the GL branch merge

2012-09-09 Thread Barry Gerdes
Hi Matthew I missed the zoom problem when I zoom in on solar system objects I don't have any fps problems. In fact it is slightly faster. However the "texture" becomes a checker board (Yellow/black). This does not happen on all my platforms and is intermittent on some. It has to be with i

Re: [Stellarium-pubdevel] Observations after the GL branch merge

2012-09-09 Thread Barry Gerdes
Hi Matthew I get all the same problems on some computers and not others so the problems are platform based. At present I test on two Win 7 platforms and six XP platforms (it takes about 6 hours to do clean builds) I also build in Linux on 3 platforms I addition I also have virtuals on the Win 7

Re: [Stellarium-pubdevel] Observations after the GL branch merge

2012-09-09 Thread Ferdinand Majerech
2: Several people have reported this problem but I couldn't reproduce it yet. Not sure what is the cause. 3: Still working on this. We seem to get better FPS on some GPU/drivers, worse on others. Particularly bad seem to be Atoms with integrated GPU. What are you using (CPU/GPU/OS/driver)? 3a: N