Oliver & Bernd : will there be ( or existing now and I haven't noticed ) an EinsteinGammaAdapter class or somesuch for the Fermi-LAT work, to complement our existing queries of the science application suite?
[ How hard would it be for me to have a crack at it? What would I need to know ? :-) ]
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
I believe that a moderator moved my earlier comments on the screensaver problem to some other thread, but without mentioning which other thread; I still haven't found it.
I did indeed. Sorry, you should have gotten an automated email about this, including the thread where your posts got moved to.
@Mike: thanks for providing the link to robertmiles.
- Earth now rotates as per query of system clock to give UTC time thus angular offset for transform.
- Sun now available, but looks a bit 2D-ish at present so will investigate lighting. Most likely idea to solve is : place a light source at our viewing position pointing toward the Sun, which OpenGL ought give a limb darkening effect on it's reflection ( base problem is how do you 'shade' a ball of gas which is supposed to be a light source? ). Good news is I have found a free ( as in beer + copyright OK for this usage ) good texture with nice 'flame' and 'sunspot' features ! :-)
- orbital model for Sun, I gave up and called it a circle. Place it according to UTC day within the year. Problem with an exaggerated ellipse ( actual orbital eccentricity is 0.0167 ie. stuff all ) is that the Sun won't be at the right places at the right times. Speeds up when closest to Earth and slows down when furthest away, so if implemented as such the passage along the ecliptic would be way off. Good news is that now I've got the Sun's position then during Earth rendering I can put a light source where the Sun is : so Earth can have a dark side and a lit side ! :-)
- now that I'm relatively a whiz with textures all sorts of 'advertising' ideas come to mind. I can finally use my Einstein-With-Tongue-Poking-Out bitmap ... I might paste it onto the orbital plane b/w Sun and Earth. :-)
- planning a new HUD class, along the lines of a Java style layout manager with sub-panels ( North/South/East/West/Center ie. BorderLayout ) which will make re-sizing to all manner of window dimensions, font choices ( font-height especially ) and screen resolutions a whiz. Should make testing of various styles not too painful. The panels will contain either image areas ( for say logos or FFT transforms ) and/or text areas, with text areas either fixed or scrollable horizontally ( ie. one line ticker-tape marquee ) or scrollable vertically*. The render callback routine will do the 3D scene then follow with the HUD overlay, per frame.
- found out that text rendering via OGLFT can include creating texture-object based strings ( ie. loaded once to the video card and then kept for redraws ). With an alpha channel for transparency too. This is terrific as it will give a massive efficiency boost compared with the display-list based alternatives.
[aside] One idea is that if an object of interest appears in the center view ( say the Colvin's pulsar discovery ) then that will activate some highlighting in the simulation and a scrolling story in the HUD about that. One would need to partition the celestial sphere appropriately/efficiently .... needs work. [/aside]
Cheers, Mike.
* NOT like a StarWars intro with 3D perspective fade into the distance .... only 2D .... though I did toy with that idea :-)
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Sounds good! Keep it up. I'll have a look when the OpenCL apps are out of the door...
Quote:
- found out that text rendering via OGLFT can include creating texture-object based strings ( ie. loaded once to the video card and then kept for redraws ). With an alpha channel for transparency too. This is terrific as it will give a massive efficiency boost compared with the display-list based alternatives.
Oliver & Bernd : will there be ( or existing now and I haven't noticed ) an EinsteinGammaAdapter class or somesuch for the Fermi-LAT work, to complement our existing queries of the science application suite?
Bernd wrote:
Quote:
For the FGRP search we are using the graphics application of the GW search. It is not planned to change that.
I mean, this is just because lack of time on our side. If you want to make something Fermi-specific, your are very much welcome to add code, and we'll publish the graphics binary then.
Try here Cheers. Mike.
)
Try here
Cheers. Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Oliver & Bernd : will there
)
Oliver & Bernd : will there be ( or existing now and I haven't noticed ) an EinsteinGammaAdapter class or somesuch for the Fermi-LAT work, to complement our existing queries of the science application suite?
[ How hard would it be for me to have a crack at it? What would I need to know ? :-) ]
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
RE: Try here Cheers.
)
Yes, that's where it is now.
For the FGRP search we are
)
For the FGRP search we are using the graphics application of the GW search. It is not planned to change that.
BM
BM
Fine by me, I won't have to
)
Fine by me, I won't have to worry then. :-)
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
RE: I believe that a
)
I did indeed. Sorry, you should have gotten an automated email about this, including the thread where your posts got moved to.
@Mike: thanks for providing the link to robertmiles.
Best,
Oliver
Einstein@Home Project
Progress report : - Earth
)
Progress report :
- Earth now rotates as per query of system clock to give UTC time thus angular offset for transform.
- Sun now available, but looks a bit 2D-ish at present so will investigate lighting. Most likely idea to solve is : place a light source at our viewing position pointing toward the Sun, which OpenGL ought give a limb darkening effect on it's reflection ( base problem is how do you 'shade' a ball of gas which is supposed to be a light source? ). Good news is I have found a free ( as in beer + copyright OK for this usage ) good texture with nice 'flame' and 'sunspot' features ! :-)
- orbital model for Sun, I gave up and called it a circle. Place it according to UTC day within the year. Problem with an exaggerated ellipse ( actual orbital eccentricity is 0.0167 ie. stuff all ) is that the Sun won't be at the right places at the right times. Speeds up when closest to Earth and slows down when furthest away, so if implemented as such the passage along the ecliptic would be way off. Good news is that now I've got the Sun's position then during Earth rendering I can put a light source where the Sun is : so Earth can have a dark side and a lit side ! :-)
- now that I'm relatively a whiz with textures all sorts of 'advertising' ideas come to mind. I can finally use my Einstein-With-Tongue-Poking-Out bitmap ... I might paste it onto the orbital plane b/w Sun and Earth. :-)
- planning a new HUD class, along the lines of a Java style layout manager with sub-panels ( North/South/East/West/Center ie. BorderLayout ) which will make re-sizing to all manner of window dimensions, font choices ( font-height especially ) and screen resolutions a whiz. Should make testing of various styles not too painful. The panels will contain either image areas ( for say logos or FFT transforms ) and/or text areas, with text areas either fixed or scrollable horizontally ( ie. one line ticker-tape marquee ) or scrollable vertically*. The render callback routine will do the 3D scene then follow with the HUD overlay, per frame.
- found out that text rendering via OGLFT can include creating texture-object based strings ( ie. loaded once to the video card and then kept for redraws ). With an alpha channel for transparency too. This is terrific as it will give a massive efficiency boost compared with the display-list based alternatives.
[aside] One idea is that if an object of interest appears in the center view ( say the Colvin's pulsar discovery ) then that will activate some highlighting in the simulation and a scrolling story in the HUD about that. One would need to partition the celestial sphere appropriately/efficiently .... needs work. [/aside]
Cheers, Mike.
* NOT like a StarWars intro with 3D perspective fade into the distance .... only 2D .... though I did toy with that idea :-)
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
The only graphics I see on my
)
The only graphics I see on my Linux box is that of the binary radio pulsars search, both in Einstein@home and Albert@home. OK all the same.
Tullio
RE: Progress report
)
Sounds good! Keep it up. I'll have a look when the OpenCL apps are out of the door...
Told ya you'll like it ;-)
Oliver
Einstein@Home Project
Mike
)
Mike wrote:
Bernd wrote:
I mean, this is just because lack of time on our side. If you want to make something Fermi-specific, your are very much welcome to add code, and we'll publish the graphics binary then.
BM
BM