Page 1 of 9
What would be needed to go to a full 1.5.0 release?
Posted: 29.09.2007, 17:40
by ajtribick
Leave aside the fact that we'd probably need Chris Laurel to give the go-ahead for a main release for the moment.
Supposing no new features get added to v1.5.0, what is still incomplete in the prerelease version?
How much development work would be needed to make a version that could work as a main release?
(Note I am not demanding that this is done, I am just curious to know what the current status is)
Re: What would be needed to go to a full 1.5.0 release?
Posted: 29.09.2007, 18:32
by t00fri
chaos syndrome wrote:Leave aside the fact that we'd probably need Chris Laurel to give the go-ahead for a main release for the moment.
Supposing no new features get added to v1.5.0, what is still incomplete in the prerelease version?
How much development work would be needed to make a version that could work as a main release?
(Note I am not demanding that this is done, I am just curious to know what the current status is)
Chaos,
this is obviously a GOOD question, yet a difficult one...since the answer depends strongly on the standards one wants to apply.
One could of course restrict the remaining requests to pure bug fixing. Then there would not be all that much left-over work, provided the people involved in doing the fixes are still motivated (!) (extrapolating from my own perspective, I doubt it...).
But on a more ambitious level, there is a number of half-finished approaches, that really would require further development work and dedication...
There are the unsatisfactory stars, the poor atmosphere situation (sky colors are not projected onto the surface), the (incorrect) frame-independence of orbit displays and many more such things.
Bye Fridger
Posted: 29.09.2007, 19:09
by Cham
While better than Celestia 1.4.2, version 1.5.0 is still in a sorry state. I made a list of about 25 bugs and limitations in 1.5.0 :
List of bugs and problems, or things to be adjusted, in random order :
1- The markers are still shown on the foreground. There is no depth sorting of the markers, which is very confusing if there's an object on the foreground (planet, moon, spacecraft, etc...)
2- Moving closer to the sun, from the exterior part of the solar system, is laggy. Moving out doesn't show any lag.
3- "Dust" ring CMODs (points models) don't receive any shadow from their parent planet.
4- Sprites aren't working on the Mac (crashing bug).
5- Cloud shadows aren't working on the Mac (removed feature ?).
6- The demo "D" keyboard shortcut and its menu item should be removed. The demo script file should be placed in the scripts folder. I suggest to assign "N" to the nebulae rendering (instead of "^", which doesn't work on my keyboard), and use "D" for the spacecraft instead.
7- The asteroids and comets labels need some alpha effect, varying with distance and object size, as for the stars and galaxies.
8- The atmosphere code isn't working with multiple sources of light (vivid red atmosphere).
9- Black "holes" and "stripes" on the sky, while the observer is very close to a planet surface (with an atmosphere).
10- Stars rendering is extremelly dull and unrealistic without an atmosphere to smooth out the edge. And there's a single texture for all classes, which makes all stars too similar and boring.
11- I'm still experiencing moving "holes" on background orbital paths, on a Mac, if there's a model on the foreground (asteroid or spacecraft).
12- The comet tail is still very crude (poor paraboloid fixed resolution). Its mesh resolution should be variable (adaptable) with distance, like it is for the planets (celestia's spheres).
13- The "spacecraft" labels aren't depth sorted if the object is close to a planet/moon. They should be hidden behind their parent planet or moon.
14- The star radius precision shown in the upper-left corner is ridiculous. It should be rounded. Currently, the precision shown is not credible.
15- The star radius given in the case of cool stars (red dwarfs and red giants) is inaccurate and unreliable. In many cases, Celestia gives very wrong values (example : according to Celestia, Proxima Cen is much smaller than Jupiter).
16- There's a bug which crashes Celestia when many close stars are each having a planetary object. Having 17 stars (or more) orbiting the same barycenter, each one with a planetary object, crashes Celestia. (limitation in the code)
17- Adding an URL to a star doesn't work (hard coded SIMBAD link), like it is for the planets and other objects. Adding an URL to a star STC definition should overide the default URL.
18- Too much stars in Celestia's database have an "?" as spectral class assignement. This may not be considered as a "bug" however, since it's related to the Hipparcos table and its limitations.
19- There are many double stars with the same name and doubles with different names (Fridger's binaries), and stars with inconsistent greek names (KHI, CHI, etc).
20- Celestia is only reading the first letter of the spectral class, which makes the scripted selection extremely limited.
21- We urgently need an "Clickable false" option at the SSC level, to make objects unclickable, while still visible.
22- The celestial grid is very crude and doesn't show properly all the coordinates labels (bug already identified in the code by me and acknowledged by Chris, who never corrected it, despite the fact that it's extremelly easy to do).
23- some memory leaks about textures and models loading, which may crash Celestia after a long use (many textures and models loaded into memory, without cleaning memory after a while).
24- Pluto/Charon's barycenter's orbit should be visible as an "asteroid".
I'm probably forgetting one or two more bugs I've identified.
Posted: 29.09.2007, 20:03
by Fenerit
Great work, Cham.
As "bugs sticky" seem to lack a list of bugs fixed to be update more systematically from time to time, so the user can browse in better way. Your list can be take up as proof for what would worked out, in my humble opinion.
Posted: 29.09.2007, 20:36
by selden
Quite some time ago Chris asked that all bugs be reported using SourceForge, so that there's just one official list.
https://sourceforge.net/tracker/?atid=1 ... unc=browse
There's a long list of bugs there.
Posted: 29.09.2007, 20:47
by t00fri
Except that the person who has asked for that buglist has VANISHED since exactly
three months. So what is that list worth, a "naive" user might ask?
----------------------------------
Most probably, all these bugs have long been corrected in some unknown archive of NASA or ESA for which Chris is presumably doing
paid full-time contract work since July 1st
----------------------------------
F.
Posted: 29.09.2007, 21:19
by Cham
t00fri wrote:Most probably, all these bugs have long been corrected in some unknown archive of NASA or ESA for which Chris is presumably doing paid full-time contract work since July 1st
If true, then it's a shame.
Posted: 29.09.2007, 21:25
by t00fri
Cham wrote:t00fri wrote:Most probably, all these bugs have long been corrected in some unknown archive of NASA or ESA for which Chris is presumably doing paid full-time contract work since July 1st
If true, then it's a shame.
Martin,
I don't think it's a shame
per se: Chris L. has the copyright for Celestia and so he can do with it plenty of things...
---------------------------------------------------
What is a shame, is that he does NOT care to inform the rest of us here, so as to allow everyone to draw his clearcut consequences!
---------------------------------------------------
Bye Fridger
Posted: 29.09.2007, 21:32
by Cham
t00fri wrote:I don't think it's a shame per se: Chris L. has the copyright for Celestia and so he can do with it plenty of things...
---------------------------------------------------
What is a shame, is that he does NOT care to inform the rest of us here, so as to allow everyone to draw his clearcut consequences!
---------------------------------------------------
Yeah, but the code is actually the creation of several developpers, not just Chris alone. And leaving the others in the cold like this is really savage, IMO.
Posted: 29.09.2007, 21:56
by t00fri
Cham wrote:t00fri wrote:I don't think it's a shame per se: Chris L. has the copyright for Celestia and so he can do with it plenty of things...
---------------------------------------------------
What is a shame, is that he does NOT care to inform the rest of us here, so as to allow everyone to draw his clearcut consequences!
---------------------------------------------------
Yeah, but the code is actually the creation of several developpers, not just Chris alone. And leaving the others in the cold like this is really savage, IMO.
You don't have to tell me
Bye Fridger
Posted: 29.09.2007, 22:20
by ElChristou
Guys, more time pass and more I find we are quite pathetic to spend our time on the wailing wall...
Chris L. just do what he want and finally if we are not agree with that better leave once for all because even if he comes back, he can disappear again at any time without a word. If you think it's too much for you to live with this attitude, again, better leave.
If tomorrow he want to rebuild the community for a really serious work, he will easily find the mails of the persons who represent a certain interest for his project...
Posted: 29.09.2007, 22:29
by Cham
ElChristou wrote:Guys, more time pass and more I find we are quite pathetic to spend our time on the wailing wall...
I agree.
ElChristou wrote:If you think it's too much for you to live with this attitude, again, better leave.
Well, this is exactly what I'm about to do. As I told Fridger by private email few weeks ago, I'm still giving Chris up to the end of this year before I leave, so I could build some very nice volumetric nebulae with Mathematica (I already have several prototypes, working on a PC at work). If I don't see any improvements in development, in the following months, I'll quit.
Posted: 29.09.2007, 22:52
by ElChristou
Cham wrote:ElChristou wrote:If you think it's too much for you to live with this attitude, again, better leave.
Well, this is exactly what I'm about to do. As I told Fridger by private email few weeks ago, I'm still giving Chris up to the end of this year before I leave, so I could build some very nice volumetric nebulae with Mathematica (I already have several prototypes, working on a PC at work). If I don't see any improvements in development, in the following months, I'll quit.
Perso I go till the end of the Apollo addon because of the community and principally because of Andrea and Linuxm@n, then if nothing change I suppose I will pass from active to very occasional member... (As I will continue to use Celestia, I suppose I won't be able to quit definitively the community right now...)
Posted: 29.09.2007, 22:54
by selden
I have to admit that I really don't understand all this angst.
Celestia is a software tool. It could be considered to be in the same category as a compiler or a text editor, except that it's for realtime 3D. It need not be considered an end-in-itself.
If it's not the tool best suited to the projects that you want to work on, then a more appropriate tool should be used. (And I, for one, would enjoy learning about those other tools. The few that I know about are listed on my Celestia Resources page.)
Many people prefer that their software tools be stable and that those tools not be changing frequently. Most scientists get really upset when compiler upgrades cause their analysis programs to produce different results. Administrative staff lose productivity when their office software changes and they need to learn new commands and procedures.
While I enjoy using Celestia in various ways, and one of the things I like to do is find bugs in software, Celestia's current stability is giving me a chance to work on other projects (some related to Celestia, some not) that I might not otherwise have time for.
Posted: 29.09.2007, 23:07
by Cham
Selden,
the problem is that Celestia 1.5.0 is currently full of "small" bugs and limitations, that makes it very unreliable and frustrating (see the list I published above). I agree that I don't want to see constant and frequent upgrades : I prefer to have a stable and reliable tool. Currently, Celestia contains too many sources of frustration and, AFAIK, there is no other sofwtare in the same category (except maybe MitakaPlus and OpenUniverse).
Posted: 29.09.2007, 23:33
by ElChristou
selden wrote:I have to admit that I really don't understand all this angst...
I think all depend on what each one is waiting and on what each one do with the soft. Perso for example I will be satisfied enough by a certain version the day HDR will be implemented (I cannot stand anymore seeing stars on the moon!
), and secondarily the day self shadows will be cast on models...
Posted: 29.09.2007, 23:39
by Cham
ElChristou wrote:I cannot stand anymore seeing stars on the moon!
Can you elaborate on this ? Stars on the moon ? I don't understand what this is.
Posted: 30.09.2007, 01:07
by Hungry4info
I think from the surface of the moon, the light coming from the lunar surface is bright enough to make the stars in the 'sky' invisible. The lunar light saturates your eyes or something to that effect.
I suggest giving the moon a thin black atmosphere in the mean time.
Posted: 30.09.2007, 01:23
by ElChristou
Cham wrote:ElChristou wrote:I cannot stand anymore seeing stars on the moon!
Can you elaborate on this ? Stars on the moon ? I don't understand what this is.
Yep, Hungry got it. No stars visible on the moon surface... (on the day side of course...)
Posted: 30.09.2007, 01:34
by Cham
Ok.
Personally, I don't mind much about the stars visible while standing on the moon. I prefer to see the bugs solved and the current limitations removed (according to my list above).
Of course, dropped shadows on models could be great too, but I wont count on that (especiallly considering the effect on the frame rate).