Full-time on Celestia!

[selden]

s3nn0c,

Unfortunately, Celestia's code developers simply don't have very much free time. I think we all agree that we would prefer that Chris spend his time on improving Celestia's functionality, not on Addons.

An example of an existing coherent set of Addons is Frank's Educational Activities. It probably would be appropriate for an enthusiastic user to put together another set that they think is appropriate; with help from others, of course.

Unfortunately, past attempts to do something similar never seem to have been finished. It's not an easy project.

[t00fri]

Great news, Chris, I hope Celestia fans will see the difference soon and I hope you will have a lot of fun - and not only fun - while making Celestia better.

I dont want to turn this thread into the next "what to do first" song, but your words, fsgregs, are imo very important. Celestia is a great core but it needs a lot of effort to configure and to shine. Thousands of textures, many extensions... The real power and the real beauty of Celestia is hidden, many people just use the default configuration. IMO the most important thing is not a new method of visualization of galaxies etc, but a simple, coherent system of extensions and two easy-to-download extension packs.

Developers should choose the best textures and the best extensions and should make them available as two separate packs, both easy to download and install. This is relatively easy to do and will make Celestia much more interesting for many people. Sure, we all want to see many improvements in the core, but first we should make the real power of Celestia more visible to the others.

I disagree with most of the priorities you are trying to advocate. The reasons are mostly simple:

-- The perspective by a teacher of young students is clearly different from that of many other users. It is crucial that Celestia offers a BALANCED and highly accurate framework for scientific visualization of the Universe. There are quite a number of important applications BEYOND mere educational ones.

-- For menpower reasons, add-on development is just too much for the devs.

Add-ons are also problematic, since many are developed by users with little specific knowledge. This leads to bad quality and frequent incompatibilities among our various supported operating systems as well as among different Celestia versions. Together with the total lack of quality control at the Motherlode, the result is highly unsatisfactory. At Celestial Matters three of us with great graphical experience are proceeding slowly to provide textures and models etc of excellent quality.

The creators of add-ons often do not care to properly adapt their add-ons to the new core development requirements! Looking for errors that enter via bad add-ons is VERY time consuming and cannot be pusued by the devs.

--having done many high-quality textures myself that mostly are part of Celestia's official distribution, I think I know better than many, how time consuming the production of state-of -the-art, scientifically correct HIRES textures really is. In the Celestia team we always try to do things as well as possible and THAT TAKES TIME!

So altogether, your plea has little practical aspects in my view.

Bye Fridger

[s3nn0c]

selden, in my opinion the coherent system of easy-to-download-install-and-use extensions is the fastest method to make Celestia significantly better and the most important thing to do now.

I know this program quite good, I put them on a few cover disks in Polish computer magazines and you know, what? On my computer I have only a basic version of Celestia. I just do not have time to dive into Motherlode, dont have time to test all these extensions, dont have time to read every "how to install" file, dont have time to find files, check them, check their compatibility with other extensions etc etc. Geez, it's horrible to do, believe me! If I have no time to do that, tell me, why do you think a common user will have so much time and patience?

I will write an email to Chris very soon because I want to use Celestia on the cover disk in the Polish popular science magazine (btw being published since 1928...). I want to do that, but in reality it costs me A LOT OF TIME - to find these extensions, to install them, to check them, to write how to install them. Horrible work, believe me. Horrible.

In my opinion - I'm a programmer, a journalist with 20 years of experience in computer sciences, computer graphics and popularization of science - if we want to make Celestia better, before any significant improvements we should make a simple, fast and easy to use system of extensions. It's quite easy to do for Chris, I suppose, and it will show the real power of Celestia - the power of community behind it. And then we, here, could make these packs of textures and extensions - a "star pack", a "cosmology pack", a "space vehicles pack", a "magnetic pack" etc. These things MUST be done. In reality, they are MORE important than any other things Chris can do with graphics and code. He is alone, we aren't. But now, many users can see only his efforts. Efforts of the whole community are available for fans only.

Please understand my point of view. This program has a lot of potential NOW. But we need to standarize extensions. For Chris it will be relatively easy to do. He will do that once.

But how we can make packs of textures and extensions without coherent methods of installing and controlling them? One catalog for extensions, one catalog for textures packs, one method of use them and control them in GUI. That's what is really important, much more than new Milky Way.

[t00fri]

That's what is really important, much more than new Milky Way.

Why do you think you know what is REALLY important??

Concerning Celestia, you obviously have a particular purpose in mind: you want to put it on the title page of some popular science magazine and attract MANY people this way. So apparently 'importance' for you relates to NUMBERS of people and finally to making more money by selling more exemplars. Right?

Bye Fridger

[s3nn0c]

t00fri, you dont need to be rude. There are people on this world which do not care about money. I know they are, because I'm one of them. In my country there are thousends of ways to make money and none of them is connected to popularization of science. So PLEASE dont use that "argument" next time you talk to me, ok?

I like Celestia. Personally. And I know how hard it is to make it look the best. And I know that it's possible to do if you know, where to find right things and what to do with them. The most important thing is the word "to find". Because they exist.

If you make Celestia for yourself, it's ok. It's your motivation, your choice. But I would like to see many people interested in this software. Because it makes astronomy more popular. Because it's good. And because it's made by entusiasts and I always admire people who are able to give to the otheres the most important thing in the world. No, I'm not talking about money, t00rfi. I'm talking about time.

[t00fri]

t00fri, you dont need to be rude.

Trying to sell more journals as a journalist is in no way bad. It's completely natural, actually. Making money is not bad, either. Therefore, I don't see why my presumption should be rude, notably since I don't know you. Most people are determined by money in some way.

I am just trying to understand what drives you. Since you are in no way a Celestia expert compared to us, the devs, your /insistance/ that YOU know better than we do what is important sounds a bit provocative to say the least.

The development team is aware of widely different possible applications of Celestia. So we try to do justice to them. Maximizing the number of potential Celestia users is NOT the only target to go for !!

But I would like to see many people interested in this software. Because it makes astronomy more popular. Because it's good.

Your reasons are not terribly "smashing", I must say...
Many things are good...

I'm talking about time.

Are you talking about the time of Celestia's developers?

Please don't forget that your proposals are in no way original. They came up various times during the last 5 years. We have discussed intensively the possibilities of making various packs with hires textures etc. There are so many practical difficulties if you think about it more extensively that we chose different priorities very soon.

Bye Fridger

[s3nn0c]

Yeah, my proposals aren't original, I know. I wrote them here many years ago (3?) Nothing has changed.

This is our reality: people download the basic version of Celestia, install the basic version of Celestia and judge the basic version of Celestia. And we still have thousand things - things which are ready for use - which practically aren't available for the others, because they are too hard to find, to hard to install.

When will it be a priority? After next 5 years? Do we need next 5 years of almost wasted efforts of so many people?

Show me, t00rfi, one thing which could improve Celestia so much like a good interface for extensions and easy to use packs. If you want to make Celestia really good in very short time, making an order with extensions is the best thing to do. Community will do the rest - will make packs, will rewrite extensions. But these people need a solid standard implemented by developers.

How many developers are working at Celestia, tr00fi? And how many fans do that? Why do you think that time of a few developers is more important than time of hundreds of fans? After so many years, Celestia is not yours anymore. It's ours - it belongs to Chris, to developers, to community. It's Chris property, but it belongs to us too, because people do the same developers do, tr00fi - invest their time into Celestia. And they do that because they think developers do amazing work.

I dont ask for much. I think it's possible to solve the problems with extensions in a month or so. It should be a priority now - because it never was.

[t00fri]

Yeah, my proposals aren't original, I know. I wrote them here many years ago (3?) Nothing has changed.

Why don't you just accept that we know better about the inherent problems associated with making and releasing of e.g. an official hires texture pack or of coding a cross-platform add-on manager etc.

The release of an official HIRES texture pack is an entirely different level of challenge compared to ongoing texture projects by individual users.

We have long planned to do it as a collaborative effort among experts. But it is tremendously time consuming to do it WELL. It requires expertise on many fronts. Only few have the required Know How. Chris has never been involved with the making of large textures, for example. Also he is going to be busy with many other tasks that require his unique expertise with OpenGL.

To get an idea about the required texture work, just have a look in my respective Celestial Matters articles ...

http://www.celestialmatters.org/cm/host ... itan.shtml
http://www.celestialmatters.org/cm/host ... etus.shtml
http://www.celestialmatters.org/cm/host ... thys.shtml

Among many other things, tedious reprojection and alignment of photographic images with specialized software is e.g. required. This in turn requires knowledge of all imaging parameters like distance and direction of sight, imaging scale etc .

The coding of an add-on manager is only a realistic option, once we have recoded Celestia's graphical user interface with a new, cross-platform GUI toolkit! The use of the latter is still in the planning stage.

Don't forget, Celestia is supporting ALL popular operating systems and thus any changes of the GUI requires separate implementations for each OS at present. We sure would not want to spend the work of coding a /platform-specific/ add-on manager NOW.

++++++++++++++++++++++++++++
I enumerated just a few of the underlying difficulties and considerations to illustrate that many things may look quite different and much simpler to an outsider ...
++++++++++++++++++++++++++++

Anyhow, unlike your above claim, during the last 3 years there have been MANY changes with Celestia after a lot of hard work:

--Meanwhile, Celestia has been downloaded more than 5 million times just from SourceForge!

--NASA and ESA are using Celestia and so do countless planetariums and schools all over the world.

--Last fall, Chris was invited to give a talk about Celestia at an international scientific conference by ESA in Holland, etc.

--A host of new widely demanded features has been implemented into the forthcoming Celestia distribution. We are still debugging since there is so much new code. Soon, the code will be wrapped up and version 1.5.0 will be published. Just have a look in the Changelog file what has changed.

For example,

++ A SPICE interface has been implemented that was urgently requested by the professional spacecraft/rocket community!

++ General reference frames have been introduced offering much new potential notably also for professional applications.

++ We have a completely new atmosphere code based on the general Mie scattering theory that still requires lots of further improvements.

++ My catalog of 10000+ galaxies has been massively updated with much more accurate distance data and ALL galaxies of the 'local group' are now implemented, too. This was only possible by merging and exploiting 10 professional galaxy catalogs using ~1000 lines of new PERL script code, for example.

++ A configurable educational overlay interface has been done using Lua.

--Superfast cross platform tools have been written for allowing people with normal home computers to generate huge (virtual) textures of the highest quality with normal home computers and without being experts. At the push of a button, directly from the published scientific raw data.

http://www.celestialmatters.org/cm/hosts/home_t00fri/

My respective thread here "Using the Nmtools Package"
has 21236 readers and 254 posters, certainly demonstrating people's interest and participation!

http://www.celestiaproject.net/forum/viewtopic ... 46&start=0

--Some users with a strong (professional) background in graphics have done marvellous spacecraft models. Have a look at some excellent ones here, for example.

http://www.celestialmatters.org/cm/host ... lchristou/


... just to name a few highlights.

We think this is not a bad achievement for a SMALL dev team each member of which is strongly involved with their daily professional duties besides Celestia.

++++++++++++++++
Celestia is a general FRAMEWORK for (scientific) visualization of the Universe, NOT an application specializing for a particular purpose.
++++++++++++++++

If many people use it for getting more familiar with the Universe it's great. But there are numerous other equally important applications of such a general framework that we do not intend to neglect! Chris and my main interests are certainly more in the domain of scientific (and educational) visualization!

++++++++++++++++++++++++++
Celestia is NOT exclusively targetted at users
who are constantly in a hurry as you seem to be... .
++++++++++++++++++++++++++

Rather, it is done to satisfy highest standards in accuracy and 3D graphics for a wide and different range of user demand. That takes TIME, yet this philosphy is appreciated by many. We KNOW this.

Bye Fridger

[s3nn0c]

t00fri, I dont understand your arguments, to tell the truth.

I know you - developers of Celestia - did an amazing work. But please let me tell you that there are other people too and that they have their rights to tell their opinions.

And my opinion is - especially after your latest answers - that you are unable to see the problems with Celestia which are REAL problems for common users. You are a developer. For you everything here is obvious and easy and the most important thing is a better algorithm or a good code optimization. But if you have no time to do everything, let other people do their job. What do we have now? The real power of Celestia is not the code alone - it's the community too. People do many extensions, many models, many textures. And how many people use them, t00fri? How many?

Celestia need a simple, easy to use interface for extensions. In my opinion it's more important than fireworks with graphics. If you - developers - dont have time to make everything, it is the fastest way to polish your software.

Just try to look at Celestia like a common user, t00fri. You definitely know too many advanced Celestia users...

And yes, I know how much time it costs to make a good texture and I am able to read history files...

[ElChristou]

Guys, no need to go further in the discussion, everybody is aware something must be done to at least release some "official" highres packs and later implement a system for automatic update/installs.

I think that there is only a few challenge Celestia need to win before completing the frame (at code level), so once this done all the time will go to polish and others little comforts for final users.

Now as you probably know, the dev team of Celestia is quite small, so it would be a pity to use the little dev time we have to resolve some of those polishing point meanwhile others soft (OpenSource or commercial) will take vantage and implement some of those missing features we are after...

[Cham]

s3nn0c,

currently, there's a very high priority for the dev team : to debug the code before releasing an official version of Celestia 1.5.0. So please, don't interfere with this until they got it right. It's already a very complicated matter.

[chris]

s3nn0c,

currently, there's a very high priority for the dev team : to debug the code before releasing an official version of Celestia 1.5.0. So please, don't interfere with this until they got it right. It's already a very complicated matter.

It's true that getting 1.5.0 wrapped up is top priority, but there's no need to discourage people from making suggestions. Making add-ons easier to use is a very high priority.

--Chris

[ElChristou]

A stuff that we should think about is a way to tag the addons with several info (author, version, description), just like for the scripts (title tag) in the eventuality of an addon manager...

[hank]

It's true that getting 1.5.0 wrapped up is top priority, but there's no need to discourage people from making suggestions. Making add-ons easier to use is a very high priority.

--Chris

I believe the current development roadmap for Celestia has "Dynamic loading and unloading of add-ons" and "Make add-ons easier to install" at the top of the list for Celestia 1.7.0.

- Hank

[t00fri]

I believe the current development roadmap for Celestia has "Dynamic loading and unloading of add-ons" and "Make add-ons easier to install" at the top of the list for Celestia 1.7.0.

- Hank

Right, Hank, that was also the basis for my arguments above! Celestia 1.4.1 has appeared on February 16 2006, about 1.5 years after the previous 1.3.2 release! Today we are 1.5 years later and the next version (1.5.0) has not yet appeared. With a bit of realism (which I always apply!) version 1.7.0 looks like being quite far in the future . I also do NOT assume that by that time Chris will still be working 'full-time' on Celestia ...

Incidentally, the user s3nn0c made very similar proposals already 3 years AGO.
So Chris, if you find his requests so important,

...there's no need to discourage people from making suggestions. Making add-ons easier to use is a very high priority.

why wasn't it implemented MUCH earlier already?

The real difficulties with addons are on a completely different level.

+++++++++++++++++++
In my view, the following 2 issues represent the main facts responsible for the frequent trouble with addons as well as for the inherent difficulties for Newbies to select good ones:
+++++++++++++++++++

1) Serious problems with add-ons arise frequently, since unexperienced users often produce add-ons that are technically unsatisfactory or since the addons are not properly updated by their creators along with the ongoing core development. In these typical problem cases an addon manager will NOT help!

2) As long as addons by the "community" are dumped on the Motherlode without any kind of "peer reviewing", the chaos is inevitably complete for new users. I have always been strongly opposed to this way of making addons availabe without any quality control by experts.

Originally s3nn0c requested that the devs should

...choose the best textures and the best extensions and should make them available as two separate packs, both easy to download and install. This is relatively easy to do

... make these packs of textures and extensions - a "star pack", a "cosmology pack", a "space vehicles pack", a "magnetic pack" etc. These things MUST be done.

Much fun with this! [ How about starting right away with a "cosmology pack" ? ]

As was discussed repeatedly, such an extensive "packaging" task will just be unrealistic, given the present setup of Celestia development.

Moreover, as we also discussed, a hires texture pack would have to be limited to 4k x 2k textures, in order to keep the total download volume within acceptable bounds etc.

Suppose, say, ElChristou has worked hard for half a year to produce such a 4k x 2k 'hires' pack, then I will come along (for example) with some great 8k x 4k textures or with 64k x 32k monster VT's and sure enough, people will want to use these instead of the /smaller/ official 4k x 2k ones....

...and we are back at the beginning of this discussion .

I hope that this example helps a bit to illustrate that the only practicable way with "community"-addons is peer review along with a well-known add-on site. Of course, there is also the scenario that Runnar, ElChristou and I have adopted in our CelestialMatters site. While there the material is of very high quality, the scope will always be strongly limited because we are only 3 graphics experts.


Bye Fridger

[hank]

With a bit of realism (which I always apply!) version 1.7.0 looks like being quite far in the future ;-) . I also do NOT assume that by that time Chris will still be working 'full-time' on Celestia ...

You may be right, Fridger, but I think it's reasonable to expect that the pace of development will accelerate substantially with Chris now working on Celestia full-time. So 1.7.0 may not be as far off as you think. The purpose of my post was simply to point out that improved add-on support is already on the roadmap. Of course, at this point the roadmap isn't very specific about what Chris has in mind. Hopefully he will add detail to it as his plans evolve.

- Hank

[t00fri]

With a bit of realism (which I always apply!) version 1.7.0 looks like being quite far in the future . I also do NOT assume that by that time Chris will still be working 'full-time' on Celestia ...

You may be right, Fridger, but I think it's reasonable to expect that the pace of development will accelerate substantially with Chris now working on Celestia full-time. So 1.7.0 may not be as far off as you think. The purpose of my post was simply to point out that improved add-on support is already on the roadmap. Of course, at this point the roadmap isn't very specific about what Chris has in mind. Hopefully he will add detail to it as his plans evolve.

- Hank

Hank,

sure...but let's wait and see . According to what I know there are quite a few "unknowns" in this "equation" ...

We should at least have Celestia working with ONE cross-platform toolkit (Qt) before addressing GUI add-on managers and the like.

Bye Fridger

[s3nn0c]

t00fri, please... A few years ago I wrote some points about addons, there was an answer like "we have more important things to do". Last days I wrote my suggestions once again - nothing has changed.

I understand that I definitely try too often.

[hank]

We should at least have Celestia working with ONE cross-platform toolkit (Qt) before addressing GUI add-on managers and the like.

Agreed. It doesn't make sense to devote a lot of effort to a GUI add-on manager until we have a cross-platform GUI in place. (I'm not sure if a decision has been made on whether to use Qt, or wxWindows, or something else.) I don't see the cross-platform GUI explicitly on the roadmap, but it may be included in the "Major UI improvements" targeted for Celestia 2.0.

- Hank

[t00fri]

We should at least have Celestia working with ONE cross-platform toolkit (Qt) before addressing GUI add-on managers and the like.

Agreed. It doesn't make sense to devote a lot of effort to a GUI add-on manager until we have a cross-platform GUI in place. (I'm not sure if a decision has been made on whether to use Qt, or wxWindows, or something else.) I don't see the cross-platform GUI explicitly on the roadmap, but it may be included in the "Major UI improvements" targeted for Celestia 2.0.

- Hank

As far as I understand, Christophe T. is tacitly working since some time (i.e. after our last respective discussion) on porting the Qt3 GUI from the KDE version to Qt4 which then should compile (in principle ) under Windows...

Bye Fridger