Preliminary thoughts for the next Unified UI

[selden]

Selden, no thoughts for better handling of urls, no thoughts on the differences/needs of Pro vs Hobb?

I don't know what you include when you call it "Hobb" (short for Hobbiest, I assume.)
When I think of "hobbiest" in relation to Celestia, I think of those people usually called amateur astronomers. They're already familiar with many of the astronomical concepts.

I think the best way to make Celestia more approachable for people who are not familiar with astronomical concepts would be improved documentation, especially tutorials on YouTube (or elsewhere) which show how to do things. They'd range from installing and running Celestia through downloading and viewing Addons, to creating simple and more complex Addons. I've noticed that a few Celestia demo videos have been published by individuals, but there hasn't been any organized effort.

[ANDREA]

...My spin: Make Celestia as "neutral" as possible, stable and optimized for performance....

I agree on this.
Perhaps I lost my way among the many discussions about Celestia perfoirmance, but If I'm right actually it takes no advantage from multicore CPUs use, even if actually they are going to be the highest percentage of available ones.
Optimizing Celestia for double or quad cores would allow more people with less performing systems to obtain the same results that actually need top performing systems.
Bye

Andrea

[ElChristou]

Frankly, I think deriving multiple "usage profiles" for Celestia is counter-productive in a sense. Why would an educator need different capabilities than a hobbyist or a professional? It depends on the application, not necessarily the person using it.
When one starts to define an application by the "needs" of multiple classes of users, development can become bogged down very quickly. An educator needs this, a professional needs that and a hobbyist needs something else entirely...

"You can only please some of the people some of the time..." So now, instead of developing new and exciting core functionality for everyone, the Celestia devs become mired in pleasing a particular class of user. And, it's not FUN to code anymore...

Your point here are very true. The only little problem is that time is passing and meanwhile the devs are enjoying coding new stuff, the community is vanishing and probably the people who have download Celestia and had fun a couple of hours have trashed the app because apart looking at some painted balls and white dots on a black background, there isn't much to do.
So perso I've followed with great interest this project during more than 4 years now and since this time the same problems unsolved. So it would be probably a good thing to do a pause and define what we want. Can be by targeting some public, can be without targeting any public, whatever, but the purpose of this thread and other threads to come is to solve once for all the deficiencies of Celestia, then again dev on a stable and healthy base.

Targeting audiences is a way to do something useful to some people while IMHO having something vague won't please anybody. Perhaps we can marry the needs of all, why not? This discussion is to get a clearer view of where we go.

Celestia is a bout to merge the various UIs in an unified one. It's the right moment to at least dev a more efficient UI. To do that, no way, you have to analyze a bit what people do or want to do, else you do nothing!

Look at Sketchup, a wonderful application that can be customized by using Ruby and the published Sketchup Ruby API. Or, if one is more hardcore, there is an C/C++ SDK available. There are literally hundreds of Ruby scripts for Sketchup for everything from exporting to photorealistic rendering engines , to designing staircases or aircraft. Sketchup did not "define" the user base. The users define what they want from Sketchup and they have the necessary tools to do so.

My spin: Make Celestia as "neutral" as possible, stable and optimized for performance and make sure the scripting interface is well defined, coherent and published. Let the end user customize as they see fit with the programs constraints.

In theory that would be fantastic. In practice I'd like to see because Celestia do not create something, a model an image whatever. Your example with Sketchup or any soft with a plug in architecture is good, but in the case of Celestia the only plug in we could imagine is an application. I'm not coder but I have some strong doubts on how many end users would wrote his own application to fulfill his needs. Pro can, of course. Hobbysts? 1, 2, 5%? So perso it's why I find your comment here not so constructive. Don't get me wrong, I like the idea of a neutral core and satellites applications but we need more feedback, more opinions.

If the question is one more of what UI Celestia should have, again, IMO that's a decision the devs can make based on technical expertise and UI code integration concerns.

Hey, your're funny! where do you think this thread is? Users? Textures? Purgatory?

[ElChristou]

Selden, no thoughts for better handling of urls, no thoughts on the differences/needs of Pro vs Hobb?

I don't know what you include when you call it "Hobb" (short for Hobbiest, I assume.)
When I think of "hobbiest" in relation to Celestia, I think of those people usually called amateur astronomers. They're already familiar with many of the astronomical concepts.

A few post above I've explain this; "Hobb" for Hobbyists, "Pro" for professionals. Professionals are people working in the field of astronomy, astrophysics or whatever space mission, while Hobbyists are the rest. From the 6 years old kid to the teacher, to the amateur astronomer or else. I do the difference because eventually the needs (we are talking at UI level!) can be opposite...

I think the best way to make Celestia more approachable for people who are not familiar with astronomical concepts would be improved documentation, especially tutorials on YouTube (or elsewhere) which show how to do things. They'd range from installing and running Celestia through downloading and viewing Addons, to creating simple and more complex Addons. I've noticed that a few Celestia demo videos have been published by individuals, but there hasn't been any organized effort.

That would represent a lot of work, perhaps more than actually build an intuitive UI which purpose is ease the use and so reduce the need of such docs...

[SiriusCG]

Well, let's look at something that I would consider an important element of "user friendliness": resource creation.

While I've not been the most active participant in Celestia, I have had an interest in it since about 1.4.1 or thereabouts. I've read through a number of threads and at times would have like to created some content but...

IMO, it's too complicated and it's a chore. The content creation pipeline is a too convoluted! Don't get me wrong, and I mean no offense to anyone who has had the tenacity and skill to produce content, but let's be real. The pipeline is much too complicated for the average user I think. By the time one gets the requisite programs installed and in working order, then creates the necessary meshes and textures, does the needed conversions, etc; it's more like work. And who wants to do any more of that than they have too?

If one wished to attract more users to Celestia, I think there are three things that are necessary:
1. The application must be stable and performance must be good. I'm not saying that isn't the case, just a condition.
2. Content creation tools that are user-friendly and intuitive to use. Face it, "Joe Average User" these days is not into "problem solving". They want it simple. And, given the proper tools, people like to "mod" games and sims. It's fun, or supposed to be!
3. Demos showing how content can be created, installed and managed. I believe Seldon commented on this earlier.

I'm not coder but I have some strong doubts on how many end users would wrote his own application to fulfill his needs. Pro can, of course. Hobbysts? 1, 2, 5%?

I think you have to define "application" from "script" even though the two can overlap. For instance, I would be inclined to use Sketchup and write Ruby scripts to get my goals achieved. Then share those scripts with the community. OTOH, I agree with you, I would not be inclined to write an entire 3D application just to build Celestia content.

I come from a UNIX background where it is very common to use scripting languages to "glue together" discreet programs, like the Gimp or Imagemagick, to do the actual work. I would apply the same methodology here.

Hey, your're funny! where do you think this thread is? Users? Textures? Purgatory?

Sorry, I don't understand your humor here. Probably just me...

Cheers.

[ElChristou]

Well it could be better, if you would be offered more than one link (InfoURL) to an object. See my ideas here: viewtopic.php?f=2&t=11971
And especially, if the links are still up to date or the website reachable. That could be a problem with the long time between the Celestia releases.

This is clearly a wanted feature and I think it's already on tracks. Anyway from an UI point of view, is the contextual usage the best?

Defining the Educational Activities as Hobbyists is not really the case. I think we should make 3 groups of users:
Professionals
Educators
Hobbyists

Because Educators have other needs for an UI than "normal" "Hobbyists". They need more text to be displayed, more information in an easy way to show and the make it disappear etc.

If needed we could do that, but for now basically the needs of educators or Hobbyist can be the same. More info, sound, video, whatever; in general the more the best for this target. On the contrary, Pros' needs could be much more restricted to very particular stuff depending on their activities. So for now, let's keep two groups, I guess it will be already hard to deal with them...

[ElChristou]

Well, let's look at something that I would consider an important element of "user friendliness": resource creation.

While I've not been the most active participant in Celestia, I have had an interest in it since about 1.4.1 or thereabouts. I've read through a number of threads and at times would have like to created some content but...

IMO, it's too complicated and it's a chore. The content creation pipeline is a too convoluted! Don't get me wrong, and I mean no offense to anyone who has had the tenacity and skill to produce content, but let's be real. The pipeline is much too complicated for the average user I think. By the time one gets the requisite programs installed and in working order, then creates the necessary meshes and textures, does the needed conversions, etc; it's more like work. And who wants to do any more of that than they have too?

If one wished to attract more users to Celestia, I think there are three things that are necessary:
1. The application must be stable and performance must be good. I'm not saying that isn't the case, just a condition.
2. Content creation tools that are user-friendly and intuitive to use. Face it, "Joe Average User" these days is not into "problem solving". They want it simple. And, given the proper tools, people like to "mod" games and sims. It's fun, or supposed to be!
3. Demos showing how content can be created, installed and managed. I believe Seldon commented on this earlier.

Here I do agree completely and I guess no one can say the contrary. Creating a descent intuitive UI is part of the process. It's what we are after.

Hey, your're funny! where do you think this thread is? Users? Textures? Purgatory?

Sorry, I don't understand your humor here. Probably just me...

You were just saying: "IMO that's a decision the devs can make based on technical expertise and UI code integration concerns". So I find this funny because actually what do you think this thread is for? Isn't this forum a dev only forum? If such thread is open, perhaps it's because we need feedback and more opinions... anyway, forget it's not important.

[SiriusCG]

You were just saying: "IMO that's a decision the devs can make based on technical expertise and UI code integration concerns". So I find this funny because actually what do you think this thread is for? Isn't this forum a dev only forum? If such thread is open, perhaps it's because we need feedback and more opinions... anyway, forget it's not important.

Ah, ok... I see what you meant... I had my "programmer's glasses" on when I read it. Different perspective.

Cheers.

[ElChristou]

Again from the UI point of view:

Perso I'm stuck because on one hand I'd like an intuitive easy to use UI but I fear to achieve this we need to do some compromise with efficiency and that is the principal need of Pros.
As example, imagine the eclipse finder. Sometime ago in a private exchange I was proposing to Chris to eventually switch the eclipse finder as a contextual feature.
The idea is the following; let's see first what are the actual steps between the idea to look after an eclipse and the graphical result on screen:

- you mentally pick a body on which you want to see an eclipse
- you open the eclipse finder
- you enter the name of the body
- you enter a time lap, press enter
- you get a listing of results
- you chose one, click go
- Celestia do a goto and bingo you can see the eclipse

Ok. Now the same but from a contextual point of view:

- you mentally pick a body on which you want to see an eclipse
- you go for it to right click and get a "Find Eclipse" feature in the contextual menu
- you get a panel, enter a time lap, press enter
- you get a listing of results ,
- you chose one of the results,
- Celestia change the time and you see the eclipse

So as you can imagine, the contextual solution could be considered more intuitive because you have a feature linked to the body you are already watching. Also the good point of this is that you can reduce the impact of floating panels, etc, to the minimum. Now clearly a pro will prefer a permanent panel based feature where he can perform several search immediately, compare timing etc,

My fear for now is the impossibility to satisfy both group with the same UI...
Opinions?

[chris]

A couple quick comments, because I don't have much time today:

- I don't think that there needs to be a separate code base for hobbyist and professional users. Maintaining two separate versions would cost developer time, and so far I just haven't seen the need for it. Celestia can be sufficiently customized with scripts and data catalogs, and the ability to customize Celestia can be increased by further development of features for scripting and add-ons.

- We need to address ease of use for all Celestia users. A unified user interface will help: developing new UI features becomes easier, because a new user interface element only needs to be implemented once, not four times. It's also easier to craft a user manual for the program when all versions operate nearly identically. Some things that need to be done to improve the UI are to rationalize the keyboard controls, make add-ons installable with a click, and document everything better.

- A *huge* problem for new Celestia users is simply figuring out what to do. Space is big: there's lots to see, but there are many more places that are very boring indeed. The ability to freely explore the universe is a big part of the Celestia's appeal for me, but the program should also have facilities to guide users to interesting views and provide some explanation for what's on the screen. The way to do this is by expanding the scope of add-ons: instead of just catalogs of objects, there should be a way of integrating them with scripts, camera views, and explanatory text. Already many add-ons come packaged with scripts and HTML files with text and cel URL links. This stuff should be accessible within Celestia and clearly associated with the add-on. The same facilities could be used for the standard Celestia package. It would be a boon for both hobbyist users and educators.

- The 'cruft' that has accumulated in Celestia is a significant obstacle for both new and veteran add-on developers. Aside from the issues mentioned by Andrew and Selden, there's the problem with the non-standard Celestia coordinate system and the inconvenient choice of 'micro light years' as the standard distance unit for scripts. We need to wipe the slate clean of this stuff while retaining compatibility with existing add-ons. It's a big issue, but the specifics of dealing with cruft belong in a separate thread.

--Chris

[ElChristou]

- I don't think that there needs to be a separate code base for hobbyist and professional users. Maintaining two separate versions would cost developer time, and so far I just haven't seen the need for it. Celestia can be sufficiently customized with scripts and data catalogs, and the ability to customize Celestia can be increased by further development of features for scripting and add-ons.

I agree we don't need too separate version, that don't make much sense. Now, eventually we can think in 2 skins, flavors or whatever of the UI if the needs of both target cannot be fulfill via a the same UI. That's and option if we find too much difference between the way of using the soft.

- We need to address ease of use for all Celestia users. A unified user interface will help: developing new UI features becomes easier, because a new user interface element only needs to be implemented once, not four times. It's also easier to craft a user manual for the program when all versions operate nearly identically. Some things that need to be done to improve the UI are to rationalize the keyboard controls, make add-ons installable with a click, and document everything better.

Ok, 3 ideas here; remapping the keyboard and eventually have a full customizable system, addon Installer/Manager and a nice doc to warp the whole stuff. Of course the docs have to come in last, once the rest in in place. The addon Installer/Manager is an important part of the new UI. We need to think it well because it may be related to online databases and other fancy stuff...

- A *huge* problem for new Celestia users is simply figuring out what to do. Space is big: there's lots to see, but there are many more places that are very boring indeed. The ability to freely explore the universe is a big part of the Celestia's appeal for me, but the program should also have facilities to guide users to interesting views and provide some explanation for what's on the screen. The way to do this is by expanding the scope of add-ons: instead of just catalogs of objects, there should be a way of integrating them with scripts, camera views, and explanatory text. Already many add-ons come packaged with scripts and HTML files with text and cel URL links. This stuff should be accessible within Celestia and clearly associated with the add-on. The same facilities could be used for the standard Celestia package. It would be a boon for both hobbyist users and educators.

That would be great but that means someone will have to work on official content. Now what I cannot avoid to ask me is if this content will not be cumbersome for Pro users?

[selden]

- A *huge* problem for new Celestia users is simply figuring out what to do. Space is big: there's lots to see, but there are many more places that are very boring indeed. The ability to freely explore the universe is a big part of the Celestia's appeal for me, but the program should also have facilities to guide users to interesting views and provide some explanation for what's on the screen. The way to do this is by expanding the scope of add-ons: instead of just catalogs of objects, there should be a way of integrating them with scripts, camera views, and explanatory text. Already many add-ons come packaged with scripts and HTML files with text and cel URL links. This stuff should be accessible within Celestia and clearly associated with the add-on. The same facilities could be used for the standard Celestia package. It would be a boon for both hobbyist users and educators.

That would be great but that means someone will have to work on official content. Now what I cannot avoid to ask me is if this content will not be cumbersome for Pro users?

I think perhaps you are reading too much into what ChrisL wrote. We're talking about Addons here. It's up to the Addon's author to provide those files. If they're not provided, they can't be cumbersome. I also interpreted what ChrisL wrote to mean that Celestia should be able to display to the user what scripts, html and other auxiliary files came with the Addon, and that Celestia should be able to execute the scripts or display the html (or Readme or PDF or other) files without forcing the user to iconize or otherwise get out of Celestia: Celestia would spawn a copy of an appropriate viewer. As a corollary, scripts should be in a directory associated with their Addon, they shouldn't be put in Celestia's own scripts directory.

[ElChristou]

I think perhaps you are reading too much into what ChrisL wrote. We're talking about Addons here. It's up to the Addon's author to provide those files. If they're not provided, they can't be cumbersome. I also interpreted what ChrisL wrote to mean that Celestia should be able to display to the user what scripts, html and other auxiliary files came with the Addon, and that Celestia should be able to execute the scripts or display the html (or Readme or PDF or other) files without forcing the user to iconize or otherwise get out of Celestia: Celestia would spawn a copy of an appropriate viewer. As a corollary, scripts should be in a directory associated with their Addon, they shouldn't be put in Celestia's own scripts directory.

yes sorry I misread this part and focused on "ChrisL: The same facilities could be used for the standard Celestia package."
So I was seeing some official content to let users discover some basis and possibilities...

[Guckytos]

- The 'cruft' that has accumulated in Celestia is a significant obstacle for both new and veteran add-on developers. Aside from the issues mentioned by Andrew and Selden, there's the problem with the non-standard Celestia coordinate system and the inconvenient choice of 'micro light years' as the standard distance unit for scripts. We need to wipe the slate clean of this stuff while retaining compatibility with existing add-ons. It's a big issue, but the specifics of dealing with cruft belong in a separate thread.
--Chris

Here I disagree with you, Chris.

If we are really going to tackle these things and make a clean slate, then the first priority should be the clean slate.
If it is possible in the process to retain backwards compability, it's okay. But if it is not possible, then the compability shalle be ignored. It is a lot more important to have a clean, consistant base for all units and coordinate systems!
If necessary, the addons have then to be converted. That's perhaps a lot of work for the creators depending on the changes, and we will perhaps 'loose' some good, old, beloved addons. But you can always use an older version of Celestia to use them.

But for the future we would have a clean base and won't have to think about all those discrepancies.
Don't make the MS error of implementing backwards compability at all costs.

My 2 cents.

Best regards,

Guckytos

[chris]

I think perhaps you are reading too much into what ChrisL wrote. We're talking about Addons here. It's up to the Addon's author to provide those files. If they're not provided, they can't be cumbersome. I also interpreted what ChrisL wrote to mean that Celestia should be able to display to the user what scripts, html and other auxiliary files came with the Addon, and that Celestia should be able to execute the scripts or display the html (or Readme or PDF or other) files without forcing the user to iconize or otherwise get out of Celestia: Celestia would spawn a copy of an appropriate viewer. As a corollary, scripts should be in a directory associated with their Addon, they shouldn't be put in Celestia's own scripts directory.

yes sorry I misread this part and focused on "ChrisL: The same facilities could be used for the standard Celestia package."
So I was seeing some official content to let users discover some basis and possibilities...

You didn't misread: I do think that the standard content should be supplemented with scripts, bookmarks, and other information that will guide exploration. But, I don't think that this needs to get in the way of anyone--professional or otherwise--who is using Celestia. Hide a window and the information is gone until you reactivate it via the menu or some other means.

--Chris

[t00fri]

I think perhaps you are reading too much into what ChrisL wrote. We're talking about Addons here. It's up to the Addon's author to provide those files. If they're not provided, they can't be cumbersome. I also interpreted what ChrisL wrote to mean that Celestia should be able to display to the user what scripts, html and other auxiliary files came with the Addon, and that Celestia should be able to execute the scripts or display the html (or Readme or PDF or other) files without forcing the user to iconize or otherwise get out of Celestia: Celestia would spawn a copy of an appropriate viewer. As a corollary, scripts should be in a directory associated with their Addon, they shouldn't be put in Celestia's own scripts directory.

yes sorry I misread this part and focused on "ChrisL: The same facilities could be used for the standard Celestia package."
So I was seeing some official content to let users discover some basis and possibilities...

You didn't misread: I do think that the standard content should be supplemented with scripts, bookmarks, and other information that will guide exploration. But, I don't think that this needs to get in the way of anyone--professional or otherwise--who is using Celestia. Hide a window and the information is gone until you reactivate it via the menu or some other means.

--Chris

Chris,

As far as I can tell, there is already a working counter example in existence, where you are actively involved and paid for:

++++++++++++++++
Celestia.STA
++++++++++++++++
http://www.personal.soton.ac.uk/jf1w07/STA/
http://sta.estec.esa.int/Space_Trajecto ... /Home.html

Here is the SVN reference:

http://sta.wiki.sourceforge.net/STA+BINARIES

LEGAL says:
--------------------------
STA code is open source. The code in the STA project is license under the GENERAL PUBLIC LICENSE, Version 2, June 1991 of the Free Software Foundation, commonly known as GNU GPL.
...
In the STA project, all Intellectual Property Rights (IPRs) are assigned to ESA.

STA Partners and Team:
-----------------------------------------
Universities and entities partnering with ESA to date are as follows:

* Technical University of Delft (Netherlands)
* University of Bremen (Germany)
* ZARM institude of Technology (Germany)
* University of Coimbra (Portugal)
* Complutense University (Spain)
* University of Birmingham (United Kingdom)
* Instituto Superior Technico de Lisboa (Portugal)
* University of Southampton (United Kingdom)
* Politechnico di Milano (Italy)
* Celestia team

The structure in SVN is like this (as you know very well of course):

celestia-src/ 302 5 days cjlaurel Fixes for building with QtCreator on Mac OS X.
doc/ 300 4 weeks cjlaurel Copied qt4 branch to trunk.
iconary/ 300 4 weeks cjlaurel Copied qt4 branch to trunk.
macosx/ 300 4 weeks cjlaurel Copied qt4 branch to trunk.
sta-data/ 300 4 weeks cjlaurel Copied qt4 branch to trunk.
sta-src/ 304 3 hours cjlaurel Moved qmake/QtCreator project file from sta-src to trunk.
windows/ 300 4 weeks cjlaurel Copied qt4 branch to trunk.
STA.pro 304 3 hours cjlaurel Moved qmake/QtCreator project file from sta-src to trunk.

+++++++++++++++++++++++++++++
The structure is very much how I consider further possible specialized setups:
+++++++++++++++++++++++++++++

1) Various additional code modules like a Plotter module, a Network module various dedicated calculation modules AND a code interface to Celestia proper (residing in the celestia-src directory)

2) sta-data contains largely Celestia's data.

So...why did you just argue that additional code for modularizations on the professional level is unnecessary? I don't quite understand.

Fridger

[ElChristou]

yes sorry I misread this part and focused on "ChrisL: The same facilities could be used for the standard Celestia package."
So I was seeing some official content to let users discover some basis and possibilities...

You didn't misread: I do think that the standard content should be supplemented with scripts, bookmarks, and other information that will guide exploration. But, I don't think that this needs to get in the way of anyone--professional or otherwise--who is using Celestia. Hide a window and the information is gone until you reactivate it via the menu or some other means.

Ok.
Mmhh... for now I have some trouble to imagine content and features to let the user discover more stuff about the possibilities of Celestia but a simple window that we can hide to do the job seems a bit limited to me. I still don't have a clear image of this stuff (would be cool to write down some examples), but I guess it have to be a bit more "present" to make really sense. Remember the idea to have some grids related to the solar system linked to Sol, etc, this kind of features eventually become part of the UI, it's why I was saying that such thing could become cumbersome for people (pros) who don't need this...

Edit, sorry Fridger, cross post!

[chris]

Chris,

As far as I can tell, there is already a working counter example in existence, where you are actively involved and paid for:

Celestia.STA

Here is the SVN reference:

http://sta.wiki.sourceforge.net/STA+BINARIES

This situation with STA is different... STA is an application for modeling and analyzing spaceflight trajectories. STA's 3D visualization module uses the Celestia engine--it's just one component of an application in which visualization is an important feature but not the raison d'?tre.

--Chris

[t00fri]

Chris,

As far as I can tell, there is already a working counter example in existence, where you are actively involved and paid for:

Celestia.STA

Here is the SVN reference:

http://sta.wiki.sourceforge.net/STA+BINARIES

This situation with STA is different... STA is an application for modeling and analyzing spaceflight trajectories. STA's 3D visualization module uses the Celestia engine--it's just one component of an application in which visualization is an important feature but not the raison d'?tre.

--Chris

Professional applications mostly require more specialized tools! STA seems like a typical extension along such lines.

I know about a few completely analogous and most interesting possible applications. For example, whenever, there are important intermediate calculations required, along with plotting and analysis tools.

Fridger

[ElChristou]

Fridger, Chris, discussing about a svn structure or code modif to eventually fulfill the needs of professional applications is necessary but would be nice if it could be done in it's own thread!

Before continuing the UI topic, what if we try to find a few examples of default content and other ways to make Celestia more interesting/appealing on the long run?

Here a thread to discuss such features: http://shatters.net/forum/viewtopic.php?f=10&t=13558