Celestia Forums

In Celestia 1.5, the cel URL's that were working correctly with version 1.4.1 are taking me to different "views". (It is also the case for the URL in the Nasa web site - http://learn.arc.nasa.gov/planets/)
What is happening ?

I'm using the latest release, Celestia-1.5 12-17-07.dmg, for Mac.

jrichelle wrote:In Celestia 1.5, the cel URL's that were working correctly with version 1.4.1 are taking me to different "views". (It is also the case for the URL in the Nasa web site -

The cel url's that were created with Celestia 1.4 used the Sun as the origin. The cel url's that are created with Celestia 1.5 use the Solar system barycenter as the origin. That is why they don't work.

You could probably get Celestia 1.5 to work if you edit the nearstars.stc file and define the origin as the Sun or replace the nearstars.stc with the one that comes with Celestia 1.4. I used to do this after the origin changed a while back in cvs versions. I would copy the original 1.5 version to a different folder so you could change it back if you decide to.
cartrite

cartrite wrote: The cel url's that are created with Celestia 1.5 use the Solar system barycenter as the origin. That is why they don't work.
cartrite

This has an effect on a script I derived from a cel url using "gotoloc". Is there a way to tell in the script that the solar system barycenter must be used as the origin ? or a way to write a script that would work as well under 1.4 and 1.5 ?

Is there a manner to transform 1.4 cel url's into 1.5 cel url's ?

jrichelle wrote:Is there a manner to transform 1.4 cel url's into 1.5 cel url's ?

As far as I know, Celestia 1.5.0 should detect the cel:urls from < 1.5.0 when the cel:url has no version tag (&ver=2).
But in fact, this doesn't work in 1.5.0pre5.
I'm pretty sure that this did work in earlier pre-versions...

Adirondack

Adirondack wrote:

jrichelle wrote:Is there a manner to transform 1.4 cel url's into 1.5 cel url's ?

As far as I know, Celestia 1.5.0 should detect the cel:urls from < 1.5.0 when the cel:url has no version tag (&ver=2).
But in fact, this doesn't work in 1.5.0pre5.
I'm pretty sure that this did work in earlier pre-versions...

Adirondack

I don't think that it did. I never got around to writing the code to transform between the two URL versions.

--Chris

chris wrote:I don't think that it did. I never got around to writing the code to transform between the two URL versions.

Then I must have been dreaming.
But that is the intention of the version tag, no?
That means that we now have a bug-tracker issue?

Adirondack

Adirondack wrote:

chris wrote:I don't think that it did. I never got around to writing the code to transform between the two URL versions.

Then I must have been dreaming.
But that is the intention of the version tag, no?
That means that we now have a bug-tracker issue?

That is the intention of the version tag, but the incompatibilities between URL versions 1.5 and before are numerous and difficult to address. The most severe problem is the change in the constant kilometers per light year: it's unfortunately a lot involved than just applying a scale factor. Then there's the fact that the origin was moved to the Solar System Barycenter and a bunch of precision fixes. We might be able to get a lot of old URLs to mostly work by subtracting the difference between the old origin and the new one, but that's as much work as I'm willing to do on cel URL compatibility. Fortunately, this should be the end of URL compatibility problems, so 1.5.0 URLs should work with any future version of Celestia.

--Chris

Chris:

I just discovered this post. It is the scariest problem I could imagine. In my 12 educational activities, which are in use for version 1.4.1 in hundreds of schools around the world, I have over 350 cel:urls that take students/users to very very specific places. If this post means that virtually none of them will take a user of the Activities to the places they need to go if they use 1.5.0, I am SUNK! It took me hundreds of hours to locate and place educational activity users at those places, dates, times and events. I just cannot start all over again. I can't even face the thought.

Therefore, please tell me specifically what has to be done to get around this problem. Cartrite suggested that one could replace any lines in nearstars.stc with an older entry that does not define the sun's barycenter so that the Activity cel:urls will work with 1.5.0? That would seem to be very simple. Would it work? Would it cause any other problems?

Alternatively, could I leave the new 1.5.0 nearstars.stc file alone in the data folder, but place an alternative nearstars.stc file in my educational-extras folder? That folder contains individual add-ons for each educational activity specifically, and is launched/accessed by command line commands. For example, if I want the 40 separate spacecraft add-ons I used in the Spacecraft educational activity to load (Activity 7), I can launch Celestia with a command line specifying that only the add-ons for Activity 7 are loaded. Could I place an alternative nearstars.stc file in that folder which defines the sun's center at its old location (the one used in 1.4.1)? Would it take precedence over the one in the data folder? Would this work out OK?

If so, what would that entry look like?

Thanks in advance.

HELP!!!

Frank

fsgregs wrote:Chris:

I just discovered this post. It is the scariest problem I could imagine. In my 12 educational activities, which are in use for version 1.4.1 in hundreds of schools around the world, I have over 350 cel:urls that take students/users to very very specific places. If this post means that virtually none of them will take a user of the Activities to the places they need to go if they use 1.5.0, I am SUNK! It took me hundreds of hours to locate and place educational activity users at those places, dates, times and events. I just cannot start all over again. I can't even face the thought.

Therefore, please tell me specifically what has to be done to get around this problem. Cartrite suggested that one could replace any lines in nearstars.stc with an older entry that does not define the sun's barycenter so that the Activity cel:urls will work with 1.5.0? That would seem to be very simple. Would it work? Would it cause any other problems?

Alternatively, could I leave the new 1.5.0 nearstars.stc file alone in the data folder, but place an alternative nearstars.stc file in my educational-extras folder? That folder contains individual add-ons for each educational activity specifically, and is launched/accessed by command line commands. For example, if I want the 40 separate spacecraft add-ons I used in the Spacecraft educational activity to load (Activity 7), I can launch Celestia with a command line specifying that only the add-ons for Activity 7 are loaded. Could I place an alternative nearstars.stc file in that folder which defines the sun's center at its old location (the one used in 1.4.1)? Would it take precedence over the one in the data folder? Would this work out OK?

If so, what would that entry look like?

Thanks in advance.

HELP!!!

Frank

Simply removing the Sun's orbit in nearstars.stc isn't enough. You'd have to place it at exactly the position it had in 1.4.1. This is possible to do, but someone (and I'd like it to not be me) will have to figure out where the Sun was in earlier Celestia versions and modify nearstars.stc appropriately.

This is not a completely solution, but it may suffice for most of the URLs. The other problem is the incorrect value for the number of kilometers per light year that was used in earlier versions of Celestia. I'm not sure how many of your URLs would still be affected by this once the position of the Sun was changed. Unfortunately, it's a very difficult problem to fix

I've said this many times before, but the URL code in Celestia really needs to be modified to store relative positions. Otherwise, URLs will become invalid any time that object positions change in Celestia. For example, I want to replace the current Moon orbit with a much more accurate version; unfortunately, this would break any current cel URLs that positioned the camera near the Moon. Relative cel URLs would be resistant to such changes.

--Chris

Simply removing the Sun's orbit in nearstars.stc isn't enough. You'd have to place it at exactly the position it had in 1.4.1. This is possible to do, but someone (and I'd like it to not be me) will have to figure out where the Sun was in earlier Celestia versions and modify nearstars.stc appropriately.

OH NO!!!!!!!!!!

help!!! I am great at designing activities and many add-ons, but I have NO IDEA how to tell where the sun is/was in space in Celestia 1.4.1?

Could someone ... anyone ... help me here ... PLEASE??????

Anyone ....

Frank

fsgregs wrote:

Simply removing the Sun's orbit in nearstars.stc isn't enough. You'd have to place it at exactly the position it had in 1.4.1. This is possible to do, but someone (and I'd like it to not be me) will have to figure out where the Sun was in earlier Celestia versions and modify nearstars.stc appropriately.

OH NO!!!!!!!!!!

help!!! I am great at designing activities and many add-ons, but I have NO IDEA how to tell where the sun is/was in space in Celestia 1.4.1?

Could someone ... anyone ... help me here ... PLEASE??????

Anyone ....

Frank

We'll try and figure out something for you, Frank. It's a nasty, tricky problem.

--Chris

Chris,
You mean where the sun was in the galaxy because of it's rotation around the galaxy? How would you even go about doing that with nearstars.stc? There doesn't seem to be many options available to modify nearstars. Distance?.

Isn't there also a time change involved? Time dependent url's might not even be fixed after correcting the sun's position?

A while back, maybe over a year now when url's were starting to get affected by the new code, I tried to keep up with the changes but quickly gave up. It used to be as simple as changing the suns position back to the center instead of using the solarsystem barycenter. No more I guess.
cartrite

cartrite wrote:Chris,
You mean where the sun was in the galaxy because of it's rotation around the galaxy? How would you even go about doing that with nearstars.stc? There doesn't seem to be many options available to modify nearstars. Distance?.

I mean that you can't just remove the Sun's orbit about the SSB and expect things to work. In 1.4.1, the Sun wasn't at the center of the universe. It was displaced from the origin by ~200 AU. We'd have to look at the 1.4.1 code and figure out exactly what the location of the Sun was.

Isn't there also a time change involved? Time dependent url's might not even be fixed after correcting the sun's position?

They'd work as long as the Sun's orbit about the SSB was eliminated. I'm not talking about modifying nearstars.stc for Celestia 1.5.0, but it's something Frank could do for Celestia-ED if it was required to make URLs work.

A while back, maybe over a year now when url's were starting to get affected by the new code, I tried to keep up with the changes but quickly gave up. It used to be as simple as changing the suns position back to the center instead of using the solarsystem barycenter. No more I guess.
cartrite

Unfortunately not . . . But the changes causing all this mess were absolutely necessary.

--Chris

Installing both might provide a way.
First everything goes around the sun in 1.4.1
Go to the same date in 1.5.1
Look how far the sun is from barycenter.
Add the distance to the url and rename the origin in the URL.
Paste your new URL in the right place in your script.

There is no easy way, you just have to rewrite parts of your script.

fsgregs wrote:OH NO!!!!!!!!!!

Unfortunately, the CelUrls are only the tip of the iceberg. The problem manifests itself with XYZ files, too. Spacecraft trajectories that were writtten for Celestia versions older than 1.5 Pre3 are off.

I know, i had to re-write my 86 Ran XYZ files completely twice during the development of 1.5...

But, Frank:
The CelUrls are off only in position, not in time. So you do not need to create them all over - you only need to open the old CelUrl, GoTo your object and re-orient the view. Times are correct. This makes the conversion job a LOT easier.

- rthorvald

rthorvald wrote:

fsgregs wrote:OH NO!!!!!!!!!!

Unfortunately, the CelUrls are only the tip of the iceberg. The problem manifests itself with XYZ files, too. Spacecraft trajectories that were writtten for Celestia versions older than 1.5 Pre3 are off.

I know, i had to re-write my 86 Ran XYZ files completely twice during the development of 1.5...

Runar,

If I recall correctly, this wasn't a general problem with XYZ files. I thought it had to do with the fact that the precision of LongLat values and rotation parameters was increased from single to double.

--Chris

chris wrote:

rthorvald wrote:I know, i had to re-write my 86 Ran XYZ files completely twice during the development of 1.5...

Runar,

If I recall correctly, this wasn't a general problem with XYZ files. I thought it had to do with the fact that the precision of LongLat values and rotation parameters was increased from single to double.

Just my 2 cents : Well, it is normal to expect to modify our addon if it is created at the same time as the main software itself ! I don't think we have any right to complain here ! Usually, an addon creator is supposed to wait for the finished software BEFORE creating something for it. I, too, had to change many things in my own addons, while Celestia was developped in the same time as the addons themselves.

I may even experience something similar very soon, since I noticed that Chris is currently adding the feature I was asking : an unclickable feature at the SSC (and STC ?) level. I strongly feel the urge to modify right away all my addons, to use this *VERY* important feature, while it isn't currently finished (?) ! This is the fun of Celestia development, anyway

chris wrote:[
If I recall correctly, this wasn't a general problem with XYZ files. I thought it had to do with the fact that the precision of LongLat values and rotation parameters was increased from single to double.

I think it was both:
- the XYZs was off (i don??t remember why) from 1.4.2 to 1.5
- the LongLats precision changed them again when a longlat was the origin between 1.5 Pre 2 and 1.5 Pre 3.

- rthorvald

Cham wrote:Well, it is normal to expect to modify our addon if it is created at the same time as the main software itself ! I don't think we have any right to complain here !

I know this! My statement here was certainly not a complaint. Just a comparison of Franks current challenge with my own recent experience.

- rthorvald

Chris and everyone:

Thank you for your kind offer to help me fix this problem without having to redo all 350 cel:urls over again. Just tell me what I can do to help.

I have tried out some cel:urls from my 1.4.1 activities in 1.5.0 final. The solar system is off position by 206.27 au. Is that any help in editing those cel:urls to work in 1.5.0?

Some of my cel:urls that do not involve the solar system do still appear to work. For example, the cel:urls to take readers to Proxima Cen or the edge of the Milky Way still work fine. The ones that work are the ones focused on an object, such as a star or galaxy. Sadly, none of my freeflight cel:urls work. I use a lot of them to position a viewer in space and time, with no particular object selected.

I haven't tested them all, but of the 350 or more cel:urls I use, probably 280 of the take place in our around our solar system, or use freeflight as their basis. That makes it even more difficult for me. Since freeflight cel:urls have no particular object selected, there is nothing for me to "go to", to reorient the cel:url in 1.5.0.

I will edit the text of a few cel:urls to see if we can just type a new number in them to compensate for the new means of positioning them, Just tell me what to type/try.

P.S. I assume bookmarks from 1.4.1 will also not go to the proper location in 1.5.0? Is that correct?