search

planeur-net / airspace

Fichier des espaces aériens Francais au format "openair"
5 stars 9 forks source link

Various improvements and fixes #70

Open reskume opened 2 months ago

reskume commented 2 months ago

This commit fixes several issues:

Problem is that it took quite a considerable amount of time to update the file. Recent additions are NOT included! Open for comments/discussion.

llauner commented 1 month ago

Error found at line 3347: Geometry of airspace 'TMA BALE 1' starting on line 3347 is invalid. First and last Position are not equivalent.

Do you think you could run openaip-openair-parser on your change to make sure the file is valid ? Thanks

llauner commented 1 month ago

I've started looking at the PR. Which reference are you using for the OpenAir Extended format ? If I look at: Documentation format OpenAir or Documentation format OpenAir + OpenAir Extended - Naviter The Restricted Area is AY R, and not AY RESTRICTED

Danger is defined as AY Q and not AY DANGER

I'm not sure it's right to classify a RMZ as G (uncontrolled), when it is controlled this you need to have a Radio contact.

The RMZ Le Versoud turned from temporary to fixed, but it's true that the date wasn't modified in the file. It should be kept, I'm not sure if others were deleted from your file ?

Let me know what you think...

reskume commented 1 month ago

I'll have a close look tomorrow when I have more time. I think that some types got mixed up during the rewrite. We are still and maybe continue to use the openaip format until the other one is fixed and done. The openaip types are explicit in there naming for better clarity. I'll also fix the airspace with self-intersections.

Ludovic @.***> schrieb am Mi., 17. Juli 2024, 17:11:

I've started looking at the PR. Which reference are you using for the OpenAir Extended format ? If I look at: Documentation format OpenAir http://www.winpilot.com/UsersGuide/UserAirspace.asp or Documentation format OpenAir + OpenAir Extended - Naviter https://github.com/naviter/seeyou_file_formats/blob/main/OpenAir_File_Format_Support.md The Restricted Area is AY R, and not AY RESTRICTED

Danger is defined as AY Q and not AY DANGER

I'm not sure it's right to classify a RMZ as G (uncontrolled), when it is controlled this you need to have a Radio contact.

The RMZ Le Versoud turned from temporary to fixed, but it's true that the date wasn't modified in the file. It should be kept, I'm not sure if others were deleted from your file ?

Let me know what you think...

— Reply to this email directly, view it on GitHub https://github.com/planeur-net/airspace/pull/70#issuecomment-2233562049, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABROSQCHY4EQB6WVQQ4KJG3ZM2CS7AVCNFSM6AAAAABK2HYVKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMZTGU3DEMBUHE . You are receiving this because you authored the thread.Message ID: @.***>

reskume commented 1 month ago

I pushed an updated version of the file that includes several fixes to geometries that had self-intersections. I also added a JS script to validate the openair text file using the openAIP openair parser. This requires the package.json etc to be able to easily install the required dependencies to the folder.

Regarding the RMZ: we mapped UNCLASSIFIED to G as defined in the official AIP AD. I re-added the previously removed RMZ for Le Havre.

@llauner The configuration for the parser should be adjusted to match the one you are using to test your builds. I simply copied over the one from our README file.

llauner commented 1 month ago

I'll look at iy over the wee-end. Regarding the parser, I have github action checking out this parser: https://github.com/openAIP/openaip-openair-parser.git I think it's better to check out the file rather than copying it locally...isn't that the same parser as the one you're using ?

reskume commented 1 month ago

Yes, the parser is the same. But I didn't know the exact configuration that you are using in the Github actions. So I left it on the default from our repository. I added the parser because it adds a convenient way to validate the openAIR file locally when by running the script validate-openairdefined in package.json. Otherwise each contributor would have to download the parser individually and then configure it depending on your config (which isn't documented anywhere I suppose?). If not using the parser, the validation process only takes place when there is a PR. And then issues have to be fixed by "pushing->waiting->reviewing check output->fix->push->wait..." which is quite tedious to do. But feel free to remove it if you think it doesn't fit in the repo.

llauner commented 1 month ago

I was trying to test your file with SeeYou, SeeYou Navigator and XCSoar, and I had an error with XCSoar: image

image

reskume commented 1 month ago

Typo is fixed

llauner commented 1 month ago

Thanks ! I'm not sure about the FIR region. There are several reasons for that:

reskume commented 1 month ago

I can't comment on how useful they are in France, but for Germany, they are very useful. Especially when you are around restricted areas that may be deactivated during specific times. Normally, the frequency associated with the airspace is the one that can be called and knows about activation hours. I just noticed that we didn't add additional AF command to them. I can try to find the specific frequency for each FIR - this would add more useful context to it. If this is not possible, I admit that they are of minimal use and can be removed.

llauner commented 1 month ago

I think for France we should remove them as they're not really useful. Is there any way that you could also put back all of the airspaces that you considered obsolete as we were past the expiration date ? Most of them are actually still active as they have been transformed into permanent area. My home airfield for instance: RMZ in LFLG - Grenoble Le Versoud And then I'll go through them and delete the ones that really are obsolete. Otherwise there are missing airpsaces...

reskume commented 1 month ago

I added the missing airspaces back in from the current version. THere are two TMAs that I cannot find in the AIP:

TMA AGEN Partie 1 (29 NOV 2001) TMA AGEN Partie 2 (29 NOV 2001)

I suppose they have been removed but they are still in the current version of the file (not in the PR).