search

Jusas / astrometry-api-lite

Astrometry.net API lite
MIT License
15 stars 4 forks source link

Not working with recent Kstars #8

Open papaf76 opened 3 years ago

papaf76 commented 3 years ago

Hi, not really knowing if this is still maintained or not, but sadly this project no longer works, as well as ANSVR and any other nova clone I've found. What's missing, I think, is really trivial: the plate-solve called is not working as it's now mandatory to pass the source image height and width in pixels. Kstars also expects the solved image height and width in arcsecs passed in the result. If these two things could be implemented, I'm fairly sure this would be working once again. It would be good for me as I'm trying to use this for my astro group. Many thanks!

Jusas commented 3 years ago

Hi @papaf76

This project is pretty dead, as the alternatives like ASTAP or locally installed astrometry.net exist - it's basically a compatibility layer, which hasn't had much use in the end. Since it's been 3 years since the last release, I can imagine a lot has changed, and Python2 -> Python3 change is also something the install should account for. The Windows installer is probably also outdated, since there's the WSL -> WSL2 change. This tool is pretty much deprecated by now.

However I'm still not going to leave anyone hanging if the amount of work is reasonable. Can you tell me a bit more about your use case? Mainly:

I'm still on the fence whether to keep supporting this project or not, as my time is limited - then again a crunch like this would at least breath some new life to it and give you and others some time to look for other options.

By the way, I'm currently working on a completely new solver that is based on the ASTAP solver algorithm, called WatneyAstrometry, which is currently in pre-release stage (first testable version available). I haven't made any noise about it yet since there are still small things I want to address, and I'll probably need to add a compatibility layer to allow using it via astrometry.net and/or ASTAP interfaces in the beginning to encourage its adoption - but the idea is simple: it is a solver that has no outside dependencies and is therefore easy to adopt, install and use and runs on Windows, Linux and OSX. It has been my pet project for the last year, which explains why I haven't paid much attention to any of my other projects ;)

papaf76 commented 3 years ago

Hi Jusas, first of all, the name is Fabio so at least we know each other! :) And also, thanks for taking the time to answer me, it's greatly appreciated! The use case is small, to be honest. And yet, it's something I'm kind of puzzled it's not needed by the mass. Basically, when everything is working, local astrometry or astap or even the newly internal solver Kstars has it's enough. But sometimes, especially since we're all running raspberries now, you need to solve an image blindly, and it takes forever on those small PCs. Hell, sometimes they even exceed the timeout. So my idea was to setup an astrometry clone at home, running on a decently powered machine and use the remote function at the telescope. Now, I tried to follow the instructions at nova.astrometry.net but they're outdated and sometimes outright wrong. I tried ANSVR but that's dead and incompatible too. Yours was promising but it has the same issue ANSVR has. I'm not really sure when those requirements appeared. Your software is able to get the request from kstars but the solve-field prog called fails because of those new requirements. Also, since I succeeded in modifying ANSVR to accomodate those two parameters, I also know kstars must have changed too, as it requires image height and width in arcsecs as well. The returned field size is present in the solving summary, so I'm guessing it's possible to obtain those. Feel free to obviously not do any of this, I realize it's a bit of a margin case.

Thanks again!

Il giorno mar 3 ago 2021 alle ore 14:03 Jussi Saarivirta < @.***> ha scritto:

Hi @papaf76 https://github.com/papaf76

This project is pretty dead, as the alternatives like ASTAP or locally installed astrometry.net exist - it's basically a compatibility layer, which hasn't had much use in the end. Since it's been 3 years since the last release, I can imagine a lot has changed, and Python2 -> Python3 change is also something the install should account for. The Windows installer is probably also outdated, since there's the WSL -> WSL2 change. This tool is pretty much deprecated by now.

However I'm still not going to leave anyone hanging if the amount of work is reasonable. Can you tell me a bit more about your use case? Mainly:

  • What OS are you running it on? (assuming Windows, since KStars has native astrometry.net support on Linux)
  • Which version of astrometry.net are we talking about?
  • Is this behavior (requiring image dimensions as parameters) a new requirement by astrometry.net, or just that the new KStars pushes these parameters to astrometry-api-lite, and it can't cope with unknown parameters?
  • Is the returned field size a new feature in astrometry.net?

I'm still on the fence whether to keep supporting this project or not, as my time is limited - then again a crunch like this would at least breath some new life to it and give you and others some time to look for other options.

By the way, I'm currently working on a completely new solver that is based on the ASTAP solver algorithm, called WatneyAstrometry https://github.com/Jusas/WatneyAstrometry, which is currently in pre-release stage (first testable version available). I haven't made any noise about it yet since there are still small things I want to address, and I'll probably need to add a compatibility layer to allow using it via astrometry.net and/or ASTAP interfaces in the beginning to encourage its adoption - but the idea is simple: it is a solver that has no outside dependencies and is therefore easy to adopt, install and use and runs on Windows, Linux and OSX. It has been my pet project for the last year, which explains why I haven't paid much attention to any of my other projects ;)

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Jusas/astrometry-api-lite/issues/8#issuecomment-891790032, or unsubscribe https://github.com/notifications/unsubscribe-auth/AMEK7CHYKYH5IUDNXK57HA3T27LH5ANCNFSM5BODKEKA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email .

Jusas commented 2 years ago

Hi Fabio,

I know it's been a very long time :) but I thought you might want to know about my project that I've just now fully released, the Watney Astrometric Solver. It's a whole new astrometric solver, based on the ASTAP algorithm. You can check it out here: https://github.com/Jusas/WatneyAstrometry

It's a C# library with both a command-line interface and a REST API, which includes Astrometry.net compatibility API. Pre-built binaries are available for Windows, Linux (arm, arm64 and x64) and OSX. I've run a few live test runs using KStars/Ekos with the compatibility API and didn't have any notable issues. This might be something interesting to try if you're still looking for options.