5 ms·

> the data sent back would be sitting on an FTP server somewhere for anyone to download and use I'm sure they could actually do that without too much fuss. But

> the data sent back would be sitting on an FTP server somewhere for anyone to download and use I'm sure they could actually do that without too much fuss. But it would require significant amounts of scientist time to document those datasets to enable others to use them for any arbitrary dataset. I'm sure we'll see fully open data sets from JWST appear, but lots of the stuff it collects isn't going to be interesting enough that it's reasonable to spend scientist time documenting it.

londons_explore 5y ago
I think there needs to be a distinction between something that a project "publishes", and something that is "made available". Something published has been checked by a few team members, written with care, and represents the opinion of the authors and project. Something made available has no guarantees of correctness, might not represent the projects opinion, and might just be random matlab scripts made by a JWST scientist in their lunchtime that they thought was fun. In the open source world, what is 'published' is probably the projects homepage, and code. What is 'made available' is random chatter on their discord or IRC channel. I hope that more government projects 'make available' everything done by all the workers - every file saved on every PC, with the understanding that there is no guarantee of correctness. I guess it's the same idea as being able to see into the kitchen from a restaurant. You might see the chef making mistakes or juggling the saucepans, but you'll also see the work being done as it's done, and being able to view doesn't delay the chefs work.
- dividedbyzero 5y ago
  That's a pretty bad idea IMO. Putting people in a panopticon has a strong chilling effect, no matter what disclaimer you put on the recordings. Creative, deep work needs space to make blunders in private, scientists are no exception. They'll just use their personal laptops or document every experiment and mistake and script to death, getting done a lot less actual research. Plus, it will be pretty much useless im practice since you'd have to be an expert in that niche yourself to know what's correct (you're not getting any extra docs or context) and probably most of it will be some kind of incorrect, possibly very subtly. The only people who could profit tremendously are the competition who aim to snipe that particular paper. Science is pretty dirty and ruthless often as not, I totally could see this happen.
jandrese 5y ago
It seems like it should all be automated. Some scientist generates a mission request for the JWST techs. If accepted the mission is added to the timeline with all of the metadata the original scientists had in their proposal. Stuff like the area being imaged, the sensors in use, duration of capture, etc... Once the data is collected and downloaded it is added to the catalog with all of that metadata attached. Then it's a matter of opening up that catalog to the public, although I'm guessing the downloads will be quite sizeable so the bandwidth could be an issue. The trick to making this work is to integrate the publishing into the workflow so it doesn't require any additional effort on the part of anyone.
- xioxox 5y ago
  The data from the mission will be made public after proprietary periods [1]. They have an archive [2]. I don't expect that the raw telemetry will be made available, but the raw science data in FITS format appears will be available. [1] https://jwst-docs.stsci.edu/jwst-opportunities-and-policies/jwst-general-science-policies/nasa-smd-policies-and-guidelines-for-the-operations-of-jwst-at-stsci/policy-2-data-rights-and-data-dissemination https://jwst-docs.stsci.edu/jwst-opportunities-and-policies/... [2] https://archive.stsci.edu/missions-and-data/jwst https://archive.stsci.edu/missions-and-data/jwst

londons_explore 5y ago

I think there needs to be a distinction between something that a project "publishes", and something that is "made available". Something published has been checked by a few team members, written with care, and represents the opinion of the authors and project. Something made available has no guarantees of correctness, might not represent the projects opinion, and might just be random matlab scripts made by a JWST scientist in their lunchtime that they thought was fun. In the open source world, what is 'published' is probably the projects homepage, and code. What is 'made available' is random chatter on their discord or IRC channel. I hope that more government projects 'make available' everything done by all the workers - every file saved on every PC, with the understanding that there is no guarantee of correctness. I guess it's the same idea as being able to see into the kitchen from a restaurant. You might see the chef making mistakes or juggling the saucepans, but you'll also see the work being done as it's done, and being able to view doesn't delay the chefs work.

dividedbyzero 5y ago
That's a pretty bad idea IMO. Putting people in a panopticon has a strong chilling effect, no matter what disclaimer you put on the recordings. Creative, deep work needs space to make blunders in private, scientists are no exception. They'll just use their personal laptops or document every experiment and mistake and script to death, getting done a lot less actual research. Plus, it will be pretty much useless im practice since you'd have to be an expert in that niche yourself to know what's correct (you're not getting any extra docs or context) and probably most of it will be some kind of incorrect, possibly very subtly. The only people who could profit tremendously are the competition who aim to snipe that particular paper. Science is pretty dirty and ruthless often as not, I totally could see this happen.

jandrese 5y ago

It seems like it should all be automated. Some scientist generates a mission request for the JWST techs. If accepted the mission is added to the timeline with all of the metadata the original scientists had in their proposal. Stuff like the area being imaged, the sensors in use, duration of capture, etc... Once the data is collected and downloaded it is added to the catalog with all of that metadata attached. Then it's a matter of opening up that catalog to the public, although I'm guessing the downloads will be quite sizeable so the bandwidth could be an issue. The trick to making this work is to integrate the publishing into the workflow so it doesn't require any additional effort on the part of anyone.

xioxox 5y ago
The data from the mission will be made public after proprietary periods [1]. They have an archive [2]. I don't expect that the raw telemetry will be made available, but the raw science data in FITS format appears will be available. [1] https://jwst-docs.stsci.edu/jwst-opportunities-and-policies/jwst-general-science-policies/nasa-smd-policies-and-guidelines-for-the-operations-of-jwst-at-stsci/policy-2-data-rights-and-data-dissemination https://jwst-docs.stsci.edu/jwst-opportunities-and-policies/... [2] https://archive.stsci.edu/missions-and-data/jwst https://archive.stsci.edu/missions-and-data/jwst